{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.149157404087487,
  "eval_steps": 500,
  "global_step": 1500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0014342058085335247,
      "grad_norm": 0.470686674118042,
      "learning_rate": 0.0001,
      "loss": 2.2273,
      "mean_token_accuracy": 0.570575475692749,
      "num_tokens": 2667.0,
      "step": 1
    },
    {
      "epoch": 0.014342058085335245,
      "grad_norm": 0.6386922001838684,
      "learning_rate": 9.961740793878528e-05,
      "loss": 3.3587,
      "mean_token_accuracy": 0.4295877367258072,
      "num_tokens": 18898.0,
      "step": 10
    },
    {
      "epoch": 0.02868411617067049,
      "grad_norm": 1.405023217201233,
      "learning_rate": 9.913916786226685e-05,
      "loss": 3.3664,
      "mean_token_accuracy": 0.43326892629265784,
      "num_tokens": 33358.0,
      "step": 20
    },
    {
      "epoch": 0.04302617425600574,
      "grad_norm": 1.0480499267578125,
      "learning_rate": 9.870875179340028e-05,
      "loss": 3.4708,
      "mean_token_accuracy": 0.3982369378209114,
      "num_tokens": 46959.0,
      "step": 30
    },
    {
      "epoch": 0.05736823234134098,
      "grad_norm": 1.2611873149871826,
      "learning_rate": 9.823051171688188e-05,
      "loss": 2.7339,
      "mean_token_accuracy": 0.4746448712423444,
      "num_tokens": 60421.0,
      "step": 40
    },
    {
      "epoch": 0.07171029042667623,
      "grad_norm": 1.2783650159835815,
      "learning_rate": 9.775227164036346e-05,
      "loss": 2.941,
      "mean_token_accuracy": 0.45884586926549675,
      "num_tokens": 76960.0,
      "step": 50
    },
    {
      "epoch": 0.08605234851201148,
      "grad_norm": 10.735796928405762,
      "learning_rate": 9.73218555714969e-05,
      "loss": 2.7924,
      "mean_token_accuracy": 0.4820432551205158,
      "num_tokens": 92542.0,
      "step": 60
    },
    {
      "epoch": 0.10039440659734672,
      "grad_norm": 1.3800513744354248,
      "learning_rate": 9.684361549497848e-05,
      "loss": 2.481,
      "mean_token_accuracy": 0.51774717643857,
      "num_tokens": 107955.0,
      "step": 70
    },
    {
      "epoch": 0.11473646468268196,
      "grad_norm": 2.344276189804077,
      "learning_rate": 9.636537541846008e-05,
      "loss": 2.613,
      "mean_token_accuracy": 0.49135800525546075,
      "num_tokens": 124083.0,
      "step": 80
    },
    {
      "epoch": 0.1290785227680172,
      "grad_norm": 6.935220241546631,
      "learning_rate": 9.588713534194166e-05,
      "loss": 2.6713,
      "mean_token_accuracy": 0.49849507249891756,
      "num_tokens": 136850.0,
      "step": 90
    },
    {
      "epoch": 0.14342058085335246,
      "grad_norm": 2.38398814201355,
      "learning_rate": 9.540889526542324e-05,
      "loss": 2.3803,
      "mean_token_accuracy": 0.5323172301054001,
      "num_tokens": 151617.0,
      "step": 100
    },
    {
      "epoch": 0.1577626389386877,
      "grad_norm": 2.4780445098876953,
      "learning_rate": 9.493065518890484e-05,
      "loss": 2.2107,
      "mean_token_accuracy": 0.5437798015773296,
      "num_tokens": 170078.0,
      "step": 110
    },
    {
      "epoch": 0.17210469702402295,
      "grad_norm": 1.6531819105148315,
      "learning_rate": 9.445241511238642e-05,
      "loss": 2.1105,
      "mean_token_accuracy": 0.5754976846277714,
      "num_tokens": 184960.0,
      "step": 120
    },
    {
      "epoch": 0.18644675510935818,
      "grad_norm": 9.240232467651367,
      "learning_rate": 9.397417503586801e-05,
      "loss": 2.3352,
      "mean_token_accuracy": 0.5454905904829502,
      "num_tokens": 198141.0,
      "step": 130
    },
    {
      "epoch": 0.20078881319469344,
      "grad_norm": 1.601082444190979,
      "learning_rate": 9.349593495934959e-05,
      "loss": 2.0366,
      "mean_token_accuracy": 0.598464447259903,
      "num_tokens": 213818.0,
      "step": 140
    },
    {
      "epoch": 0.2151308712800287,
      "grad_norm": 1.8371312618255615,
      "learning_rate": 9.301769488283119e-05,
      "loss": 2.2671,
      "mean_token_accuracy": 0.5460132732987404,
      "num_tokens": 228452.0,
      "step": 150
    },
    {
      "epoch": 0.22947292936536393,
      "grad_norm": 1.5072098970413208,
      "learning_rate": 9.253945480631277e-05,
      "loss": 2.0947,
      "mean_token_accuracy": 0.5814152680337429,
      "num_tokens": 242744.0,
      "step": 160
    },
    {
      "epoch": 0.24381498745069918,
      "grad_norm": 2.21504807472229,
      "learning_rate": 9.206121472979437e-05,
      "loss": 1.8873,
      "mean_token_accuracy": 0.6064566381275653,
      "num_tokens": 257036.0,
      "step": 170
    },
    {
      "epoch": 0.2581570455360344,
      "grad_norm": 1.861096739768982,
      "learning_rate": 9.158297465327594e-05,
      "loss": 2.1033,
      "mean_token_accuracy": 0.5820611655712128,
      "num_tokens": 270889.0,
      "step": 180
    },
    {
      "epoch": 0.27249910362136964,
      "grad_norm": 1.8603662252426147,
      "learning_rate": 9.110473457675754e-05,
      "loss": 2.1123,
      "mean_token_accuracy": 0.58076161891222,
      "num_tokens": 286076.0,
      "step": 190
    },
    {
      "epoch": 0.28684116170670493,
      "grad_norm": 1.733254075050354,
      "learning_rate": 9.062649450023912e-05,
      "loss": 1.8983,
      "mean_token_accuracy": 0.6083198614418507,
      "num_tokens": 303708.0,
      "step": 200
    },
    {
      "epoch": 0.30118321979204016,
      "grad_norm": 2.6913516521453857,
      "learning_rate": 9.014825442372072e-05,
      "loss": 2.0216,
      "mean_token_accuracy": 0.5970883950591087,
      "num_tokens": 319821.0,
      "step": 210
    },
    {
      "epoch": 0.3155252778773754,
      "grad_norm": 1.7351229190826416,
      "learning_rate": 8.96700143472023e-05,
      "loss": 1.8032,
      "mean_token_accuracy": 0.6248559698462486,
      "num_tokens": 339102.0,
      "step": 220
    },
    {
      "epoch": 0.3298673359627107,
      "grad_norm": 1.5186843872070312,
      "learning_rate": 8.91917742706839e-05,
      "loss": 1.7491,
      "mean_token_accuracy": 0.6350872039794921,
      "num_tokens": 355472.0,
      "step": 230
    },
    {
      "epoch": 0.3442093940480459,
      "grad_norm": 1.639203429222107,
      "learning_rate": 8.871353419416547e-05,
      "loss": 1.6937,
      "mean_token_accuracy": 0.6430230528116226,
      "num_tokens": 370676.0,
      "step": 240
    },
    {
      "epoch": 0.35855145213338113,
      "grad_norm": 3.4076974391937256,
      "learning_rate": 8.823529411764706e-05,
      "loss": 1.8259,
      "mean_token_accuracy": 0.6324844613671303,
      "num_tokens": 387764.0,
      "step": 250
    },
    {
      "epoch": 0.37289351021871636,
      "grad_norm": 1.737406611442566,
      "learning_rate": 8.775705404112865e-05,
      "loss": 1.8649,
      "mean_token_accuracy": 0.6135493002831935,
      "num_tokens": 402937.0,
      "step": 260
    },
    {
      "epoch": 0.38723556830405165,
      "grad_norm": 2.120741605758667,
      "learning_rate": 8.727881396461025e-05,
      "loss": 1.7671,
      "mean_token_accuracy": 0.6292251981794834,
      "num_tokens": 418228.0,
      "step": 270
    },
    {
      "epoch": 0.4015776263893869,
      "grad_norm": 1.5619535446166992,
      "learning_rate": 8.680057388809183e-05,
      "loss": 1.8934,
      "mean_token_accuracy": 0.6246664986014366,
      "num_tokens": 432792.0,
      "step": 280
    },
    {
      "epoch": 0.4159196844747221,
      "grad_norm": 6.643039703369141,
      "learning_rate": 8.632233381157341e-05,
      "loss": 1.7215,
      "mean_token_accuracy": 0.6430931359529495,
      "num_tokens": 450508.0,
      "step": 290
    },
    {
      "epoch": 0.4302617425600574,
      "grad_norm": 3.8804564476013184,
      "learning_rate": 8.5844093735055e-05,
      "loss": 1.7634,
      "mean_token_accuracy": 0.63612090498209,
      "num_tokens": 463939.0,
      "step": 300
    },
    {
      "epoch": 0.4446038006453926,
      "grad_norm": 2.418219566345215,
      "learning_rate": 8.53658536585366e-05,
      "loss": 1.6666,
      "mean_token_accuracy": 0.6541348457336426,
      "num_tokens": 481126.0,
      "step": 310
    },
    {
      "epoch": 0.45894585873072785,
      "grad_norm": 2.0463144779205322,
      "learning_rate": 8.488761358201818e-05,
      "loss": 1.7649,
      "mean_token_accuracy": 0.6427375547587871,
      "num_tokens": 496113.0,
      "step": 320
    },
    {
      "epoch": 0.4732879168160631,
      "grad_norm": 3.4844088554382324,
      "learning_rate": 8.440937350549976e-05,
      "loss": 1.6673,
      "mean_token_accuracy": 0.6421275071799755,
      "num_tokens": 513793.0,
      "step": 330
    },
    {
      "epoch": 0.48762997490139837,
      "grad_norm": 2.4026458263397217,
      "learning_rate": 8.393113342898136e-05,
      "loss": 1.721,
      "mean_token_accuracy": 0.6389043562114238,
      "num_tokens": 529289.0,
      "step": 340
    },
    {
      "epoch": 0.5019720329867337,
      "grad_norm": 1.8850486278533936,
      "learning_rate": 8.345289335246294e-05,
      "loss": 1.8284,
      "mean_token_accuracy": 0.6292856223881245,
      "num_tokens": 545136.0,
      "step": 350
    },
    {
      "epoch": 0.5163140910720688,
      "grad_norm": 3.17570161819458,
      "learning_rate": 8.297465327594453e-05,
      "loss": 1.7378,
      "mean_token_accuracy": 0.6412064053118229,
      "num_tokens": 562947.0,
      "step": 360
    },
    {
      "epoch": 0.5306561491574041,
      "grad_norm": 2.047070026397705,
      "learning_rate": 8.249641319942611e-05,
      "loss": 1.6242,
      "mean_token_accuracy": 0.6595074772834778,
      "num_tokens": 577017.0,
      "step": 370
    },
    {
      "epoch": 0.5449982072427393,
      "grad_norm": 2.9317102432250977,
      "learning_rate": 8.20181731229077e-05,
      "loss": 1.6056,
      "mean_token_accuracy": 0.6700303673744201,
      "num_tokens": 592786.0,
      "step": 380
    },
    {
      "epoch": 0.5593402653280746,
      "grad_norm": 2.8381388187408447,
      "learning_rate": 8.153993304638929e-05,
      "loss": 1.6943,
      "mean_token_accuracy": 0.6448078341782093,
      "num_tokens": 609927.0,
      "step": 390
    },
    {
      "epoch": 0.5736823234134099,
      "grad_norm": 2.2730963230133057,
      "learning_rate": 8.106169296987089e-05,
      "loss": 1.5817,
      "mean_token_accuracy": 0.6591131195425988,
      "num_tokens": 625711.0,
      "step": 400
    },
    {
      "epoch": 0.588024381498745,
      "grad_norm": 2.802509307861328,
      "learning_rate": 8.058345289335246e-05,
      "loss": 1.6625,
      "mean_token_accuracy": 0.6519417136907577,
      "num_tokens": 642904.0,
      "step": 410
    },
    {
      "epoch": 0.6023664395840803,
      "grad_norm": 2.065239429473877,
      "learning_rate": 8.010521281683405e-05,
      "loss": 1.6026,
      "mean_token_accuracy": 0.656041894853115,
      "num_tokens": 659332.0,
      "step": 420
    },
    {
      "epoch": 0.6167084976694156,
      "grad_norm": 1.4606506824493408,
      "learning_rate": 7.962697274031564e-05,
      "loss": 1.5676,
      "mean_token_accuracy": 0.6626035153865815,
      "num_tokens": 675202.0,
      "step": 430
    },
    {
      "epoch": 0.6310505557547508,
      "grad_norm": 4.072043418884277,
      "learning_rate": 7.914873266379723e-05,
      "loss": 1.6388,
      "mean_token_accuracy": 0.6583093114197254,
      "num_tokens": 690023.0,
      "step": 440
    },
    {
      "epoch": 0.6453926138400861,
      "grad_norm": 1.800016164779663,
      "learning_rate": 7.867049258727882e-05,
      "loss": 1.8092,
      "mean_token_accuracy": 0.6206394098699093,
      "num_tokens": 706376.0,
      "step": 450
    },
    {
      "epoch": 0.6597346719254213,
      "grad_norm": 3.809736490249634,
      "learning_rate": 7.819225251076042e-05,
      "loss": 1.5875,
      "mean_token_accuracy": 0.6476077564060688,
      "num_tokens": 721324.0,
      "step": 460
    },
    {
      "epoch": 0.6740767300107565,
      "grad_norm": 2.1231305599212646,
      "learning_rate": 7.771401243424199e-05,
      "loss": 1.6035,
      "mean_token_accuracy": 0.6592143125832081,
      "num_tokens": 738919.0,
      "step": 470
    },
    {
      "epoch": 0.6884187880960918,
      "grad_norm": 2.784208297729492,
      "learning_rate": 7.723577235772358e-05,
      "loss": 1.6384,
      "mean_token_accuracy": 0.6568839088082313,
      "num_tokens": 754580.0,
      "step": 480
    },
    {
      "epoch": 0.7027608461814271,
      "grad_norm": 1.5424902439117432,
      "learning_rate": 7.675753228120517e-05,
      "loss": 1.5661,
      "mean_token_accuracy": 0.6690633282065391,
      "num_tokens": 769337.0,
      "step": 490
    },
    {
      "epoch": 0.7171029042667623,
      "grad_norm": 2.4127919673919678,
      "learning_rate": 7.627929220468676e-05,
      "loss": 1.7273,
      "mean_token_accuracy": 0.6398490175604821,
      "num_tokens": 784894.0,
      "step": 500
    },
    {
      "epoch": 0.7314449623520975,
      "grad_norm": 1.8765316009521484,
      "learning_rate": 7.580105212816835e-05,
      "loss": 1.72,
      "mean_token_accuracy": 0.644102756679058,
      "num_tokens": 800931.0,
      "step": 510
    },
    {
      "epoch": 0.7457870204374327,
      "grad_norm": 1.97030508518219,
      "learning_rate": 7.532281205164993e-05,
      "loss": 1.6752,
      "mean_token_accuracy": 0.6556637264788151,
      "num_tokens": 816581.0,
      "step": 520
    },
    {
      "epoch": 0.760129078522768,
      "grad_norm": 2.5584030151367188,
      "learning_rate": 7.484457197513151e-05,
      "loss": 1.6071,
      "mean_token_accuracy": 0.6555875569581986,
      "num_tokens": 832985.0,
      "step": 530
    },
    {
      "epoch": 0.7744711366081033,
      "grad_norm": 2.1065962314605713,
      "learning_rate": 7.436633189861311e-05,
      "loss": 1.6353,
      "mean_token_accuracy": 0.6672606863081455,
      "num_tokens": 848592.0,
      "step": 540
    },
    {
      "epoch": 0.7888131946934385,
      "grad_norm": 2.460101366043091,
      "learning_rate": 7.38880918220947e-05,
      "loss": 1.6701,
      "mean_token_accuracy": 0.6500819236040115,
      "num_tokens": 862542.0,
      "step": 550
    },
    {
      "epoch": 0.8031552527787738,
      "grad_norm": 1.6919441223144531,
      "learning_rate": 7.340985174557629e-05,
      "loss": 1.6249,
      "mean_token_accuracy": 0.6479768246412277,
      "num_tokens": 877457.0,
      "step": 560
    },
    {
      "epoch": 0.817497310864109,
      "grad_norm": 1.7528458833694458,
      "learning_rate": 7.293161166905788e-05,
      "loss": 1.5832,
      "mean_token_accuracy": 0.6721912071108818,
      "num_tokens": 892474.0,
      "step": 570
    },
    {
      "epoch": 0.8318393689494442,
      "grad_norm": 2.1731760501861572,
      "learning_rate": 7.245337159253946e-05,
      "loss": 1.5953,
      "mean_token_accuracy": 0.6713483422994614,
      "num_tokens": 910071.0,
      "step": 580
    },
    {
      "epoch": 0.8461814270347795,
      "grad_norm": 1.6760331392288208,
      "learning_rate": 7.197513151602104e-05,
      "loss": 1.647,
      "mean_token_accuracy": 0.6591413423418999,
      "num_tokens": 927445.0,
      "step": 590
    },
    {
      "epoch": 0.8605234851201148,
      "grad_norm": 1.4302948713302612,
      "learning_rate": 7.149689143950264e-05,
      "loss": 1.6653,
      "mean_token_accuracy": 0.6485698565840721,
      "num_tokens": 942011.0,
      "step": 600
    },
    {
      "epoch": 0.87486554320545,
      "grad_norm": 2.2263236045837402,
      "learning_rate": 7.101865136298422e-05,
      "loss": 1.4487,
      "mean_token_accuracy": 0.6805191606283187,
      "num_tokens": 957309.0,
      "step": 610
    },
    {
      "epoch": 0.8892076012907852,
      "grad_norm": 2.6513445377349854,
      "learning_rate": 7.054041128646581e-05,
      "loss": 1.661,
      "mean_token_accuracy": 0.6481011644005775,
      "num_tokens": 970774.0,
      "step": 620
    },
    {
      "epoch": 0.9035496593761205,
      "grad_norm": 1.9575634002685547,
      "learning_rate": 7.006217120994739e-05,
      "loss": 1.5155,
      "mean_token_accuracy": 0.6912697181105614,
      "num_tokens": 985630.0,
      "step": 630
    },
    {
      "epoch": 0.9178917174614557,
      "grad_norm": 1.7692954540252686,
      "learning_rate": 6.958393113342897e-05,
      "loss": 1.5249,
      "mean_token_accuracy": 0.6790330305695533,
      "num_tokens": 1001557.0,
      "step": 640
    },
    {
      "epoch": 0.932233775546791,
      "grad_norm": 1.6159223318099976,
      "learning_rate": 6.910569105691057e-05,
      "loss": 1.4862,
      "mean_token_accuracy": 0.6802081070840359,
      "num_tokens": 1017875.0,
      "step": 650
    },
    {
      "epoch": 0.9465758336321262,
      "grad_norm": 1.7482064962387085,
      "learning_rate": 6.862745098039216e-05,
      "loss": 1.7641,
      "mean_token_accuracy": 0.6506462752819061,
      "num_tokens": 1035082.0,
      "step": 660
    },
    {
      "epoch": 0.9609178917174614,
      "grad_norm": 1.83534574508667,
      "learning_rate": 6.814921090387375e-05,
      "loss": 1.5879,
      "mean_token_accuracy": 0.6688900426030159,
      "num_tokens": 1050136.0,
      "step": 670
    },
    {
      "epoch": 0.9752599498027967,
      "grad_norm": 4.526679515838623,
      "learning_rate": 6.767097082735534e-05,
      "loss": 1.6348,
      "mean_token_accuracy": 0.6631537288427353,
      "num_tokens": 1064297.0,
      "step": 680
    },
    {
      "epoch": 0.9896020078881319,
      "grad_norm": 3.217916965484619,
      "learning_rate": 6.719273075083692e-05,
      "loss": 1.6184,
      "mean_token_accuracy": 0.6580263905227184,
      "num_tokens": 1082302.0,
      "step": 690
    },
    {
      "epoch": 1.002868411617067,
      "grad_norm": 2.3952481746673584,
      "learning_rate": 6.67144906743185e-05,
      "loss": 1.552,
      "mean_token_accuracy": 0.6819982995858064,
      "num_tokens": 1095261.0,
      "step": 700
    },
    {
      "epoch": 1.0172104697024023,
      "grad_norm": 1.892841100692749,
      "learning_rate": 6.62362505978001e-05,
      "loss": 1.571,
      "mean_token_accuracy": 0.6693050697445869,
      "num_tokens": 1110422.0,
      "step": 710
    },
    {
      "epoch": 1.0315525277877375,
      "grad_norm": 22.51019287109375,
      "learning_rate": 6.575801052128168e-05,
      "loss": 1.5682,
      "mean_token_accuracy": 0.6653205692768097,
      "num_tokens": 1127096.0,
      "step": 720
    },
    {
      "epoch": 1.0458945858730728,
      "grad_norm": 2.4129638671875,
      "learning_rate": 6.527977044476328e-05,
      "loss": 1.5858,
      "mean_token_accuracy": 0.6521917231380939,
      "num_tokens": 1143137.0,
      "step": 730
    },
    {
      "epoch": 1.060236643958408,
      "grad_norm": 2.1772189140319824,
      "learning_rate": 6.480153036824486e-05,
      "loss": 1.6734,
      "mean_token_accuracy": 0.6441927455365658,
      "num_tokens": 1157904.0,
      "step": 740
    },
    {
      "epoch": 1.0745787020437434,
      "grad_norm": 1.4643489122390747,
      "learning_rate": 6.432329029172645e-05,
      "loss": 1.4623,
      "mean_token_accuracy": 0.6849726974964142,
      "num_tokens": 1176653.0,
      "step": 750
    },
    {
      "epoch": 1.0889207601290785,
      "grad_norm": 1.3659816980361938,
      "learning_rate": 6.384505021520803e-05,
      "loss": 1.434,
      "mean_token_accuracy": 0.6869873121380806,
      "num_tokens": 1193019.0,
      "step": 760
    },
    {
      "epoch": 1.1032628182144137,
      "grad_norm": 1.920989990234375,
      "learning_rate": 6.336681013868963e-05,
      "loss": 1.6151,
      "mean_token_accuracy": 0.6590917229652404,
      "num_tokens": 1209004.0,
      "step": 770
    },
    {
      "epoch": 1.117604876299749,
      "grad_norm": 2.5810189247131348,
      "learning_rate": 6.288857006217121e-05,
      "loss": 1.5681,
      "mean_token_accuracy": 0.658272746950388,
      "num_tokens": 1223332.0,
      "step": 780
    },
    {
      "epoch": 1.1319469343850843,
      "grad_norm": 2.319913864135742,
      "learning_rate": 6.241032998565281e-05,
      "loss": 1.4777,
      "mean_token_accuracy": 0.6888576149940491,
      "num_tokens": 1237845.0,
      "step": 790
    },
    {
      "epoch": 1.1462889924704196,
      "grad_norm": 3.982288122177124,
      "learning_rate": 6.193208990913438e-05,
      "loss": 1.5009,
      "mean_token_accuracy": 0.6694469384849071,
      "num_tokens": 1254783.0,
      "step": 800
    },
    {
      "epoch": 1.1606310505557547,
      "grad_norm": 1.5996657609939575,
      "learning_rate": 6.145384983261598e-05,
      "loss": 1.5317,
      "mean_token_accuracy": 0.6781374484300613,
      "num_tokens": 1271210.0,
      "step": 810
    },
    {
      "epoch": 1.17497310864109,
      "grad_norm": 3.356278419494629,
      "learning_rate": 6.097560975609756e-05,
      "loss": 1.5377,
      "mean_token_accuracy": 0.6668447598814964,
      "num_tokens": 1284772.0,
      "step": 820
    },
    {
      "epoch": 1.1893151667264252,
      "grad_norm": 1.802544355392456,
      "learning_rate": 6.049736967957915e-05,
      "loss": 1.5716,
      "mean_token_accuracy": 0.6638300716876984,
      "num_tokens": 1299024.0,
      "step": 830
    },
    {
      "epoch": 1.2036572248117605,
      "grad_norm": 2.500906467437744,
      "learning_rate": 6.001912960306074e-05,
      "loss": 1.5394,
      "mean_token_accuracy": 0.6792485311627388,
      "num_tokens": 1316230.0,
      "step": 840
    },
    {
      "epoch": 1.2179992828970958,
      "grad_norm": 3.2831778526306152,
      "learning_rate": 5.954088952654233e-05,
      "loss": 1.3411,
      "mean_token_accuracy": 0.7026472240686417,
      "num_tokens": 1335171.0,
      "step": 850
    },
    {
      "epoch": 1.232341340982431,
      "grad_norm": 1.632434368133545,
      "learning_rate": 5.906264945002391e-05,
      "loss": 1.6163,
      "mean_token_accuracy": 0.6708522118628025,
      "num_tokens": 1350786.0,
      "step": 860
    },
    {
      "epoch": 1.2466833990677662,
      "grad_norm": 2.128074884414673,
      "learning_rate": 5.85844093735055e-05,
      "loss": 1.5226,
      "mean_token_accuracy": 0.6910128638148307,
      "num_tokens": 1367664.0,
      "step": 870
    },
    {
      "epoch": 1.2610254571531014,
      "grad_norm": 1.786818265914917,
      "learning_rate": 5.810616929698709e-05,
      "loss": 1.4107,
      "mean_token_accuracy": 0.6914131015539169,
      "num_tokens": 1384317.0,
      "step": 880
    },
    {
      "epoch": 1.2753675152384367,
      "grad_norm": 2.589280128479004,
      "learning_rate": 5.762792922046868e-05,
      "loss": 1.4736,
      "mean_token_accuracy": 0.6800985768437385,
      "num_tokens": 1402238.0,
      "step": 890
    },
    {
      "epoch": 1.289709573323772,
      "grad_norm": 2.6926186084747314,
      "learning_rate": 5.714968914395027e-05,
      "loss": 1.6272,
      "mean_token_accuracy": 0.6593966111540794,
      "num_tokens": 1418578.0,
      "step": 900
    },
    {
      "epoch": 1.304051631409107,
      "grad_norm": 2.787445306777954,
      "learning_rate": 5.667144906743186e-05,
      "loss": 1.432,
      "mean_token_accuracy": 0.6899330474436283,
      "num_tokens": 1434941.0,
      "step": 910
    },
    {
      "epoch": 1.3183936894944424,
      "grad_norm": 5.517631530761719,
      "learning_rate": 5.619320899091344e-05,
      "loss": 1.5307,
      "mean_token_accuracy": 0.6874054931104183,
      "num_tokens": 1448665.0,
      "step": 920
    },
    {
      "epoch": 1.3327357475797776,
      "grad_norm": 1.9263485670089722,
      "learning_rate": 5.571496891439503e-05,
      "loss": 1.5261,
      "mean_token_accuracy": 0.679547868669033,
      "num_tokens": 1462120.0,
      "step": 930
    },
    {
      "epoch": 1.347077805665113,
      "grad_norm": 2.604401111602783,
      "learning_rate": 5.523672883787662e-05,
      "loss": 1.5099,
      "mean_token_accuracy": 0.6857024997472763,
      "num_tokens": 1476792.0,
      "step": 940
    },
    {
      "epoch": 1.3614198637504482,
      "grad_norm": 2.0061893463134766,
      "learning_rate": 5.480631276901005e-05,
      "loss": 1.6256,
      "mean_token_accuracy": 0.6685486875474453,
      "num_tokens": 1493863.0,
      "step": 950
    },
    {
      "epoch": 1.3757619218357835,
      "grad_norm": 1.8100831508636475,
      "learning_rate": 5.432807269249164e-05,
      "loss": 1.4529,
      "mean_token_accuracy": 0.6858629032969474,
      "num_tokens": 1510236.0,
      "step": 960
    },
    {
      "epoch": 1.3901039799211188,
      "grad_norm": 1.7307875156402588,
      "learning_rate": 5.3849832615973216e-05,
      "loss": 1.5551,
      "mean_token_accuracy": 0.6620718680322171,
      "num_tokens": 1526526.0,
      "step": 970
    },
    {
      "epoch": 1.4044460380064538,
      "grad_norm": 2.136469602584839,
      "learning_rate": 5.337159253945481e-05,
      "loss": 1.4498,
      "mean_token_accuracy": 0.6863499596714974,
      "num_tokens": 1541400.0,
      "step": 980
    },
    {
      "epoch": 1.4187880960917891,
      "grad_norm": 1.960593819618225,
      "learning_rate": 5.28933524629364e-05,
      "loss": 1.6216,
      "mean_token_accuracy": 0.6671703614294529,
      "num_tokens": 1555980.0,
      "step": 990
    },
    {
      "epoch": 1.4331301541771244,
      "grad_norm": 2.4239704608917236,
      "learning_rate": 5.241511238641799e-05,
      "loss": 1.5087,
      "mean_token_accuracy": 0.6855074353516102,
      "num_tokens": 1571350.0,
      "step": 1000
    },
    {
      "epoch": 1.4474722122624597,
      "grad_norm": 3.153437852859497,
      "learning_rate": 5.193687230989958e-05,
      "loss": 1.4422,
      "mean_token_accuracy": 0.7070553943514823,
      "num_tokens": 1585106.0,
      "step": 1010
    },
    {
      "epoch": 1.461814270347795,
      "grad_norm": 1.828374981880188,
      "learning_rate": 5.1458632233381155e-05,
      "loss": 1.5071,
      "mean_token_accuracy": 0.6775861948728561,
      "num_tokens": 1600582.0,
      "step": 1020
    },
    {
      "epoch": 1.47615632843313,
      "grad_norm": 4.680983543395996,
      "learning_rate": 5.0980392156862745e-05,
      "loss": 1.5504,
      "mean_token_accuracy": 0.6747892014682293,
      "num_tokens": 1614677.0,
      "step": 1030
    },
    {
      "epoch": 1.4904983865184653,
      "grad_norm": 2.083742380142212,
      "learning_rate": 5.0502152080344336e-05,
      "loss": 1.4545,
      "mean_token_accuracy": 0.6944699421525001,
      "num_tokens": 1630046.0,
      "step": 1040
    },
    {
      "epoch": 1.5048404446038006,
      "grad_norm": 3.8617031574249268,
      "learning_rate": 5.0023912003825926e-05,
      "loss": 1.5685,
      "mean_token_accuracy": 0.6765852883458138,
      "num_tokens": 1645660.0,
      "step": 1050
    },
    {
      "epoch": 1.519182502689136,
      "grad_norm": 2.4847116470336914,
      "learning_rate": 4.9545671927307516e-05,
      "loss": 1.5439,
      "mean_token_accuracy": 0.6794092074036598,
      "num_tokens": 1661080.0,
      "step": 1060
    },
    {
      "epoch": 1.5335245607744712,
      "grad_norm": 3.753831148147583,
      "learning_rate": 4.90674318507891e-05,
      "loss": 1.5337,
      "mean_token_accuracy": 0.6865640193223953,
      "num_tokens": 1675490.0,
      "step": 1070
    },
    {
      "epoch": 1.5478666188598065,
      "grad_norm": 1.5879663228988647,
      "learning_rate": 4.858919177427069e-05,
      "loss": 1.4808,
      "mean_token_accuracy": 0.681853985786438,
      "num_tokens": 1689533.0,
      "step": 1080
    },
    {
      "epoch": 1.5622086769451418,
      "grad_norm": 1.138118028640747,
      "learning_rate": 4.8110951697752274e-05,
      "loss": 1.4723,
      "mean_token_accuracy": 0.6909654997289181,
      "num_tokens": 1706915.0,
      "step": 1090
    },
    {
      "epoch": 1.5765507350304768,
      "grad_norm": 2.1534876823425293,
      "learning_rate": 4.763271162123386e-05,
      "loss": 1.6073,
      "mean_token_accuracy": 0.6727578081190586,
      "num_tokens": 1722527.0,
      "step": 1100
    },
    {
      "epoch": 1.590892793115812,
      "grad_norm": 1.9966998100280762,
      "learning_rate": 4.715447154471545e-05,
      "loss": 1.4986,
      "mean_token_accuracy": 0.676116967946291,
      "num_tokens": 1737987.0,
      "step": 1110
    },
    {
      "epoch": 1.6052348512011474,
      "grad_norm": 1.9551148414611816,
      "learning_rate": 4.667623146819703e-05,
      "loss": 1.4541,
      "mean_token_accuracy": 0.6881252631545067,
      "num_tokens": 1754778.0,
      "step": 1120
    },
    {
      "epoch": 1.6195769092864825,
      "grad_norm": 2.3044798374176025,
      "learning_rate": 4.619799139167862e-05,
      "loss": 1.5929,
      "mean_token_accuracy": 0.6656625166535377,
      "num_tokens": 1769795.0,
      "step": 1130
    },
    {
      "epoch": 1.6339189673718177,
      "grad_norm": 2.1620051860809326,
      "learning_rate": 4.571975131516021e-05,
      "loss": 1.5498,
      "mean_token_accuracy": 0.6724934101104736,
      "num_tokens": 1785990.0,
      "step": 1140
    },
    {
      "epoch": 1.648261025457153,
      "grad_norm": 1.2986624240875244,
      "learning_rate": 4.5241511238641796e-05,
      "loss": 1.4972,
      "mean_token_accuracy": 0.681456783413887,
      "num_tokens": 1804143.0,
      "step": 1150
    },
    {
      "epoch": 1.6626030835424883,
      "grad_norm": 1.6331572532653809,
      "learning_rate": 4.4763271162123386e-05,
      "loss": 1.5982,
      "mean_token_accuracy": 0.6560898676514626,
      "num_tokens": 1818754.0,
      "step": 1160
    },
    {
      "epoch": 1.6769451416278236,
      "grad_norm": 1.3731579780578613,
      "learning_rate": 4.428503108560498e-05,
      "loss": 1.4023,
      "mean_token_accuracy": 0.7019235521554947,
      "num_tokens": 1834913.0,
      "step": 1170
    },
    {
      "epoch": 1.6912871997131589,
      "grad_norm": 1.8848557472229004,
      "learning_rate": 4.380679100908656e-05,
      "loss": 1.4246,
      "mean_token_accuracy": 0.6995187073945999,
      "num_tokens": 1849201.0,
      "step": 1180
    },
    {
      "epoch": 1.7056292577984942,
      "grad_norm": 2.893299102783203,
      "learning_rate": 4.332855093256815e-05,
      "loss": 1.4975,
      "mean_token_accuracy": 0.6855712987482547,
      "num_tokens": 1863036.0,
      "step": 1190
    },
    {
      "epoch": 1.7199713158838295,
      "grad_norm": 2.4176976680755615,
      "learning_rate": 4.285031085604974e-05,
      "loss": 1.5163,
      "mean_token_accuracy": 0.6739328794181347,
      "num_tokens": 1878746.0,
      "step": 1200
    },
    {
      "epoch": 1.7343133739691647,
      "grad_norm": 2.116476535797119,
      "learning_rate": 4.2372070779531325e-05,
      "loss": 1.5439,
      "mean_token_accuracy": 0.6762977614998817,
      "num_tokens": 1892601.0,
      "step": 1210
    },
    {
      "epoch": 1.7486554320544998,
      "grad_norm": 2.4355766773223877,
      "learning_rate": 4.1893830703012915e-05,
      "loss": 1.6207,
      "mean_token_accuracy": 0.6575247406959533,
      "num_tokens": 1909258.0,
      "step": 1220
    },
    {
      "epoch": 1.762997490139835,
      "grad_norm": 2.1678593158721924,
      "learning_rate": 4.1415590626494505e-05,
      "loss": 1.5225,
      "mean_token_accuracy": 0.6775545470416546,
      "num_tokens": 1924766.0,
      "step": 1230
    },
    {
      "epoch": 1.7773395482251702,
      "grad_norm": 2.594438314437866,
      "learning_rate": 4.093735054997609e-05,
      "loss": 1.4556,
      "mean_token_accuracy": 0.7052863880991935,
      "num_tokens": 1942931.0,
      "step": 1240
    },
    {
      "epoch": 1.7916816063105054,
      "grad_norm": 1.8568722009658813,
      "learning_rate": 4.045911047345768e-05,
      "loss": 1.484,
      "mean_token_accuracy": 0.6940086185932159,
      "num_tokens": 1959918.0,
      "step": 1250
    },
    {
      "epoch": 1.8060236643958407,
      "grad_norm": 2.9484922885894775,
      "learning_rate": 3.998087039693926e-05,
      "loss": 1.5105,
      "mean_token_accuracy": 0.683978445827961,
      "num_tokens": 1974063.0,
      "step": 1260
    },
    {
      "epoch": 1.820365722481176,
      "grad_norm": 2.427079916000366,
      "learning_rate": 3.9502630320420853e-05,
      "loss": 1.4813,
      "mean_token_accuracy": 0.7070754542946815,
      "num_tokens": 1988748.0,
      "step": 1270
    },
    {
      "epoch": 1.8347077805665113,
      "grad_norm": 2.605332136154175,
      "learning_rate": 3.9024390243902444e-05,
      "loss": 1.4335,
      "mean_token_accuracy": 0.6803991563618184,
      "num_tokens": 2005256.0,
      "step": 1280
    },
    {
      "epoch": 1.8490498386518466,
      "grad_norm": 1.4314664602279663,
      "learning_rate": 3.854615016738403e-05,
      "loss": 1.4365,
      "mean_token_accuracy": 0.6869161948561668,
      "num_tokens": 2020151.0,
      "step": 1290
    },
    {
      "epoch": 1.8633918967371819,
      "grad_norm": 2.391542673110962,
      "learning_rate": 3.806791009086562e-05,
      "loss": 1.3465,
      "mean_token_accuracy": 0.6867123879492283,
      "num_tokens": 2038112.0,
      "step": 1300
    },
    {
      "epoch": 1.8777339548225171,
      "grad_norm": 1.8515162467956543,
      "learning_rate": 3.758967001434721e-05,
      "loss": 1.4549,
      "mean_token_accuracy": 0.6883729994297028,
      "num_tokens": 2053699.0,
      "step": 1310
    },
    {
      "epoch": 1.8920760129078524,
      "grad_norm": 1.989882230758667,
      "learning_rate": 3.711142993782879e-05,
      "loss": 1.4244,
      "mean_token_accuracy": 0.7050331614911556,
      "num_tokens": 2070663.0,
      "step": 1320
    },
    {
      "epoch": 1.9064180709931875,
      "grad_norm": 10.000530242919922,
      "learning_rate": 3.663318986131038e-05,
      "loss": 1.4198,
      "mean_token_accuracy": 0.6946797490119934,
      "num_tokens": 2086286.0,
      "step": 1330
    },
    {
      "epoch": 1.9207601290785228,
      "grad_norm": 3.4234516620635986,
      "learning_rate": 3.615494978479197e-05,
      "loss": 1.4259,
      "mean_token_accuracy": 0.700089217722416,
      "num_tokens": 2102045.0,
      "step": 1340
    },
    {
      "epoch": 1.935102187163858,
      "grad_norm": 2.789325714111328,
      "learning_rate": 3.5676709708273556e-05,
      "loss": 1.4597,
      "mean_token_accuracy": 0.6905134588479995,
      "num_tokens": 2115351.0,
      "step": 1350
    },
    {
      "epoch": 1.9494442452491931,
      "grad_norm": 2.9415955543518066,
      "learning_rate": 3.519846963175514e-05,
      "loss": 1.4747,
      "mean_token_accuracy": 0.6978569567203522,
      "num_tokens": 2131695.0,
      "step": 1360
    },
    {
      "epoch": 1.9637863033345284,
      "grad_norm": 2.0069663524627686,
      "learning_rate": 3.472022955523673e-05,
      "loss": 1.5132,
      "mean_token_accuracy": 0.6854235969483853,
      "num_tokens": 2146911.0,
      "step": 1370
    },
    {
      "epoch": 1.9781283614198637,
      "grad_norm": 2.3505399227142334,
      "learning_rate": 3.4241989478718314e-05,
      "loss": 1.3899,
      "mean_token_accuracy": 0.7020819112658501,
      "num_tokens": 2162164.0,
      "step": 1380
    },
    {
      "epoch": 1.992470419505199,
      "grad_norm": 1.4830210208892822,
      "learning_rate": 3.3763749402199904e-05,
      "loss": 1.5027,
      "mean_token_accuracy": 0.6788195565342903,
      "num_tokens": 2176995.0,
      "step": 1390
    },
    {
      "epoch": 2.005736823234134,
      "grad_norm": 2.140188217163086,
      "learning_rate": 3.328550932568149e-05,
      "loss": 1.4032,
      "mean_token_accuracy": 0.693047933481835,
      "num_tokens": 2190128.0,
      "step": 1400
    },
    {
      "epoch": 2.020078881319469,
      "grad_norm": 1.661476492881775,
      "learning_rate": 3.280726924916308e-05,
      "loss": 1.4841,
      "mean_token_accuracy": 0.6926379069685936,
      "num_tokens": 2205346.0,
      "step": 1410
    },
    {
      "epoch": 2.0344209394048045,
      "grad_norm": 2.564972400665283,
      "learning_rate": 3.232902917264467e-05,
      "loss": 1.5078,
      "mean_token_accuracy": 0.6939757362008094,
      "num_tokens": 2219381.0,
      "step": 1420
    },
    {
      "epoch": 2.04876299749014,
      "grad_norm": 4.097690582275391,
      "learning_rate": 3.185078909612625e-05,
      "loss": 1.5106,
      "mean_token_accuracy": 0.69179158359766,
      "num_tokens": 2236634.0,
      "step": 1430
    },
    {
      "epoch": 2.063105055575475,
      "grad_norm": 2.772319793701172,
      "learning_rate": 3.137254901960784e-05,
      "loss": 1.5549,
      "mean_token_accuracy": 0.6739653892815113,
      "num_tokens": 2253528.0,
      "step": 1440
    },
    {
      "epoch": 2.0774471136608104,
      "grad_norm": 1.8374364376068115,
      "learning_rate": 3.089430894308943e-05,
      "loss": 1.3455,
      "mean_token_accuracy": 0.7092796668410302,
      "num_tokens": 2268239.0,
      "step": 1450
    },
    {
      "epoch": 2.0917891717461456,
      "grad_norm": 2.8052825927734375,
      "learning_rate": 3.0416068866571017e-05,
      "loss": 1.424,
      "mean_token_accuracy": 0.6826408110558987,
      "num_tokens": 2285160.0,
      "step": 1460
    },
    {
      "epoch": 2.106131229831481,
      "grad_norm": 1.8539540767669678,
      "learning_rate": 2.9937828790052607e-05,
      "loss": 1.397,
      "mean_token_accuracy": 0.683707358688116,
      "num_tokens": 2302380.0,
      "step": 1470
    },
    {
      "epoch": 2.120473287916816,
      "grad_norm": 2.8811655044555664,
      "learning_rate": 2.9459588713534197e-05,
      "loss": 1.4091,
      "mean_token_accuracy": 0.7010952278971672,
      "num_tokens": 2316415.0,
      "step": 1480
    },
    {
      "epoch": 2.1348153460021515,
      "grad_norm": 1.6305341720581055,
      "learning_rate": 2.898134863701578e-05,
      "loss": 1.4301,
      "mean_token_accuracy": 0.6907539337873458,
      "num_tokens": 2331973.0,
      "step": 1490
    },
    {
      "epoch": 2.149157404087487,
      "grad_norm": 3.8295979499816895,
      "learning_rate": 2.850310856049737e-05,
      "loss": 1.4743,
      "mean_token_accuracy": 0.6858786374330521,
      "num_tokens": 2347285.0,
      "step": 1500
    }
  ],
  "logging_steps": 10,
  "max_steps": 2091,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4.058431818937344e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}