{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 510,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00984009840098401,
      "grad_norm": 4.0806973528031865,
      "learning_rate": 0.0,
      "loss": 1.3287,
      "num_tokens": 423926.0,
      "step": 1
    },
    {
      "epoch": 0.01968019680196802,
      "grad_norm": 3.67747034614091,
      "learning_rate": 6.25e-07,
      "loss": 1.2135,
      "num_tokens": 900553.0,
      "step": 2
    },
    {
      "epoch": 0.02952029520295203,
      "grad_norm": 3.8171114533122408,
      "learning_rate": 1.25e-06,
      "loss": 1.2489,
      "num_tokens": 1354101.0,
      "step": 3
    },
    {
      "epoch": 0.03936039360393604,
      "grad_norm": 3.601160453205568,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 1.2057,
      "num_tokens": 1831336.0,
      "step": 4
    },
    {
      "epoch": 0.04920049200492005,
      "grad_norm": 3.5064220048415957,
      "learning_rate": 2.5e-06,
      "loss": 1.2032,
      "num_tokens": 2299763.0,
      "step": 5
    },
    {
      "epoch": 0.05904059040590406,
      "grad_norm": 3.1555410715542345,
      "learning_rate": 3.125e-06,
      "loss": 1.2117,
      "num_tokens": 2739130.0,
      "step": 6
    },
    {
      "epoch": 0.06888068880688807,
      "grad_norm": 2.268607117088524,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 1.0872,
      "num_tokens": 3186646.0,
      "step": 7
    },
    {
      "epoch": 0.07872078720787208,
      "grad_norm": 2.2099777833656136,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 1.0835,
      "num_tokens": 3629799.0,
      "step": 8
    },
    {
      "epoch": 0.08856088560885608,
      "grad_norm": 1.8443437689000939,
      "learning_rate": 5e-06,
      "loss": 0.8645,
      "num_tokens": 4083162.0,
      "step": 9
    },
    {
      "epoch": 0.0984009840098401,
      "grad_norm": 1.9131458467328875,
      "learning_rate": 5.625e-06,
      "loss": 0.884,
      "num_tokens": 4514571.0,
      "step": 10
    },
    {
      "epoch": 0.10824108241082411,
      "grad_norm": 1.6995323396674717,
      "learning_rate": 6.25e-06,
      "loss": 0.7743,
      "num_tokens": 4990034.0,
      "step": 11
    },
    {
      "epoch": 0.11808118081180811,
      "grad_norm": 1.9772572417592447,
      "learning_rate": 6.875e-06,
      "loss": 0.53,
      "num_tokens": 5446252.0,
      "step": 12
    },
    {
      "epoch": 0.12792127921279212,
      "grad_norm": 2.235467875434991,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.4443,
      "num_tokens": 5880398.0,
      "step": 13
    },
    {
      "epoch": 0.13776137761377613,
      "grad_norm": 1.779888369557262,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.3535,
      "num_tokens": 6317527.0,
      "step": 14
    },
    {
      "epoch": 0.14760147601476015,
      "grad_norm": 1.4174646561758966,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.257,
      "num_tokens": 6768988.0,
      "step": 15
    },
    {
      "epoch": 0.15744157441574416,
      "grad_norm": 0.825210223865691,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.1351,
      "num_tokens": 7212525.0,
      "step": 16
    },
    {
      "epoch": 0.16728167281672818,
      "grad_norm": 0.5272789553691977,
      "learning_rate": 1e-05,
      "loss": 0.1092,
      "num_tokens": 7660418.0,
      "step": 17
    },
    {
      "epoch": 0.17712177121771217,
      "grad_norm": 0.4895013664227461,
      "learning_rate": 9.999909003036192e-06,
      "loss": 0.3213,
      "num_tokens": 8107247.0,
      "step": 18
    },
    {
      "epoch": 0.18696186961869618,
      "grad_norm": 0.2368744364092584,
      "learning_rate": 9.99963601582496e-06,
      "loss": 0.0892,
      "num_tokens": 8547828.0,
      "step": 19
    },
    {
      "epoch": 0.1968019680196802,
      "grad_norm": 0.17827168672763916,
      "learning_rate": 9.999181049406756e-06,
      "loss": 0.0814,
      "num_tokens": 8995945.0,
      "step": 20
    },
    {
      "epoch": 0.2066420664206642,
      "grad_norm": 0.1790643674004139,
      "learning_rate": 9.998544122181829e-06,
      "loss": 0.07,
      "num_tokens": 9428292.0,
      "step": 21
    },
    {
      "epoch": 0.21648216482164823,
      "grad_norm": 0.17378868386224117,
      "learning_rate": 9.997725259909487e-06,
      "loss": 0.0822,
      "num_tokens": 9872391.0,
      "step": 22
    },
    {
      "epoch": 0.22632226322263221,
      "grad_norm": 0.16172795763425857,
      "learning_rate": 9.996724495707056e-06,
      "loss": 0.0695,
      "num_tokens": 10321265.0,
      "step": 23
    },
    {
      "epoch": 0.23616236162361623,
      "grad_norm": 0.15374263186536802,
      "learning_rate": 9.995541870048537e-06,
      "loss": 0.0699,
      "num_tokens": 10756941.0,
      "step": 24
    },
    {
      "epoch": 0.24600246002460024,
      "grad_norm": 0.13904547469173764,
      "learning_rate": 9.994177430762971e-06,
      "loss": 0.0654,
      "num_tokens": 11211102.0,
      "step": 25
    },
    {
      "epoch": 0.25584255842558423,
      "grad_norm": 0.1313831102836863,
      "learning_rate": 9.992631233032507e-06,
      "loss": 0.0575,
      "num_tokens": 11655921.0,
      "step": 26
    },
    {
      "epoch": 0.2656826568265683,
      "grad_norm": 0.12358184868008147,
      "learning_rate": 9.990903339390164e-06,
      "loss": 0.0654,
      "num_tokens": 12123406.0,
      "step": 27
    },
    {
      "epoch": 0.27552275522755226,
      "grad_norm": 0.13102947055348427,
      "learning_rate": 9.988993819717312e-06,
      "loss": 0.0576,
      "num_tokens": 12571100.0,
      "step": 28
    },
    {
      "epoch": 0.2853628536285363,
      "grad_norm": 0.12370667408412227,
      "learning_rate": 9.986902751240836e-06,
      "loss": 0.0672,
      "num_tokens": 13004426.0,
      "step": 29
    },
    {
      "epoch": 0.2952029520295203,
      "grad_norm": 0.11805846922575773,
      "learning_rate": 9.984630218530014e-06,
      "loss": 0.0653,
      "num_tokens": 13453098.0,
      "step": 30
    },
    {
      "epoch": 0.3050430504305043,
      "grad_norm": 0.12596387207866033,
      "learning_rate": 9.982176313493108e-06,
      "loss": 0.0531,
      "num_tokens": 13910634.0,
      "step": 31
    },
    {
      "epoch": 0.3148831488314883,
      "grad_norm": 0.10677675254365898,
      "learning_rate": 9.979541135373628e-06,
      "loss": 0.0422,
      "num_tokens": 14359546.0,
      "step": 32
    },
    {
      "epoch": 0.3247232472324723,
      "grad_norm": 0.11973442758561412,
      "learning_rate": 9.976724790746333e-06,
      "loss": 0.0681,
      "num_tokens": 14810794.0,
      "step": 33
    },
    {
      "epoch": 0.33456334563345635,
      "grad_norm": 0.09765768004768276,
      "learning_rate": 9.973727393512921e-06,
      "loss": 0.0513,
      "num_tokens": 15282306.0,
      "step": 34
    },
    {
      "epoch": 0.34440344403444034,
      "grad_norm": 0.09901017562060792,
      "learning_rate": 9.970549064897407e-06,
      "loss": 0.0461,
      "num_tokens": 15719140.0,
      "step": 35
    },
    {
      "epoch": 0.35424354243542433,
      "grad_norm": 0.08726438112480077,
      "learning_rate": 9.967189933441243e-06,
      "loss": 0.0546,
      "num_tokens": 16160794.0,
      "step": 36
    },
    {
      "epoch": 0.3640836408364084,
      "grad_norm": 0.0862429330782484,
      "learning_rate": 9.9636501349981e-06,
      "loss": 0.0482,
      "num_tokens": 16594103.0,
      "step": 37
    },
    {
      "epoch": 0.37392373923739236,
      "grad_norm": 0.08703197270042312,
      "learning_rate": 9.95992981272838e-06,
      "loss": 0.0431,
      "num_tokens": 17050487.0,
      "step": 38
    },
    {
      "epoch": 0.3837638376383764,
      "grad_norm": 0.09107427503262026,
      "learning_rate": 9.956029117093432e-06,
      "loss": 0.0437,
      "num_tokens": 17495118.0,
      "step": 39
    },
    {
      "epoch": 0.3936039360393604,
      "grad_norm": 0.0937444613029256,
      "learning_rate": 9.951948205849457e-06,
      "loss": 0.0454,
      "num_tokens": 17970094.0,
      "step": 40
    },
    {
      "epoch": 0.4034440344403444,
      "grad_norm": 0.08311104466178655,
      "learning_rate": 9.947687244041143e-06,
      "loss": 0.035,
      "num_tokens": 18410339.0,
      "step": 41
    },
    {
      "epoch": 0.4132841328413284,
      "grad_norm": 0.08793054938548499,
      "learning_rate": 9.943246403994969e-06,
      "loss": 0.0398,
      "num_tokens": 18827350.0,
      "step": 42
    },
    {
      "epoch": 0.4231242312423124,
      "grad_norm": 0.0870730212694011,
      "learning_rate": 9.938625865312252e-06,
      "loss": 0.0399,
      "num_tokens": 19263988.0,
      "step": 43
    },
    {
      "epoch": 0.43296432964329645,
      "grad_norm": 0.09378264156348792,
      "learning_rate": 9.933825814861877e-06,
      "loss": 0.0417,
      "num_tokens": 19720928.0,
      "step": 44
    },
    {
      "epoch": 0.44280442804428044,
      "grad_norm": 0.09291535955010499,
      "learning_rate": 9.928846446772737e-06,
      "loss": 0.0503,
      "num_tokens": 20158462.0,
      "step": 45
    },
    {
      "epoch": 0.45264452644526443,
      "grad_norm": 0.09075012323845133,
      "learning_rate": 9.923687962425895e-06,
      "loss": 0.0383,
      "num_tokens": 20601362.0,
      "step": 46
    },
    {
      "epoch": 0.46248462484624847,
      "grad_norm": 0.08004209717985471,
      "learning_rate": 9.91835057044642e-06,
      "loss": 0.037,
      "num_tokens": 21072330.0,
      "step": 47
    },
    {
      "epoch": 0.47232472324723246,
      "grad_norm": 0.08493198697322972,
      "learning_rate": 9.912834486694963e-06,
      "loss": 0.0391,
      "num_tokens": 21525288.0,
      "step": 48
    },
    {
      "epoch": 0.4821648216482165,
      "grad_norm": 0.09568253792719325,
      "learning_rate": 9.907139934259025e-06,
      "loss": 0.0343,
      "num_tokens": 21953640.0,
      "step": 49
    },
    {
      "epoch": 0.4920049200492005,
      "grad_norm": 0.08303424847503821,
      "learning_rate": 9.90126714344393e-06,
      "loss": 0.0376,
      "num_tokens": 22435895.0,
      "step": 50
    },
    {
      "epoch": 0.5018450184501845,
      "grad_norm": 0.08402276200759969,
      "learning_rate": 9.895216351763515e-06,
      "loss": 0.0329,
      "num_tokens": 22883079.0,
      "step": 51
    },
    {
      "epoch": 0.5116851168511685,
      "grad_norm": 0.08739650107232434,
      "learning_rate": 9.888987803930523e-06,
      "loss": 0.0434,
      "num_tokens": 23334165.0,
      "step": 52
    },
    {
      "epoch": 0.5215252152521526,
      "grad_norm": 0.08000069499125344,
      "learning_rate": 9.882581751846707e-06,
      "loss": 0.0292,
      "num_tokens": 23797014.0,
      "step": 53
    },
    {
      "epoch": 0.5313653136531366,
      "grad_norm": 0.08649372481980258,
      "learning_rate": 9.87599845459264e-06,
      "loss": 0.0347,
      "num_tokens": 24288663.0,
      "step": 54
    },
    {
      "epoch": 0.5412054120541205,
      "grad_norm": 0.08523000136251592,
      "learning_rate": 9.869238178417235e-06,
      "loss": 0.0341,
      "num_tokens": 24710186.0,
      "step": 55
    },
    {
      "epoch": 0.5510455104551045,
      "grad_norm": 0.07465459840939084,
      "learning_rate": 9.862301196726988e-06,
      "loss": 0.0304,
      "num_tokens": 25160056.0,
      "step": 56
    },
    {
      "epoch": 0.5608856088560885,
      "grad_norm": 0.08503608654454574,
      "learning_rate": 9.855187790074906e-06,
      "loss": 0.0396,
      "num_tokens": 25598946.0,
      "step": 57
    },
    {
      "epoch": 0.5707257072570726,
      "grad_norm": 0.09884072621149297,
      "learning_rate": 9.847898246149173e-06,
      "loss": 0.0379,
      "num_tokens": 26066191.0,
      "step": 58
    },
    {
      "epoch": 0.5805658056580566,
      "grad_norm": 0.08138600445295054,
      "learning_rate": 9.840432859761504e-06,
      "loss": 0.0346,
      "num_tokens": 26518390.0,
      "step": 59
    },
    {
      "epoch": 0.5904059040590406,
      "grad_norm": 0.0914956119948303,
      "learning_rate": 9.832791932835232e-06,
      "loss": 0.0338,
      "num_tokens": 26971703.0,
      "step": 60
    },
    {
      "epoch": 0.6002460024600246,
      "grad_norm": 0.07667618947866506,
      "learning_rate": 9.824975774393089e-06,
      "loss": 0.0311,
      "num_tokens": 27413893.0,
      "step": 61
    },
    {
      "epoch": 0.6100861008610086,
      "grad_norm": 0.07239904062409902,
      "learning_rate": 9.816984700544714e-06,
      "loss": 0.0281,
      "num_tokens": 27868849.0,
      "step": 62
    },
    {
      "epoch": 0.6199261992619927,
      "grad_norm": 0.07354472191772266,
      "learning_rate": 9.808819034473869e-06,
      "loss": 0.0286,
      "num_tokens": 28286253.0,
      "step": 63
    },
    {
      "epoch": 0.6297662976629766,
      "grad_norm": 0.07644175488051343,
      "learning_rate": 9.800479106425356e-06,
      "loss": 0.0259,
      "num_tokens": 28699755.0,
      "step": 64
    },
    {
      "epoch": 0.6396063960639606,
      "grad_norm": 0.07705355988067097,
      "learning_rate": 9.791965253691687e-06,
      "loss": 0.0273,
      "num_tokens": 29142779.0,
      "step": 65
    },
    {
      "epoch": 0.6494464944649446,
      "grad_norm": 0.09001368693559736,
      "learning_rate": 9.783277820599408e-06,
      "loss": 0.0324,
      "num_tokens": 29564256.0,
      "step": 66
    },
    {
      "epoch": 0.6592865928659286,
      "grad_norm": 0.07998998288422189,
      "learning_rate": 9.774417158495208e-06,
      "loss": 0.0311,
      "num_tokens": 30035831.0,
      "step": 67
    },
    {
      "epoch": 0.6691266912669127,
      "grad_norm": 0.08094973090726079,
      "learning_rate": 9.765383625731683e-06,
      "loss": 0.0306,
      "num_tokens": 30471581.0,
      "step": 68
    },
    {
      "epoch": 0.6789667896678967,
      "grad_norm": 0.07575991095164336,
      "learning_rate": 9.756177587652857e-06,
      "loss": 0.0259,
      "num_tokens": 30914761.0,
      "step": 69
    },
    {
      "epoch": 0.6888068880688807,
      "grad_norm": 0.08255507372333148,
      "learning_rate": 9.746799416579403e-06,
      "loss": 0.0288,
      "num_tokens": 31351700.0,
      "step": 70
    },
    {
      "epoch": 0.6986469864698647,
      "grad_norm": 0.07383179182825,
      "learning_rate": 9.737249491793587e-06,
      "loss": 0.0322,
      "num_tokens": 31832339.0,
      "step": 71
    },
    {
      "epoch": 0.7084870848708487,
      "grad_norm": 0.0779807473775348,
      "learning_rate": 9.727528199523923e-06,
      "loss": 0.0384,
      "num_tokens": 32297687.0,
      "step": 72
    },
    {
      "epoch": 0.7183271832718328,
      "grad_norm": 0.9900117683369877,
      "learning_rate": 9.717635932929556e-06,
      "loss": 0.2712,
      "num_tokens": 32795919.0,
      "step": 73
    },
    {
      "epoch": 0.7281672816728167,
      "grad_norm": 0.0991868351288853,
      "learning_rate": 9.707573092084368e-06,
      "loss": 0.0419,
      "num_tokens": 33250913.0,
      "step": 74
    },
    {
      "epoch": 0.7380073800738007,
      "grad_norm": 0.08222641622667337,
      "learning_rate": 9.697340083960785e-06,
      "loss": 0.0289,
      "num_tokens": 33700970.0,
      "step": 75
    },
    {
      "epoch": 0.7478474784747847,
      "grad_norm": 0.0807953175546777,
      "learning_rate": 9.686937322413325e-06,
      "loss": 0.0286,
      "num_tokens": 34160528.0,
      "step": 76
    },
    {
      "epoch": 0.7576875768757687,
      "grad_norm": 0.07804479191069146,
      "learning_rate": 9.676365228161869e-06,
      "loss": 0.0288,
      "num_tokens": 34612256.0,
      "step": 77
    },
    {
      "epoch": 0.7675276752767528,
      "grad_norm": 0.0766842054326196,
      "learning_rate": 9.66562422877462e-06,
      "loss": 0.0232,
      "num_tokens": 35040279.0,
      "step": 78
    },
    {
      "epoch": 0.7773677736777368,
      "grad_norm": 0.0891517560198986,
      "learning_rate": 9.654714758650844e-06,
      "loss": 0.0297,
      "num_tokens": 35479921.0,
      "step": 79
    },
    {
      "epoch": 0.7872078720787208,
      "grad_norm": 0.07866162600023792,
      "learning_rate": 9.643637259003276e-06,
      "loss": 0.0246,
      "num_tokens": 35916545.0,
      "step": 80
    },
    {
      "epoch": 0.7970479704797048,
      "grad_norm": 0.09304350719642641,
      "learning_rate": 9.632392177840286e-06,
      "loss": 0.034,
      "num_tokens": 36366132.0,
      "step": 81
    },
    {
      "epoch": 0.8068880688806888,
      "grad_norm": 0.07700162361294739,
      "learning_rate": 9.620979969947759e-06,
      "loss": 0.0253,
      "num_tokens": 36826531.0,
      "step": 82
    },
    {
      "epoch": 0.8167281672816729,
      "grad_norm": 0.07988889172787667,
      "learning_rate": 9.609401096870707e-06,
      "loss": 0.0218,
      "num_tokens": 37250803.0,
      "step": 83
    },
    {
      "epoch": 0.8265682656826568,
      "grad_norm": 0.07662932628344148,
      "learning_rate": 9.597656026894591e-06,
      "loss": 0.0246,
      "num_tokens": 37707551.0,
      "step": 84
    },
    {
      "epoch": 0.8364083640836408,
      "grad_norm": 0.08311607561866212,
      "learning_rate": 9.585745235026391e-06,
      "loss": 0.0288,
      "num_tokens": 38157805.0,
      "step": 85
    },
    {
      "epoch": 0.8462484624846248,
      "grad_norm": 0.08297888399480316,
      "learning_rate": 9.5736692029754e-06,
      "loss": 0.04,
      "num_tokens": 38632172.0,
      "step": 86
    },
    {
      "epoch": 0.8560885608856088,
      "grad_norm": 0.07156880458651095,
      "learning_rate": 9.561428419133723e-06,
      "loss": 0.0244,
      "num_tokens": 39106655.0,
      "step": 87
    },
    {
      "epoch": 0.8659286592865929,
      "grad_norm": 0.07908760460903244,
      "learning_rate": 9.549023378556548e-06,
      "loss": 0.0283,
      "num_tokens": 39539107.0,
      "step": 88
    },
    {
      "epoch": 0.8757687576875769,
      "grad_norm": 0.07754860550749794,
      "learning_rate": 9.53645458294211e-06,
      "loss": 0.0267,
      "num_tokens": 39991573.0,
      "step": 89
    },
    {
      "epoch": 0.8856088560885609,
      "grad_norm": 0.07382048908783427,
      "learning_rate": 9.523722540611403e-06,
      "loss": 0.0302,
      "num_tokens": 40424055.0,
      "step": 90
    },
    {
      "epoch": 0.8954489544895449,
      "grad_norm": 0.07273833859487783,
      "learning_rate": 9.510827766487625e-06,
      "loss": 0.0256,
      "num_tokens": 40896577.0,
      "step": 91
    },
    {
      "epoch": 0.9052890528905289,
      "grad_norm": 0.07964369978108742,
      "learning_rate": 9.497770782075353e-06,
      "loss": 0.035,
      "num_tokens": 41366216.0,
      "step": 92
    },
    {
      "epoch": 0.915129151291513,
      "grad_norm": 0.07401188987280478,
      "learning_rate": 9.484552115439445e-06,
      "loss": 0.0245,
      "num_tokens": 41809244.0,
      "step": 93
    },
    {
      "epoch": 0.9249692496924969,
      "grad_norm": 0.07840443321039077,
      "learning_rate": 9.471172301183695e-06,
      "loss": 0.0289,
      "num_tokens": 42248913.0,
      "step": 94
    },
    {
      "epoch": 0.9348093480934809,
      "grad_norm": 0.08325366129705902,
      "learning_rate": 9.4576318804292e-06,
      "loss": 0.0231,
      "num_tokens": 42692031.0,
      "step": 95
    },
    {
      "epoch": 0.9446494464944649,
      "grad_norm": 0.07523171897596297,
      "learning_rate": 9.443931400792486e-06,
      "loss": 0.0231,
      "num_tokens": 43152827.0,
      "step": 96
    },
    {
      "epoch": 0.9544895448954489,
      "grad_norm": 0.08414955879961411,
      "learning_rate": 9.430071416363352e-06,
      "loss": 0.0361,
      "num_tokens": 43589003.0,
      "step": 97
    },
    {
      "epoch": 0.964329643296433,
      "grad_norm": 0.07780017154939398,
      "learning_rate": 9.416052487682465e-06,
      "loss": 0.0325,
      "num_tokens": 44039888.0,
      "step": 98
    },
    {
      "epoch": 0.974169741697417,
      "grad_norm": 0.07812625211660923,
      "learning_rate": 9.401875181718686e-06,
      "loss": 0.0269,
      "num_tokens": 44477755.0,
      "step": 99
    },
    {
      "epoch": 0.984009840098401,
      "grad_norm": 0.08986184340836745,
      "learning_rate": 9.387540071846155e-06,
      "loss": 0.028,
      "num_tokens": 44942597.0,
      "step": 100
    },
    {
      "epoch": 0.993849938499385,
      "grad_norm": 0.07602588052869161,
      "learning_rate": 9.373047737821078e-06,
      "loss": 0.0263,
      "num_tokens": 45381474.0,
      "step": 101
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.07602588052869161,
      "learning_rate": 9.358398765758296e-06,
      "loss": 0.0233,
      "num_tokens": 45594452.0,
      "step": 102
    },
    {
      "epoch": 1.009840098400984,
      "grad_norm": 0.10983513675872132,
      "learning_rate": 9.34359374810758e-06,
      "loss": 0.0203,
      "num_tokens": 46030876.0,
      "step": 103
    },
    {
      "epoch": 1.019680196801968,
      "grad_norm": 0.06573014575776571,
      "learning_rate": 9.328633283629666e-06,
      "loss": 0.0227,
      "num_tokens": 46486828.0,
      "step": 104
    },
    {
      "epoch": 1.029520295202952,
      "grad_norm": 0.08034715377363097,
      "learning_rate": 9.31351797737204e-06,
      "loss": 0.0318,
      "num_tokens": 46952208.0,
      "step": 105
    },
    {
      "epoch": 1.039360393603936,
      "grad_norm": 0.07429105498387009,
      "learning_rate": 9.29824844064447e-06,
      "loss": 0.0303,
      "num_tokens": 47410999.0,
      "step": 106
    },
    {
      "epoch": 1.04920049200492,
      "grad_norm": 0.08198831738294611,
      "learning_rate": 9.282825290994282e-06,
      "loss": 0.0214,
      "num_tokens": 47859777.0,
      "step": 107
    },
    {
      "epoch": 1.0590405904059041,
      "grad_norm": 0.08882279229683457,
      "learning_rate": 9.267249152181379e-06,
      "loss": 0.0348,
      "num_tokens": 48313998.0,
      "step": 108
    },
    {
      "epoch": 1.068880688806888,
      "grad_norm": 0.07979621896563821,
      "learning_rate": 9.251520654153028e-06,
      "loss": 0.0217,
      "num_tokens": 48768989.0,
      "step": 109
    },
    {
      "epoch": 1.0787207872078721,
      "grad_norm": 0.07487008128556895,
      "learning_rate": 9.235640433018363e-06,
      "loss": 0.0312,
      "num_tokens": 49228144.0,
      "step": 110
    },
    {
      "epoch": 1.088560885608856,
      "grad_norm": 0.07679612898308323,
      "learning_rate": 9.219609131022684e-06,
      "loss": 0.0931,
      "num_tokens": 49660848.0,
      "step": 111
    },
    {
      "epoch": 1.09840098400984,
      "grad_norm": 0.2760363965156351,
      "learning_rate": 9.203427396521454e-06,
      "loss": 0.0199,
      "num_tokens": 50101951.0,
      "step": 112
    },
    {
      "epoch": 1.1082410824108242,
      "grad_norm": 0.0785172085157219,
      "learning_rate": 9.187095883954104e-06,
      "loss": 0.0249,
      "num_tokens": 50561884.0,
      "step": 113
    },
    {
      "epoch": 1.118081180811808,
      "grad_norm": 0.07511192184574486,
      "learning_rate": 9.170615253817547e-06,
      "loss": 0.0202,
      "num_tokens": 51023121.0,
      "step": 114
    },
    {
      "epoch": 1.1279212792127922,
      "grad_norm": 0.0788369940517159,
      "learning_rate": 9.153986172639474e-06,
      "loss": 0.0242,
      "num_tokens": 51475717.0,
      "step": 115
    },
    {
      "epoch": 1.137761377613776,
      "grad_norm": 0.08125072620630153,
      "learning_rate": 9.137209312951395e-06,
      "loss": 0.023,
      "num_tokens": 51936171.0,
      "step": 116
    },
    {
      "epoch": 1.1476014760147601,
      "grad_norm": 0.07976122735928637,
      "learning_rate": 9.12028535326144e-06,
      "loss": 0.0243,
      "num_tokens": 52386546.0,
      "step": 117
    },
    {
      "epoch": 1.1574415744157442,
      "grad_norm": 0.08346012008591713,
      "learning_rate": 9.103214978026922e-06,
      "loss": 0.0325,
      "num_tokens": 52836033.0,
      "step": 118
    },
    {
      "epoch": 1.1672816728167281,
      "grad_norm": 0.11906435467388006,
      "learning_rate": 9.085998877626644e-06,
      "loss": 0.0322,
      "num_tokens": 53268007.0,
      "step": 119
    },
    {
      "epoch": 1.1771217712177122,
      "grad_norm": 0.07249411043595039,
      "learning_rate": 9.068637748332993e-06,
      "loss": 0.022,
      "num_tokens": 53704420.0,
      "step": 120
    },
    {
      "epoch": 1.186961869618696,
      "grad_norm": 0.07720155000392087,
      "learning_rate": 9.051132292283772e-06,
      "loss": 0.0175,
      "num_tokens": 54139993.0,
      "step": 121
    },
    {
      "epoch": 1.1968019680196802,
      "grad_norm": 0.07940921390702492,
      "learning_rate": 9.033483217453801e-06,
      "loss": 0.028,
      "num_tokens": 54570992.0,
      "step": 122
    },
    {
      "epoch": 1.2066420664206643,
      "grad_norm": 0.07451157010744122,
      "learning_rate": 9.015691237626292e-06,
      "loss": 0.0372,
      "num_tokens": 55010129.0,
      "step": 123
    },
    {
      "epoch": 1.2164821648216482,
      "grad_norm": 0.0847625901422177,
      "learning_rate": 8.997757072363976e-06,
      "loss": 0.026,
      "num_tokens": 55484006.0,
      "step": 124
    },
    {
      "epoch": 1.2263222632226323,
      "grad_norm": 0.0734447382667372,
      "learning_rate": 8.979681446980002e-06,
      "loss": 0.0187,
      "num_tokens": 55916003.0,
      "step": 125
    },
    {
      "epoch": 1.2361623616236161,
      "grad_norm": 0.07720210069754488,
      "learning_rate": 8.961465092508607e-06,
      "loss": 0.0204,
      "num_tokens": 56367051.0,
      "step": 126
    },
    {
      "epoch": 1.2460024600246002,
      "grad_norm": 0.07005770488907263,
      "learning_rate": 8.943108745675542e-06,
      "loss": 0.0191,
      "num_tokens": 56809919.0,
      "step": 127
    },
    {
      "epoch": 1.2558425584255843,
      "grad_norm": 0.070753857700872,
      "learning_rate": 8.92461314886829e-06,
      "loss": 0.0166,
      "num_tokens": 57239307.0,
      "step": 128
    },
    {
      "epoch": 1.2656826568265682,
      "grad_norm": 0.07412229123559273,
      "learning_rate": 8.905979050106029e-06,
      "loss": 0.0182,
      "num_tokens": 57659139.0,
      "step": 129
    },
    {
      "epoch": 1.2755227552275523,
      "grad_norm": 0.07409420522842904,
      "learning_rate": 8.887207203009385e-06,
      "loss": 0.0182,
      "num_tokens": 58106309.0,
      "step": 130
    },
    {
      "epoch": 1.2853628536285364,
      "grad_norm": 0.07792555021983492,
      "learning_rate": 8.868298366769956e-06,
      "loss": 0.02,
      "num_tokens": 58533301.0,
      "step": 131
    },
    {
      "epoch": 1.2952029520295203,
      "grad_norm": 0.07755261181847745,
      "learning_rate": 8.849253306119601e-06,
      "loss": 0.018,
      "num_tokens": 58966567.0,
      "step": 132
    },
    {
      "epoch": 1.3050430504305042,
      "grad_norm": 0.07236047683074463,
      "learning_rate": 8.83007279129952e-06,
      "loss": 0.0231,
      "num_tokens": 59420421.0,
      "step": 133
    },
    {
      "epoch": 1.3148831488314883,
      "grad_norm": 0.07698814451544558,
      "learning_rate": 8.810757598029094e-06,
      "loss": 0.0171,
      "num_tokens": 59841646.0,
      "step": 134
    },
    {
      "epoch": 1.3247232472324724,
      "grad_norm": 0.07814813953683186,
      "learning_rate": 8.79130850747452e-06,
      "loss": 0.02,
      "num_tokens": 60262792.0,
      "step": 135
    },
    {
      "epoch": 1.3345633456334562,
      "grad_norm": 0.06873838482423762,
      "learning_rate": 8.771726306217217e-06,
      "loss": 0.0185,
      "num_tokens": 60732922.0,
      "step": 136
    },
    {
      "epoch": 1.3444034440344403,
      "grad_norm": 0.09687913725758614,
      "learning_rate": 8.752011786222011e-06,
      "loss": 0.0176,
      "num_tokens": 61205715.0,
      "step": 137
    },
    {
      "epoch": 1.3542435424354244,
      "grad_norm": 0.07245054518431464,
      "learning_rate": 8.732165744805107e-06,
      "loss": 0.0235,
      "num_tokens": 61643580.0,
      "step": 138
    },
    {
      "epoch": 1.3640836408364083,
      "grad_norm": 0.0751731684247356,
      "learning_rate": 8.712188984601845e-06,
      "loss": 0.0204,
      "num_tokens": 62086664.0,
      "step": 139
    },
    {
      "epoch": 1.3739237392373924,
      "grad_norm": 0.0765957432328115,
      "learning_rate": 8.692082313534233e-06,
      "loss": 0.0246,
      "num_tokens": 62556237.0,
      "step": 140
    },
    {
      "epoch": 1.3837638376383765,
      "grad_norm": 0.07942748233411417,
      "learning_rate": 8.671846544778284e-06,
      "loss": 0.0262,
      "num_tokens": 63014753.0,
      "step": 141
    },
    {
      "epoch": 1.3936039360393604,
      "grad_norm": 0.07220660807543852,
      "learning_rate": 8.651482496731116e-06,
      "loss": 0.0206,
      "num_tokens": 63462646.0,
      "step": 142
    },
    {
      "epoch": 1.4034440344403443,
      "grad_norm": 0.07123468502943975,
      "learning_rate": 8.630990992977854e-06,
      "loss": 0.0171,
      "num_tokens": 63953362.0,
      "step": 143
    },
    {
      "epoch": 1.4132841328413284,
      "grad_norm": 0.06845259344800432,
      "learning_rate": 8.61037286225834e-06,
      "loss": 0.0176,
      "num_tokens": 64407104.0,
      "step": 144
    },
    {
      "epoch": 1.4231242312423125,
      "grad_norm": 0.0712728854892911,
      "learning_rate": 8.589628938433587e-06,
      "loss": 0.0198,
      "num_tokens": 64843926.0,
      "step": 145
    },
    {
      "epoch": 1.4329643296432963,
      "grad_norm": 0.07477061506965738,
      "learning_rate": 8.56876006045208e-06,
      "loss": 0.0186,
      "num_tokens": 65296311.0,
      "step": 146
    },
    {
      "epoch": 1.4428044280442804,
      "grad_norm": 0.06777146676082023,
      "learning_rate": 8.547767072315835e-06,
      "loss": 0.019,
      "num_tokens": 65756092.0,
      "step": 147
    },
    {
      "epoch": 1.4526445264452645,
      "grad_norm": 0.07114975862038111,
      "learning_rate": 8.526650823046266e-06,
      "loss": 0.0178,
      "num_tokens": 66217555.0,
      "step": 148
    },
    {
      "epoch": 1.4624846248462484,
      "grad_norm": 0.07525940320393967,
      "learning_rate": 8.505412166649847e-06,
      "loss": 0.0179,
      "num_tokens": 66685889.0,
      "step": 149
    },
    {
      "epoch": 1.4723247232472325,
      "grad_norm": 0.06408400112895582,
      "learning_rate": 8.484051962083579e-06,
      "loss": 0.0218,
      "num_tokens": 67133723.0,
      "step": 150
    },
    {
      "epoch": 1.4821648216482166,
      "grad_norm": 0.28687790559136056,
      "learning_rate": 8.462571073220243e-06,
      "loss": 0.0384,
      "num_tokens": 67578744.0,
      "step": 151
    },
    {
      "epoch": 1.4920049200492005,
      "grad_norm": 0.0793310962273527,
      "learning_rate": 8.44097036881347e-06,
      "loss": 0.0229,
      "num_tokens": 68014169.0,
      "step": 152
    },
    {
      "epoch": 1.5018450184501844,
      "grad_norm": 0.06714316191575523,
      "learning_rate": 8.419250722462603e-06,
      "loss": 0.0171,
      "num_tokens": 68463185.0,
      "step": 153
    },
    {
      "epoch": 1.5116851168511685,
      "grad_norm": 0.07317824154830675,
      "learning_rate": 8.39741301257736e-06,
      "loss": 0.0168,
      "num_tokens": 68915214.0,
      "step": 154
    },
    {
      "epoch": 1.5215252152521526,
      "grad_norm": 0.07498605748005185,
      "learning_rate": 8.375458122342317e-06,
      "loss": 0.0218,
      "num_tokens": 69372682.0,
      "step": 155
    },
    {
      "epoch": 1.5313653136531364,
      "grad_norm": 0.07930974741179314,
      "learning_rate": 8.353386939681186e-06,
      "loss": 0.0197,
      "num_tokens": 69856887.0,
      "step": 156
    },
    {
      "epoch": 1.5412054120541205,
      "grad_norm": 0.06557036848250264,
      "learning_rate": 8.331200357220908e-06,
      "loss": 0.0169,
      "num_tokens": 70281087.0,
      "step": 157
    },
    {
      "epoch": 1.5510455104551046,
      "grad_norm": 0.06911145883255942,
      "learning_rate": 8.308899272255542e-06,
      "loss": 0.0258,
      "num_tokens": 70703636.0,
      "step": 158
    },
    {
      "epoch": 1.5608856088560885,
      "grad_norm": 0.07694897194945108,
      "learning_rate": 8.286484586709989e-06,
      "loss": 0.0195,
      "num_tokens": 71151769.0,
      "step": 159
    },
    {
      "epoch": 1.5707257072570726,
      "grad_norm": 0.07509068702973927,
      "learning_rate": 8.263957207103506e-06,
      "loss": 0.0163,
      "num_tokens": 71608512.0,
      "step": 160
    },
    {
      "epoch": 1.5805658056580567,
      "grad_norm": 0.06708625688834685,
      "learning_rate": 8.241318044513046e-06,
      "loss": 0.0182,
      "num_tokens": 72064241.0,
      "step": 161
    },
    {
      "epoch": 1.5904059040590406,
      "grad_norm": 0.07409751654155339,
      "learning_rate": 8.218568014536414e-06,
      "loss": 0.0214,
      "num_tokens": 72529560.0,
      "step": 162
    },
    {
      "epoch": 1.6002460024600245,
      "grad_norm": 0.08398773553611014,
      "learning_rate": 8.195708037255233e-06,
      "loss": 0.0219,
      "num_tokens": 72983545.0,
      "step": 163
    },
    {
      "epoch": 1.6100861008610086,
      "grad_norm": 0.0924163975250339,
      "learning_rate": 8.172739037197739e-06,
      "loss": 0.0199,
      "num_tokens": 73420020.0,
      "step": 164
    },
    {
      "epoch": 1.6199261992619927,
      "grad_norm": 0.07341361509627292,
      "learning_rate": 8.149661943301382e-06,
      "loss": 0.0193,
      "num_tokens": 73867089.0,
      "step": 165
    },
    {
      "epoch": 1.6297662976629765,
      "grad_norm": 0.09734659605562107,
      "learning_rate": 8.126477688875262e-06,
      "loss": 0.0393,
      "num_tokens": 74337960.0,
      "step": 166
    },
    {
      "epoch": 1.6396063960639606,
      "grad_norm": 0.0653922306331562,
      "learning_rate": 8.103187211562386e-06,
      "loss": 0.0169,
      "num_tokens": 74779558.0,
      "step": 167
    },
    {
      "epoch": 1.6494464944649447,
      "grad_norm": 0.07532255669592061,
      "learning_rate": 8.079791453301742e-06,
      "loss": 0.0181,
      "num_tokens": 75249832.0,
      "step": 168
    },
    {
      "epoch": 1.6592865928659286,
      "grad_norm": 0.06592599083206258,
      "learning_rate": 8.056291360290202e-06,
      "loss": 0.0189,
      "num_tokens": 75680939.0,
      "step": 169
    },
    {
      "epoch": 1.6691266912669127,
      "grad_norm": 0.07466148209100053,
      "learning_rate": 8.032687882944264e-06,
      "loss": 0.0197,
      "num_tokens": 76131450.0,
      "step": 170
    },
    {
      "epoch": 1.6789667896678968,
      "grad_norm": 0.06887573727890169,
      "learning_rate": 8.0089819758616e-06,
      "loss": 0.0192,
      "num_tokens": 76593750.0,
      "step": 171
    },
    {
      "epoch": 1.6888068880688807,
      "grad_norm": 0.06693500545746488,
      "learning_rate": 7.985174597782469e-06,
      "loss": 0.0184,
      "num_tokens": 77050866.0,
      "step": 172
    },
    {
      "epoch": 1.6986469864698646,
      "grad_norm": 0.0743814094089007,
      "learning_rate": 7.961266711550922e-06,
      "loss": 0.018,
      "num_tokens": 77464839.0,
      "step": 173
    },
    {
      "epoch": 1.7084870848708487,
      "grad_norm": 0.07491525806289706,
      "learning_rate": 7.937259284075872e-06,
      "loss": 0.0242,
      "num_tokens": 77916642.0,
      "step": 174
    },
    {
      "epoch": 1.7183271832718328,
      "grad_norm": 0.06409329354600994,
      "learning_rate": 7.913153286291995e-06,
      "loss": 0.0159,
      "num_tokens": 78389781.0,
      "step": 175
    },
    {
      "epoch": 1.7281672816728166,
      "grad_norm": 0.07234553574283793,
      "learning_rate": 7.888949693120443e-06,
      "loss": 0.0185,
      "num_tokens": 78840356.0,
      "step": 176
    },
    {
      "epoch": 1.7380073800738007,
      "grad_norm": 0.06775336526687877,
      "learning_rate": 7.864649483429442e-06,
      "loss": 0.0222,
      "num_tokens": 79280780.0,
      "step": 177
    },
    {
      "epoch": 1.7478474784747848,
      "grad_norm": 0.06875670670455086,
      "learning_rate": 7.840253639994676e-06,
      "loss": 0.2282,
      "num_tokens": 79776798.0,
      "step": 178
    },
    {
      "epoch": 1.7576875768757687,
      "grad_norm": 0.6287911227006427,
      "learning_rate": 7.815763149459563e-06,
      "loss": 0.0278,
      "num_tokens": 80233063.0,
      "step": 179
    },
    {
      "epoch": 1.7675276752767528,
      "grad_norm": 0.0846354199885941,
      "learning_rate": 7.791179002295334e-06,
      "loss": 0.0176,
      "num_tokens": 80677429.0,
      "step": 180
    },
    {
      "epoch": 1.777367773677737,
      "grad_norm": 0.07041805151413434,
      "learning_rate": 7.766502192760995e-06,
      "loss": 0.019,
      "num_tokens": 81122406.0,
      "step": 181
    },
    {
      "epoch": 1.7872078720787208,
      "grad_norm": 0.0643818791121022,
      "learning_rate": 7.741733718863096e-06,
      "loss": 0.021,
      "num_tokens": 81570725.0,
      "step": 182
    },
    {
      "epoch": 1.7970479704797047,
      "grad_norm": 0.09698204542146324,
      "learning_rate": 7.71687458231538e-06,
      "loss": 0.0164,
      "num_tokens": 82009994.0,
      "step": 183
    },
    {
      "epoch": 1.8068880688806888,
      "grad_norm": 0.06854732624004896,
      "learning_rate": 7.69192578849827e-06,
      "loss": 0.0166,
      "num_tokens": 82455315.0,
      "step": 184
    },
    {
      "epoch": 1.8167281672816729,
      "grad_norm": 0.08363256711677616,
      "learning_rate": 7.666888346418205e-06,
      "loss": 0.026,
      "num_tokens": 82901866.0,
      "step": 185
    },
    {
      "epoch": 1.8265682656826567,
      "grad_norm": 0.07243825921308177,
      "learning_rate": 7.641763268666832e-06,
      "loss": 0.019,
      "num_tokens": 83346981.0,
      "step": 186
    },
    {
      "epoch": 1.8364083640836408,
      "grad_norm": 0.06930096159612371,
      "learning_rate": 7.616551571380061e-06,
      "loss": 0.0214,
      "num_tokens": 83806206.0,
      "step": 187
    },
    {
      "epoch": 1.846248462484625,
      "grad_norm": 0.0681854243495615,
      "learning_rate": 7.5912542741969585e-06,
      "loss": 0.0163,
      "num_tokens": 84274704.0,
      "step": 188
    },
    {
      "epoch": 1.8560885608856088,
      "grad_norm": 0.07242016185622592,
      "learning_rate": 7.5658724002185215e-06,
      "loss": 0.0176,
      "num_tokens": 84711534.0,
      "step": 189
    },
    {
      "epoch": 1.865928659286593,
      "grad_norm": 0.06579105822967504,
      "learning_rate": 7.54040697596629e-06,
      "loss": 0.0283,
      "num_tokens": 85160170.0,
      "step": 190
    },
    {
      "epoch": 1.875768757687577,
      "grad_norm": 0.07108784887648237,
      "learning_rate": 7.514859031340835e-06,
      "loss": 0.0396,
      "num_tokens": 85613402.0,
      "step": 191
    },
    {
      "epoch": 1.8856088560885609,
      "grad_norm": 0.1388795021837048,
      "learning_rate": 7.489229599580111e-06,
      "loss": 0.0246,
      "num_tokens": 86052381.0,
      "step": 192
    },
    {
      "epoch": 1.8954489544895448,
      "grad_norm": 0.07205397955257539,
      "learning_rate": 7.463519717217663e-06,
      "loss": 0.0168,
      "num_tokens": 86528412.0,
      "step": 193
    },
    {
      "epoch": 1.9052890528905289,
      "grad_norm": 0.06405950191637912,
      "learning_rate": 7.437730424040702e-06,
      "loss": 0.0172,
      "num_tokens": 86961573.0,
      "step": 194
    },
    {
      "epoch": 1.915129151291513,
      "grad_norm": 0.06777861356344914,
      "learning_rate": 7.411862763048068e-06,
      "loss": 0.2543,
      "num_tokens": 87398765.0,
      "step": 195
    },
    {
      "epoch": 1.9249692496924968,
      "grad_norm": 0.6838887307240747,
      "learning_rate": 7.38591778040803e-06,
      "loss": 0.0173,
      "num_tokens": 87844520.0,
      "step": 196
    },
    {
      "epoch": 1.934809348093481,
      "grad_norm": 0.0776240582840387,
      "learning_rate": 7.359896525415986e-06,
      "loss": 0.017,
      "num_tokens": 88283922.0,
      "step": 197
    },
    {
      "epoch": 1.944649446494465,
      "grad_norm": 0.07000567204619491,
      "learning_rate": 7.333800050452024e-06,
      "loss": 0.017,
      "num_tokens": 88751425.0,
      "step": 198
    },
    {
      "epoch": 1.954489544895449,
      "grad_norm": 0.06845270762016023,
      "learning_rate": 7.307629410938364e-06,
      "loss": 0.0156,
      "num_tokens": 89205645.0,
      "step": 199
    },
    {
      "epoch": 1.964329643296433,
      "grad_norm": 0.06368141241550528,
      "learning_rate": 7.281385665296663e-06,
      "loss": 0.0241,
      "num_tokens": 89660304.0,
      "step": 200
    },
    {
      "epoch": 1.974169741697417,
      "grad_norm": 0.07288122003316398,
      "learning_rate": 7.255069874905221e-06,
      "loss": 0.0177,
      "num_tokens": 90107418.0,
      "step": 201
    },
    {
      "epoch": 1.984009840098401,
      "grad_norm": 0.06287805035179808,
      "learning_rate": 7.228683104056051e-06,
      "loss": 0.0162,
      "num_tokens": 90539742.0,
      "step": 202
    },
    {
      "epoch": 1.9938499384993849,
      "grad_norm": 0.0689550244527601,
      "learning_rate": 7.202226419911832e-06,
      "loss": 0.0158,
      "num_tokens": 90980046.0,
      "step": 203
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.08227420282080453,
      "learning_rate": 7.175700892462757e-06,
      "loss": 0.0146,
      "num_tokens": 91185077.0,
      "step": 204
    },
    {
      "epoch": 2.009840098400984,
      "grad_norm": 0.07497487377417848,
      "learning_rate": 7.149107594483251e-06,
      "loss": 0.0179,
      "num_tokens": 91614882.0,
      "step": 205
    },
    {
      "epoch": 2.019680196801968,
      "grad_norm": 0.06696426999273204,
      "learning_rate": 7.122447601488592e-06,
      "loss": 0.0138,
      "num_tokens": 92080175.0,
      "step": 206
    },
    {
      "epoch": 2.029520295202952,
      "grad_norm": 0.6914326541667255,
      "learning_rate": 7.095721991691411e-06,
      "loss": 0.0906,
      "num_tokens": 92537610.0,
      "step": 207
    },
    {
      "epoch": 2.039360393603936,
      "grad_norm": 0.07698410304489973,
      "learning_rate": 7.0689318459580845e-06,
      "loss": 0.014,
      "num_tokens": 93006593.0,
      "step": 208
    },
    {
      "epoch": 2.0492004920049203,
      "grad_norm": 0.06605656882870746,
      "learning_rate": 7.042078247765019e-06,
      "loss": 0.0135,
      "num_tokens": 93476972.0,
      "step": 209
    },
    {
      "epoch": 2.059040590405904,
      "grad_norm": 0.20519577755096738,
      "learning_rate": 7.015162283154843e-06,
      "loss": 0.0448,
      "num_tokens": 93923596.0,
      "step": 210
    },
    {
      "epoch": 2.068880688806888,
      "grad_norm": 0.07138995831165522,
      "learning_rate": 6.988185040692469e-06,
      "loss": 0.0147,
      "num_tokens": 94363955.0,
      "step": 211
    },
    {
      "epoch": 2.078720787207872,
      "grad_norm": 0.07243575038647761,
      "learning_rate": 6.961147611421076e-06,
      "loss": 0.0133,
      "num_tokens": 94822454.0,
      "step": 212
    },
    {
      "epoch": 2.088560885608856,
      "grad_norm": 0.06470408361373992,
      "learning_rate": 6.934051088817988e-06,
      "loss": 0.0129,
      "num_tokens": 95281989.0,
      "step": 213
    },
    {
      "epoch": 2.09840098400984,
      "grad_norm": 0.06874194166286132,
      "learning_rate": 6.906896568750441e-06,
      "loss": 0.0133,
      "num_tokens": 95718050.0,
      "step": 214
    },
    {
      "epoch": 2.108241082410824,
      "grad_norm": 0.06876988398180649,
      "learning_rate": 6.87968514943127e-06,
      "loss": 0.0131,
      "num_tokens": 96162240.0,
      "step": 215
    },
    {
      "epoch": 2.1180811808118083,
      "grad_norm": 0.06407419572232166,
      "learning_rate": 6.852417931374494e-06,
      "loss": 0.0154,
      "num_tokens": 96606411.0,
      "step": 216
    },
    {
      "epoch": 2.127921279212792,
      "grad_norm": 0.07884002350292682,
      "learning_rate": 6.825096017350807e-06,
      "loss": 0.0145,
      "num_tokens": 97056671.0,
      "step": 217
    },
    {
      "epoch": 2.137761377613776,
      "grad_norm": 0.9981711556732346,
      "learning_rate": 6.797720512342967e-06,
      "loss": 0.2386,
      "num_tokens": 97514395.0,
      "step": 218
    },
    {
      "epoch": 2.14760147601476,
      "grad_norm": 0.08409368119046862,
      "learning_rate": 6.77029252350113e-06,
      "loss": 0.0131,
      "num_tokens": 97963358.0,
      "step": 219
    },
    {
      "epoch": 2.1574415744157442,
      "grad_norm": 0.07248619207767666,
      "learning_rate": 6.742813160098054e-06,
      "loss": 0.0209,
      "num_tokens": 98418943.0,
      "step": 220
    },
    {
      "epoch": 2.167281672816728,
      "grad_norm": 0.0644855598598356,
      "learning_rate": 6.715283533484242e-06,
      "loss": 0.013,
      "num_tokens": 98877761.0,
      "step": 221
    },
    {
      "epoch": 2.177121771217712,
      "grad_norm": 0.06622877203564365,
      "learning_rate": 6.6877047570430044e-06,
      "loss": 0.0142,
      "num_tokens": 99319699.0,
      "step": 222
    },
    {
      "epoch": 2.1869618696186963,
      "grad_norm": 0.07112191399232934,
      "learning_rate": 6.660077946145412e-06,
      "loss": 0.0141,
      "num_tokens": 99741161.0,
      "step": 223
    },
    {
      "epoch": 2.19680196801968,
      "grad_norm": 0.06421873614899783,
      "learning_rate": 6.632404218105205e-06,
      "loss": 0.0125,
      "num_tokens": 100166108.0,
      "step": 224
    },
    {
      "epoch": 2.206642066420664,
      "grad_norm": 0.10033810460897824,
      "learning_rate": 6.604684692133597e-06,
      "loss": 0.0252,
      "num_tokens": 100619482.0,
      "step": 225
    },
    {
      "epoch": 2.2164821648216484,
      "grad_norm": 0.06530150422818083,
      "learning_rate": 6.576920489294011e-06,
      "loss": 0.0127,
      "num_tokens": 101046330.0,
      "step": 226
    },
    {
      "epoch": 2.2263222632226323,
      "grad_norm": 0.07185919142366148,
      "learning_rate": 6.549112732456739e-06,
      "loss": 0.016,
      "num_tokens": 101495504.0,
      "step": 227
    },
    {
      "epoch": 2.236162361623616,
      "grad_norm": 0.06818636079419191,
      "learning_rate": 6.5212625462535365e-06,
      "loss": 0.0213,
      "num_tokens": 101968574.0,
      "step": 228
    },
    {
      "epoch": 2.2460024600246005,
      "grad_norm": 0.06992643360327205,
      "learning_rate": 6.493371057032129e-06,
      "loss": 0.0129,
      "num_tokens": 102400777.0,
      "step": 229
    },
    {
      "epoch": 2.2558425584255843,
      "grad_norm": 0.07704204356207317,
      "learning_rate": 6.465439392810664e-06,
      "loss": 0.015,
      "num_tokens": 102843075.0,
      "step": 230
    },
    {
      "epoch": 2.265682656826568,
      "grad_norm": 0.06544957942318019,
      "learning_rate": 6.4374686832320944e-06,
      "loss": 0.0123,
      "num_tokens": 103261288.0,
      "step": 231
    },
    {
      "epoch": 2.275522755227552,
      "grad_norm": 0.062334728387161344,
      "learning_rate": 6.409460059518482e-06,
      "loss": 0.0129,
      "num_tokens": 103717825.0,
      "step": 232
    },
    {
      "epoch": 2.2853628536285364,
      "grad_norm": 0.06332858275476305,
      "learning_rate": 6.381414654425261e-06,
      "loss": 0.0206,
      "num_tokens": 104165764.0,
      "step": 233
    },
    {
      "epoch": 2.2952029520295203,
      "grad_norm": 0.09448999187172064,
      "learning_rate": 6.353333602195414e-06,
      "loss": 0.0124,
      "num_tokens": 104606652.0,
      "step": 234
    },
    {
      "epoch": 2.305043050430504,
      "grad_norm": 0.0668177771084121,
      "learning_rate": 6.325218038513604e-06,
      "loss": 0.0126,
      "num_tokens": 105039670.0,
      "step": 235
    },
    {
      "epoch": 2.3148831488314885,
      "grad_norm": 0.06130143066732146,
      "learning_rate": 6.2970691004602425e-06,
      "loss": 0.0128,
      "num_tokens": 105498133.0,
      "step": 236
    },
    {
      "epoch": 2.3247232472324724,
      "grad_norm": 0.06353571099299025,
      "learning_rate": 6.26888792646551e-06,
      "loss": 0.0132,
      "num_tokens": 105955337.0,
      "step": 237
    },
    {
      "epoch": 2.3345633456334562,
      "grad_norm": 0.07072744488591255,
      "learning_rate": 6.240675656263303e-06,
      "loss": 0.0206,
      "num_tokens": 106398065.0,
      "step": 238
    },
    {
      "epoch": 2.34440344403444,
      "grad_norm": 0.0701393753248028,
      "learning_rate": 6.212433430845145e-06,
      "loss": 0.0207,
      "num_tokens": 106835350.0,
      "step": 239
    },
    {
      "epoch": 2.3542435424354244,
      "grad_norm": 0.06848300507150402,
      "learning_rate": 6.184162392414044e-06,
      "loss": 0.0145,
      "num_tokens": 107296896.0,
      "step": 240
    },
    {
      "epoch": 2.3640836408364083,
      "grad_norm": 0.07142535679112567,
      "learning_rate": 6.155863684338294e-06,
      "loss": 0.0142,
      "num_tokens": 107740689.0,
      "step": 241
    },
    {
      "epoch": 2.373923739237392,
      "grad_norm": 0.061123970206678674,
      "learning_rate": 6.127538451105232e-06,
      "loss": 0.0127,
      "num_tokens": 108196343.0,
      "step": 242
    },
    {
      "epoch": 2.3837638376383765,
      "grad_norm": 0.0699480717937599,
      "learning_rate": 6.099187838274959e-06,
      "loss": 0.0136,
      "num_tokens": 108646263.0,
      "step": 243
    },
    {
      "epoch": 2.3936039360393604,
      "grad_norm": 0.06593159068712609,
      "learning_rate": 6.070812992434003e-06,
      "loss": 0.0127,
      "num_tokens": 109089367.0,
      "step": 244
    },
    {
      "epoch": 2.4034440344403443,
      "grad_norm": 0.07004342519131408,
      "learning_rate": 6.042415061148954e-06,
      "loss": 0.015,
      "num_tokens": 109534992.0,
      "step": 245
    },
    {
      "epoch": 2.4132841328413286,
      "grad_norm": 0.08012326478595262,
      "learning_rate": 6.013995192920044e-06,
      "loss": 0.0181,
      "num_tokens": 110020906.0,
      "step": 246
    },
    {
      "epoch": 2.4231242312423125,
      "grad_norm": 0.06662874441753898,
      "learning_rate": 5.985554537134702e-06,
      "loss": 0.0125,
      "num_tokens": 110456126.0,
      "step": 247
    },
    {
      "epoch": 2.4329643296432963,
      "grad_norm": 0.06935993203158428,
      "learning_rate": 5.957094244021071e-06,
      "loss": 0.0124,
      "num_tokens": 110889342.0,
      "step": 248
    },
    {
      "epoch": 2.4428044280442807,
      "grad_norm": 0.06948156606308496,
      "learning_rate": 5.928615464601497e-06,
      "loss": 0.0143,
      "num_tokens": 111329838.0,
      "step": 249
    },
    {
      "epoch": 2.4526445264452645,
      "grad_norm": 0.08174081643789106,
      "learning_rate": 5.900119350645956e-06,
      "loss": 0.0158,
      "num_tokens": 111789840.0,
      "step": 250
    },
    {
      "epoch": 2.4624846248462484,
      "grad_norm": 0.07150394939916561,
      "learning_rate": 5.871607054625497e-06,
      "loss": 0.0189,
      "num_tokens": 112240250.0,
      "step": 251
    },
    {
      "epoch": 2.4723247232472323,
      "grad_norm": 0.0795332410978433,
      "learning_rate": 5.8430797296656125e-06,
      "loss": 0.0211,
      "num_tokens": 112673933.0,
      "step": 252
    },
    {
      "epoch": 2.4821648216482166,
      "grad_norm": 0.0689891322716241,
      "learning_rate": 5.814538529499622e-06,
      "loss": 0.0148,
      "num_tokens": 113134668.0,
      "step": 253
    },
    {
      "epoch": 2.4920049200492005,
      "grad_norm": 0.07982357004600245,
      "learning_rate": 5.785984608421993e-06,
      "loss": 0.016,
      "num_tokens": 113596645.0,
      "step": 254
    },
    {
      "epoch": 2.5018450184501844,
      "grad_norm": 0.07075114056461625,
      "learning_rate": 5.757419121241667e-06,
      "loss": 0.0184,
      "num_tokens": 114038969.0,
      "step": 255
    },
    {
      "epoch": 2.5116851168511687,
      "grad_norm": 0.07123055043084275,
      "learning_rate": 5.7288432232353615e-06,
      "loss": 0.0172,
      "num_tokens": 114476716.0,
      "step": 256
    },
    {
      "epoch": 2.5215252152521526,
      "grad_norm": 0.06238876313147613,
      "learning_rate": 5.7002580701008325e-06,
      "loss": 0.0134,
      "num_tokens": 114958206.0,
      "step": 257
    },
    {
      "epoch": 2.5313653136531364,
      "grad_norm": 0.06444872364781601,
      "learning_rate": 5.6716648179101445e-06,
      "loss": 0.0122,
      "num_tokens": 115415134.0,
      "step": 258
    },
    {
      "epoch": 2.5412054120541203,
      "grad_norm": 0.06263958198396931,
      "learning_rate": 5.64306462306291e-06,
      "loss": 0.0142,
      "num_tokens": 115898609.0,
      "step": 259
    },
    {
      "epoch": 2.5510455104551046,
      "grad_norm": 0.07604741062856764,
      "learning_rate": 5.614458642239534e-06,
      "loss": 0.0121,
      "num_tokens": 116348874.0,
      "step": 260
    },
    {
      "epoch": 2.5608856088560885,
      "grad_norm": 0.07535603766213132,
      "learning_rate": 5.585848032354411e-06,
      "loss": 0.0187,
      "num_tokens": 116822359.0,
      "step": 261
    },
    {
      "epoch": 2.570725707257073,
      "grad_norm": 0.07043618175110689,
      "learning_rate": 5.557233950509159e-06,
      "loss": 0.0332,
      "num_tokens": 117246814.0,
      "step": 262
    },
    {
      "epoch": 2.5805658056580567,
      "grad_norm": 0.20105726012087,
      "learning_rate": 5.528617553945807e-06,
      "loss": 0.0128,
      "num_tokens": 117708546.0,
      "step": 263
    },
    {
      "epoch": 2.5904059040590406,
      "grad_norm": 0.07781337694485964,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.0138,
      "num_tokens": 118145045.0,
      "step": 264
    },
    {
      "epoch": 2.6002460024600245,
      "grad_norm": 0.06125312067867148,
      "learning_rate": 5.4713824460541964e-06,
      "loss": 0.0122,
      "num_tokens": 118586755.0,
      "step": 265
    },
    {
      "epoch": 2.6100861008610083,
      "grad_norm": 0.06778243090705331,
      "learning_rate": 5.442766049490843e-06,
      "loss": 0.0242,
      "num_tokens": 119025525.0,
      "step": 266
    },
    {
      "epoch": 2.6199261992619927,
      "grad_norm": 0.06153415408443227,
      "learning_rate": 5.414151967645591e-06,
      "loss": 0.0125,
      "num_tokens": 119479378.0,
      "step": 267
    },
    {
      "epoch": 2.6297662976629765,
      "grad_norm": 0.06942223993965259,
      "learning_rate": 5.385541357760469e-06,
      "loss": 0.0243,
      "num_tokens": 119937308.0,
      "step": 268
    },
    {
      "epoch": 2.639606396063961,
      "grad_norm": 0.06214738625923835,
      "learning_rate": 5.35693537693709e-06,
      "loss": 0.0131,
      "num_tokens": 120395822.0,
      "step": 269
    },
    {
      "epoch": 2.6494464944649447,
      "grad_norm": 0.0669104820126879,
      "learning_rate": 5.3283351820898586e-06,
      "loss": 0.013,
      "num_tokens": 120868183.0,
      "step": 270
    },
    {
      "epoch": 2.6592865928659286,
      "grad_norm": 0.0647734711741491,
      "learning_rate": 5.299741929899171e-06,
      "loss": 0.0123,
      "num_tokens": 121321350.0,
      "step": 271
    },
    {
      "epoch": 2.6691266912669125,
      "grad_norm": 0.07530586770614496,
      "learning_rate": 5.27115677676464e-06,
      "loss": 0.0148,
      "num_tokens": 121767518.0,
      "step": 272
    },
    {
      "epoch": 2.678966789667897,
      "grad_norm": 0.060785296202459226,
      "learning_rate": 5.242580878758334e-06,
      "loss": 0.0113,
      "num_tokens": 122214099.0,
      "step": 273
    },
    {
      "epoch": 2.6888068880688807,
      "grad_norm": 0.06051260316507515,
      "learning_rate": 5.21401539157801e-06,
      "loss": 0.0118,
      "num_tokens": 122641818.0,
      "step": 274
    },
    {
      "epoch": 2.6986469864698646,
      "grad_norm": 0.06697146315594861,
      "learning_rate": 5.1854614705003796e-06,
      "loss": 0.0129,
      "num_tokens": 123073540.0,
      "step": 275
    },
    {
      "epoch": 2.708487084870849,
      "grad_norm": 0.06445807540084068,
      "learning_rate": 5.156920270334389e-06,
      "loss": 0.011,
      "num_tokens": 123534092.0,
      "step": 276
    },
    {
      "epoch": 2.7183271832718328,
      "grad_norm": 0.06525833653570826,
      "learning_rate": 5.1283929453745055e-06,
      "loss": 0.0134,
      "num_tokens": 123958744.0,
      "step": 277
    },
    {
      "epoch": 2.7281672816728166,
      "grad_norm": 0.05852868843858968,
      "learning_rate": 5.099880649354044e-06,
      "loss": 0.0111,
      "num_tokens": 124403705.0,
      "step": 278
    },
    {
      "epoch": 2.7380073800738005,
      "grad_norm": 0.07241672197036282,
      "learning_rate": 5.071384535398505e-06,
      "loss": 0.0127,
      "num_tokens": 124889693.0,
      "step": 279
    },
    {
      "epoch": 2.747847478474785,
      "grad_norm": 0.06559889162939202,
      "learning_rate": 5.04290575597893e-06,
      "loss": 0.0141,
      "num_tokens": 125351837.0,
      "step": 280
    },
    {
      "epoch": 2.7576875768757687,
      "grad_norm": 0.06353700618816872,
      "learning_rate": 5.0144454628653015e-06,
      "loss": 0.0119,
      "num_tokens": 125793504.0,
      "step": 281
    },
    {
      "epoch": 2.767527675276753,
      "grad_norm": 0.06683615854915595,
      "learning_rate": 4.986004807079959e-06,
      "loss": 0.0129,
      "num_tokens": 126251447.0,
      "step": 282
    },
    {
      "epoch": 2.777367773677737,
      "grad_norm": 0.07018116547650956,
      "learning_rate": 4.957584938851048e-06,
      "loss": 0.0302,
      "num_tokens": 126702532.0,
      "step": 283
    },
    {
      "epoch": 2.787207872078721,
      "grad_norm": 0.0779053076124061,
      "learning_rate": 4.929187007565996e-06,
      "loss": 0.0135,
      "num_tokens": 127176249.0,
      "step": 284
    },
    {
      "epoch": 2.7970479704797047,
      "grad_norm": 0.05851676194024864,
      "learning_rate": 4.9008121617250425e-06,
      "loss": 0.0119,
      "num_tokens": 127624082.0,
      "step": 285
    },
    {
      "epoch": 2.8068880688806885,
      "grad_norm": 0.06198574153406469,
      "learning_rate": 4.87246154889477e-06,
      "loss": 0.0122,
      "num_tokens": 128044479.0,
      "step": 286
    },
    {
      "epoch": 2.816728167281673,
      "grad_norm": 0.06543548304815791,
      "learning_rate": 4.8441363156617085e-06,
      "loss": 0.0116,
      "num_tokens": 128480925.0,
      "step": 287
    },
    {
      "epoch": 2.8265682656826567,
      "grad_norm": 0.06877942595907044,
      "learning_rate": 4.815837607585957e-06,
      "loss": 0.0129,
      "num_tokens": 128924136.0,
      "step": 288
    },
    {
      "epoch": 2.836408364083641,
      "grad_norm": 0.07130567586435052,
      "learning_rate": 4.787566569154855e-06,
      "loss": 0.014,
      "num_tokens": 129349471.0,
      "step": 289
    },
    {
      "epoch": 2.846248462484625,
      "grad_norm": 0.059470596149140166,
      "learning_rate": 4.759324343736698e-06,
      "loss": 0.0114,
      "num_tokens": 129789536.0,
      "step": 290
    },
    {
      "epoch": 2.856088560885609,
      "grad_norm": 0.06813104970883538,
      "learning_rate": 4.731112073534491e-06,
      "loss": 0.0134,
      "num_tokens": 130231279.0,
      "step": 291
    },
    {
      "epoch": 2.8659286592865927,
      "grad_norm": 0.44224478447995313,
      "learning_rate": 4.70293089953976e-06,
      "loss": 0.2098,
      "num_tokens": 130687907.0,
      "step": 292
    },
    {
      "epoch": 2.875768757687577,
      "grad_norm": 0.06521058331073082,
      "learning_rate": 4.674781961486399e-06,
      "loss": 0.0121,
      "num_tokens": 131139639.0,
      "step": 293
    },
    {
      "epoch": 2.885608856088561,
      "grad_norm": 0.0669166944603376,
      "learning_rate": 4.646666397804586e-06,
      "loss": 0.0115,
      "num_tokens": 131558019.0,
      "step": 294
    },
    {
      "epoch": 2.8954489544895448,
      "grad_norm": 0.05844491203018172,
      "learning_rate": 4.618585345574741e-06,
      "loss": 0.0116,
      "num_tokens": 132015109.0,
      "step": 295
    },
    {
      "epoch": 2.905289052890529,
      "grad_norm": 0.06568591959992806,
      "learning_rate": 4.5905399404815196e-06,
      "loss": 0.0139,
      "num_tokens": 132473250.0,
      "step": 296
    },
    {
      "epoch": 2.915129151291513,
      "grad_norm": 0.0740965654131703,
      "learning_rate": 4.562531316767908e-06,
      "loss": 0.0142,
      "num_tokens": 132924223.0,
      "step": 297
    },
    {
      "epoch": 2.924969249692497,
      "grad_norm": 0.06249511498433694,
      "learning_rate": 4.534560607189338e-06,
      "loss": 0.0116,
      "num_tokens": 133401598.0,
      "step": 298
    },
    {
      "epoch": 2.9348093480934807,
      "grad_norm": 0.06283720689157825,
      "learning_rate": 4.506628942967874e-06,
      "loss": 0.0111,
      "num_tokens": 133858797.0,
      "step": 299
    },
    {
      "epoch": 2.944649446494465,
      "grad_norm": 0.056149678938915315,
      "learning_rate": 4.478737453746464e-06,
      "loss": 0.0104,
      "num_tokens": 134308841.0,
      "step": 300
    },
    {
      "epoch": 2.954489544895449,
      "grad_norm": 0.09917979372278035,
      "learning_rate": 4.450887267543261e-06,
      "loss": 0.016,
      "num_tokens": 134737382.0,
      "step": 301
    },
    {
      "epoch": 2.9643296432964332,
      "grad_norm": 0.0640380581075781,
      "learning_rate": 4.423079510705992e-06,
      "loss": 0.0122,
      "num_tokens": 135161876.0,
      "step": 302
    },
    {
      "epoch": 2.974169741697417,
      "grad_norm": 0.06467022053195733,
      "learning_rate": 4.395315307866404e-06,
      "loss": 0.0165,
      "num_tokens": 135642143.0,
      "step": 303
    },
    {
      "epoch": 2.984009840098401,
      "grad_norm": 0.06355259176583768,
      "learning_rate": 4.3675957818947965e-06,
      "loss": 0.0122,
      "num_tokens": 136085866.0,
      "step": 304
    },
    {
      "epoch": 2.993849938499385,
      "grad_norm": 0.06019771768539855,
      "learning_rate": 4.33992205385459e-06,
      "loss": 0.0192,
      "num_tokens": 136558605.0,
      "step": 305
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.06019771768539855,
      "learning_rate": 4.312295242956998e-06,
      "loss": 0.0196,
      "num_tokens": 136769120.0,
      "step": 306
    },
    {
      "epoch": 3.009840098400984,
      "grad_norm": 0.09281584655295379,
      "learning_rate": 4.284716466515759e-06,
      "loss": 0.015,
      "num_tokens": 137221619.0,
      "step": 307
    },
    {
      "epoch": 3.019680196801968,
      "grad_norm": 0.06328942540239005,
      "learning_rate": 4.257186839901948e-06,
      "loss": 0.0242,
      "num_tokens": 137668031.0,
      "step": 308
    },
    {
      "epoch": 3.029520295202952,
      "grad_norm": 0.1013363943890701,
      "learning_rate": 4.229707476498871e-06,
      "loss": 0.0098,
      "num_tokens": 138123076.0,
      "step": 309
    },
    {
      "epoch": 3.039360393603936,
      "grad_norm": 0.05368561892248246,
      "learning_rate": 4.2022794876570335e-06,
      "loss": 0.0092,
      "num_tokens": 138584469.0,
      "step": 310
    },
    {
      "epoch": 3.0492004920049203,
      "grad_norm": 0.055594058845563385,
      "learning_rate": 4.1749039826491956e-06,
      "loss": 0.0097,
      "num_tokens": 139040295.0,
      "step": 311
    },
    {
      "epoch": 3.059040590405904,
      "grad_norm": 0.061365365973882896,
      "learning_rate": 4.1475820686255055e-06,
      "loss": 0.0112,
      "num_tokens": 139467398.0,
      "step": 312
    },
    {
      "epoch": 3.068880688806888,
      "grad_norm": 0.07576400254078287,
      "learning_rate": 4.120314850568731e-06,
      "loss": 0.0112,
      "num_tokens": 139915959.0,
      "step": 313
    },
    {
      "epoch": 3.078720787207872,
      "grad_norm": 0.06609887375125617,
      "learning_rate": 4.093103431249563e-06,
      "loss": 0.0093,
      "num_tokens": 140383093.0,
      "step": 314
    },
    {
      "epoch": 3.088560885608856,
      "grad_norm": 0.06075307137850355,
      "learning_rate": 4.065948911182015e-06,
      "loss": 0.0099,
      "num_tokens": 140822538.0,
      "step": 315
    },
    {
      "epoch": 3.09840098400984,
      "grad_norm": 0.062126435434214644,
      "learning_rate": 4.038852388578925e-06,
      "loss": 0.0099,
      "num_tokens": 141294868.0,
      "step": 316
    },
    {
      "epoch": 3.108241082410824,
      "grad_norm": 0.06817279237293263,
      "learning_rate": 4.011814959307533e-06,
      "loss": 0.0168,
      "num_tokens": 141736810.0,
      "step": 317
    },
    {
      "epoch": 3.1180811808118083,
      "grad_norm": 0.06959931879800538,
      "learning_rate": 3.984837716845157e-06,
      "loss": 0.0098,
      "num_tokens": 142184603.0,
      "step": 318
    },
    {
      "epoch": 3.127921279212792,
      "grad_norm": 0.06353490194683095,
      "learning_rate": 3.957921752234982e-06,
      "loss": 0.0167,
      "num_tokens": 142622524.0,
      "step": 319
    },
    {
      "epoch": 3.137761377613776,
      "grad_norm": 0.06947695018548046,
      "learning_rate": 3.931068154041919e-06,
      "loss": 0.0097,
      "num_tokens": 143083493.0,
      "step": 320
    },
    {
      "epoch": 3.14760147601476,
      "grad_norm": 0.059043509631747315,
      "learning_rate": 3.904278008308589e-06,
      "loss": 0.0104,
      "num_tokens": 143533589.0,
      "step": 321
    },
    {
      "epoch": 3.1574415744157442,
      "grad_norm": 0.06496286878562947,
      "learning_rate": 3.877552398511409e-06,
      "loss": 0.0096,
      "num_tokens": 143992909.0,
      "step": 322
    },
    {
      "epoch": 3.167281672816728,
      "grad_norm": 0.05772743455179762,
      "learning_rate": 3.85089240551675e-06,
      "loss": 0.0116,
      "num_tokens": 144432232.0,
      "step": 323
    },
    {
      "epoch": 3.177121771217712,
      "grad_norm": 0.06067527859066343,
      "learning_rate": 3.8242991075372436e-06,
      "loss": 0.0095,
      "num_tokens": 144872986.0,
      "step": 324
    },
    {
      "epoch": 3.1869618696186963,
      "grad_norm": 0.06092970409513084,
      "learning_rate": 3.7977735800881687e-06,
      "loss": 0.0096,
      "num_tokens": 145338330.0,
      "step": 325
    },
    {
      "epoch": 3.19680196801968,
      "grad_norm": 0.0599181858398082,
      "learning_rate": 3.7713168959439515e-06,
      "loss": 0.0105,
      "num_tokens": 145779615.0,
      "step": 326
    },
    {
      "epoch": 3.206642066420664,
      "grad_norm": 0.06222020359396952,
      "learning_rate": 3.74493012509478e-06,
      "loss": 0.0111,
      "num_tokens": 146242526.0,
      "step": 327
    },
    {
      "epoch": 3.2164821648216484,
      "grad_norm": 0.06718711950500762,
      "learning_rate": 3.718614334703339e-06,
      "loss": 0.011,
      "num_tokens": 146706718.0,
      "step": 328
    },
    {
      "epoch": 3.2263222632226323,
      "grad_norm": 0.06582492337813067,
      "learning_rate": 3.692370589061639e-06,
      "loss": 0.0176,
      "num_tokens": 147136544.0,
      "step": 329
    },
    {
      "epoch": 3.236162361623616,
      "grad_norm": 0.11649214601244208,
      "learning_rate": 3.6661999495479772e-06,
      "loss": 0.0143,
      "num_tokens": 147581972.0,
      "step": 330
    },
    {
      "epoch": 3.2460024600246005,
      "grad_norm": 0.07623567009845389,
      "learning_rate": 3.640103474584016e-06,
      "loss": 0.0109,
      "num_tokens": 148046352.0,
      "step": 331
    },
    {
      "epoch": 3.2558425584255843,
      "grad_norm": 0.10991461635206767,
      "learning_rate": 3.614082219591972e-06,
      "loss": 0.0125,
      "num_tokens": 148512074.0,
      "step": 332
    },
    {
      "epoch": 3.265682656826568,
      "grad_norm": 0.0553413140200664,
      "learning_rate": 3.588137236951934e-06,
      "loss": 0.0112,
      "num_tokens": 148974289.0,
      "step": 333
    },
    {
      "epoch": 3.275522755227552,
      "grad_norm": 0.0647893470114748,
      "learning_rate": 3.5622695759592996e-06,
      "loss": 0.0093,
      "num_tokens": 149432821.0,
      "step": 334
    },
    {
      "epoch": 3.2853628536285364,
      "grad_norm": 0.05909492699311915,
      "learning_rate": 3.5364802827823397e-06,
      "loss": 0.0095,
      "num_tokens": 149885306.0,
      "step": 335
    },
    {
      "epoch": 3.2952029520295203,
      "grad_norm": 0.06091784855812255,
      "learning_rate": 3.5107704004198904e-06,
      "loss": 0.012,
      "num_tokens": 150329823.0,
      "step": 336
    },
    {
      "epoch": 3.305043050430504,
      "grad_norm": 0.0653285630713288,
      "learning_rate": 3.485140968659166e-06,
      "loss": 0.0092,
      "num_tokens": 150762193.0,
      "step": 337
    },
    {
      "epoch": 3.3148831488314885,
      "grad_norm": 0.059323407369363576,
      "learning_rate": 3.4595930240337115e-06,
      "loss": 0.0095,
      "num_tokens": 151238234.0,
      "step": 338
    },
    {
      "epoch": 3.3247232472324724,
      "grad_norm": 0.057695587877706345,
      "learning_rate": 3.4341275997814795e-06,
      "loss": 0.0088,
      "num_tokens": 151665989.0,
      "step": 339
    },
    {
      "epoch": 3.3345633456334562,
      "grad_norm": 0.05732212076245395,
      "learning_rate": 3.408745725803042e-06,
      "loss": 0.0098,
      "num_tokens": 152102577.0,
      "step": 340
    },
    {
      "epoch": 3.34440344403444,
      "grad_norm": 0.0595487114131121,
      "learning_rate": 3.383448428619941e-06,
      "loss": 0.0097,
      "num_tokens": 152531136.0,
      "step": 341
    },
    {
      "epoch": 3.3542435424354244,
      "grad_norm": 0.06136757636985544,
      "learning_rate": 3.3582367313331692e-06,
      "loss": 0.0096,
      "num_tokens": 152996747.0,
      "step": 342
    },
    {
      "epoch": 3.3640836408364083,
      "grad_norm": 0.05958558106516815,
      "learning_rate": 3.3331116535817974e-06,
      "loss": 0.0095,
      "num_tokens": 153435759.0,
      "step": 343
    },
    {
      "epoch": 3.373923739237392,
      "grad_norm": 0.05874672952511224,
      "learning_rate": 3.308074211501732e-06,
      "loss": 0.0095,
      "num_tokens": 153870082.0,
      "step": 344
    },
    {
      "epoch": 3.3837638376383765,
      "grad_norm": 0.06528615293660638,
      "learning_rate": 3.2831254176846205e-06,
      "loss": 0.0099,
      "num_tokens": 154317864.0,
      "step": 345
    },
    {
      "epoch": 3.3936039360393604,
      "grad_norm": 0.06940648713659847,
      "learning_rate": 3.258266281136905e-06,
      "loss": 0.0152,
      "num_tokens": 154756769.0,
      "step": 346
    },
    {
      "epoch": 3.4034440344403443,
      "grad_norm": 0.061886970984610115,
      "learning_rate": 3.233497807239008e-06,
      "loss": 0.0097,
      "num_tokens": 155205894.0,
      "step": 347
    },
    {
      "epoch": 3.4132841328413286,
      "grad_norm": 0.06025802060896222,
      "learning_rate": 3.2088209977046657e-06,
      "loss": 0.0104,
      "num_tokens": 155650235.0,
      "step": 348
    },
    {
      "epoch": 3.4231242312423125,
      "grad_norm": 0.079505096478584,
      "learning_rate": 3.1842368505404388e-06,
      "loss": 0.0093,
      "num_tokens": 156097609.0,
      "step": 349
    },
    {
      "epoch": 3.4329643296432963,
      "grad_norm": 0.05748478767611536,
      "learning_rate": 3.1597463600053258e-06,
      "loss": 0.0099,
      "num_tokens": 156538234.0,
      "step": 350
    },
    {
      "epoch": 3.4428044280442807,
      "grad_norm": 0.06024044990886995,
      "learning_rate": 3.135350516570559e-06,
      "loss": 0.0089,
      "num_tokens": 156983361.0,
      "step": 351
    },
    {
      "epoch": 3.4526445264452645,
      "grad_norm": 0.05743125939934545,
      "learning_rate": 3.111050306879556e-06,
      "loss": 0.0092,
      "num_tokens": 157418360.0,
      "step": 352
    },
    {
      "epoch": 3.4624846248462484,
      "grad_norm": 0.05732681074233502,
      "learning_rate": 3.0868467137080075e-06,
      "loss": 0.0095,
      "num_tokens": 157832487.0,
      "step": 353
    },
    {
      "epoch": 3.4723247232472323,
      "grad_norm": 0.05983498966527194,
      "learning_rate": 3.0627407159241273e-06,
      "loss": 0.0134,
      "num_tokens": 158300222.0,
      "step": 354
    },
    {
      "epoch": 3.4821648216482166,
      "grad_norm": 0.11937622237492422,
      "learning_rate": 3.0387332884490806e-06,
      "loss": 0.0101,
      "num_tokens": 158757341.0,
      "step": 355
    },
    {
      "epoch": 3.4920049200492005,
      "grad_norm": 0.06131645683210505,
      "learning_rate": 3.014825402217533e-06,
      "loss": 0.0162,
      "num_tokens": 159204107.0,
      "step": 356
    },
    {
      "epoch": 3.5018450184501844,
      "grad_norm": 0.07487264837463524,
      "learning_rate": 2.9910180241384014e-06,
      "loss": 0.0094,
      "num_tokens": 159649846.0,
      "step": 357
    },
    {
      "epoch": 3.5116851168511687,
      "grad_norm": 0.05978234516994627,
      "learning_rate": 2.9673121170557396e-06,
      "loss": 0.0093,
      "num_tokens": 160075112.0,
      "step": 358
    },
    {
      "epoch": 3.5215252152521526,
      "grad_norm": 0.05949296353246161,
      "learning_rate": 2.9437086397097996e-06,
      "loss": 0.0654,
      "num_tokens": 160529018.0,
      "step": 359
    },
    {
      "epoch": 3.5313653136531364,
      "grad_norm": 0.49227073434692264,
      "learning_rate": 2.92020854669826e-06,
      "loss": 0.0099,
      "num_tokens": 160983786.0,
      "step": 360
    },
    {
      "epoch": 3.5412054120541203,
      "grad_norm": 0.06622940201032448,
      "learning_rate": 2.896812788437615e-06,
      "loss": 0.01,
      "num_tokens": 161416791.0,
      "step": 361
    },
    {
      "epoch": 3.5510455104551046,
      "grad_norm": 0.06226773442108338,
      "learning_rate": 2.8735223111247402e-06,
      "loss": 0.0086,
      "num_tokens": 161878966.0,
      "step": 362
    },
    {
      "epoch": 3.5608856088560885,
      "grad_norm": 0.052597494365304226,
      "learning_rate": 2.850338056698621e-06,
      "loss": 0.0105,
      "num_tokens": 162327927.0,
      "step": 363
    },
    {
      "epoch": 3.570725707257073,
      "grad_norm": 0.07012270383857575,
      "learning_rate": 2.827260962802263e-06,
      "loss": 0.0147,
      "num_tokens": 162786081.0,
      "step": 364
    },
    {
      "epoch": 3.5805658056580567,
      "grad_norm": 0.055063275741049224,
      "learning_rate": 2.804291962744768e-06,
      "loss": 0.0094,
      "num_tokens": 163225787.0,
      "step": 365
    },
    {
      "epoch": 3.5904059040590406,
      "grad_norm": 0.05925152600247852,
      "learning_rate": 2.7814319854635875e-06,
      "loss": 0.0103,
      "num_tokens": 163657501.0,
      "step": 366
    },
    {
      "epoch": 3.6002460024600245,
      "grad_norm": 0.06722393012249571,
      "learning_rate": 2.758681955486955e-06,
      "loss": 0.0144,
      "num_tokens": 164141801.0,
      "step": 367
    },
    {
      "epoch": 3.6100861008610083,
      "grad_norm": 0.06258536118771861,
      "learning_rate": 2.736042792896495e-06,
      "loss": 0.0095,
      "num_tokens": 164583522.0,
      "step": 368
    },
    {
      "epoch": 3.6199261992619927,
      "grad_norm": 0.058745565835135405,
      "learning_rate": 2.7135154132900133e-06,
      "loss": 0.2081,
      "num_tokens": 165059903.0,
      "step": 369
    },
    {
      "epoch": 3.6297662976629765,
      "grad_norm": 0.45572028974101275,
      "learning_rate": 2.691100727744458e-06,
      "loss": 0.0093,
      "num_tokens": 165501531.0,
      "step": 370
    },
    {
      "epoch": 3.639606396063961,
      "grad_norm": 0.06413891247366597,
      "learning_rate": 2.668799642779093e-06,
      "loss": 0.0092,
      "num_tokens": 165943767.0,
      "step": 371
    },
    {
      "epoch": 3.6494464944649447,
      "grad_norm": 0.06190519506649364,
      "learning_rate": 2.6466130603188157e-06,
      "loss": 0.0093,
      "num_tokens": 166367316.0,
      "step": 372
    },
    {
      "epoch": 3.6592865928659286,
      "grad_norm": 0.05986849692187042,
      "learning_rate": 2.624541877657685e-06,
      "loss": 0.0089,
      "num_tokens": 166834863.0,
      "step": 373
    },
    {
      "epoch": 3.6691266912669125,
      "grad_norm": 0.05660848820396227,
      "learning_rate": 2.602586987422643e-06,
      "loss": 0.0091,
      "num_tokens": 167276182.0,
      "step": 374
    },
    {
      "epoch": 3.678966789667897,
      "grad_norm": 0.06018671949798086,
      "learning_rate": 2.580749277537399e-06,
      "loss": 0.0089,
      "num_tokens": 167713298.0,
      "step": 375
    },
    {
      "epoch": 3.6888068880688807,
      "grad_norm": 0.05866547399139435,
      "learning_rate": 2.5590296311865294e-06,
      "loss": 0.0089,
      "num_tokens": 168174626.0,
      "step": 376
    },
    {
      "epoch": 3.6986469864698646,
      "grad_norm": 0.05484667079392053,
      "learning_rate": 2.537428926779758e-06,
      "loss": 0.0084,
      "num_tokens": 168614618.0,
      "step": 377
    },
    {
      "epoch": 3.708487084870849,
      "grad_norm": 0.05383142631084436,
      "learning_rate": 2.515948037916423e-06,
      "loss": 0.0093,
      "num_tokens": 169020836.0,
      "step": 378
    },
    {
      "epoch": 3.7183271832718328,
      "grad_norm": 0.06002011881446048,
      "learning_rate": 2.494587833350153e-06,
      "loss": 0.0087,
      "num_tokens": 169486382.0,
      "step": 379
    },
    {
      "epoch": 3.7281672816728166,
      "grad_norm": 0.05744381429509693,
      "learning_rate": 2.473349176953736e-06,
      "loss": 0.0095,
      "num_tokens": 169927283.0,
      "step": 380
    },
    {
      "epoch": 3.7380073800738005,
      "grad_norm": 0.060208708524887654,
      "learning_rate": 2.4522329276841664e-06,
      "loss": 0.0199,
      "num_tokens": 170406677.0,
      "step": 381
    },
    {
      "epoch": 3.747847478474785,
      "grad_norm": 0.07568634621301189,
      "learning_rate": 2.431239939547921e-06,
      "loss": 0.0091,
      "num_tokens": 170848564.0,
      "step": 382
    },
    {
      "epoch": 3.7576875768757687,
      "grad_norm": 0.05568928633830086,
      "learning_rate": 2.4103710615664145e-06,
      "loss": 0.0087,
      "num_tokens": 171300323.0,
      "step": 383
    },
    {
      "epoch": 3.767527675276753,
      "grad_norm": 0.05797358351819658,
      "learning_rate": 2.389627137741662e-06,
      "loss": 0.009,
      "num_tokens": 171756442.0,
      "step": 384
    },
    {
      "epoch": 3.777367773677737,
      "grad_norm": 0.05921869424481757,
      "learning_rate": 2.369009007022146e-06,
      "loss": 0.0093,
      "num_tokens": 172205755.0,
      "step": 385
    },
    {
      "epoch": 3.787207872078721,
      "grad_norm": 0.05977666911207599,
      "learning_rate": 2.3485175032688865e-06,
      "loss": 0.0093,
      "num_tokens": 172657182.0,
      "step": 386
    },
    {
      "epoch": 3.7970479704797047,
      "grad_norm": 0.05932546877174744,
      "learning_rate": 2.328153455221717e-06,
      "loss": 0.0093,
      "num_tokens": 173095889.0,
      "step": 387
    },
    {
      "epoch": 3.8068880688806885,
      "grad_norm": 0.06042663443540837,
      "learning_rate": 2.3079176864657673e-06,
      "loss": 0.0126,
      "num_tokens": 173547333.0,
      "step": 388
    },
    {
      "epoch": 3.816728167281673,
      "grad_norm": 0.15319976378374434,
      "learning_rate": 2.2878110153981565e-06,
      "loss": 0.0127,
      "num_tokens": 174010201.0,
      "step": 389
    },
    {
      "epoch": 3.8265682656826567,
      "grad_norm": 0.06053526576608028,
      "learning_rate": 2.267834255194894e-06,
      "loss": 0.0091,
      "num_tokens": 174451094.0,
      "step": 390
    },
    {
      "epoch": 3.836408364083641,
      "grad_norm": 0.056105090205726134,
      "learning_rate": 2.2479882137779903e-06,
      "loss": 0.0086,
      "num_tokens": 174912611.0,
      "step": 391
    },
    {
      "epoch": 3.846248462484625,
      "grad_norm": 0.05451931809215225,
      "learning_rate": 2.228273693782784e-06,
      "loss": 0.181,
      "num_tokens": 175406884.0,
      "step": 392
    },
    {
      "epoch": 3.856088560885609,
      "grad_norm": 0.21723127868567568,
      "learning_rate": 2.208691492525481e-06,
      "loss": 0.0201,
      "num_tokens": 175906140.0,
      "step": 393
    },
    {
      "epoch": 3.8659286592865927,
      "grad_norm": 0.0674311563870576,
      "learning_rate": 2.189242401970908e-06,
      "loss": 0.0109,
      "num_tokens": 176321003.0,
      "step": 394
    },
    {
      "epoch": 3.875768757687577,
      "grad_norm": 0.06352093720920256,
      "learning_rate": 2.169927208700482e-06,
      "loss": 0.0105,
      "num_tokens": 176774697.0,
      "step": 395
    },
    {
      "epoch": 3.885608856088561,
      "grad_norm": 0.07829494568911512,
      "learning_rate": 2.1507466938804013e-06,
      "loss": 0.009,
      "num_tokens": 177239096.0,
      "step": 396
    },
    {
      "epoch": 3.8954489544895448,
      "grad_norm": 0.05295977087497385,
      "learning_rate": 2.131701633230045e-06,
      "loss": 0.0108,
      "num_tokens": 177705420.0,
      "step": 397
    },
    {
      "epoch": 3.905289052890529,
      "grad_norm": 0.06253022869487425,
      "learning_rate": 2.112792796990616e-06,
      "loss": 0.011,
      "num_tokens": 178159440.0,
      "step": 398
    },
    {
      "epoch": 3.915129151291513,
      "grad_norm": 0.06311743420915392,
      "learning_rate": 2.0940209498939732e-06,
      "loss": 0.0087,
      "num_tokens": 178600759.0,
      "step": 399
    },
    {
      "epoch": 3.924969249692497,
      "grad_norm": 0.05932091295852782,
      "learning_rate": 2.075386851131711e-06,
      "loss": 0.022,
      "num_tokens": 179053675.0,
      "step": 400
    },
    {
      "epoch": 3.9348093480934807,
      "grad_norm": 0.07035371257735856,
      "learning_rate": 2.056891254324459e-06,
      "loss": 0.0099,
      "num_tokens": 179505855.0,
      "step": 401
    },
    {
      "epoch": 3.944649446494465,
      "grad_norm": 0.06484414999103219,
      "learning_rate": 2.038534907491396e-06,
      "loss": 0.0088,
      "num_tokens": 179936013.0,
      "step": 402
    },
    {
      "epoch": 3.954489544895449,
      "grad_norm": 0.060396196023032395,
      "learning_rate": 2.0203185530199983e-06,
      "loss": 0.0088,
      "num_tokens": 180382727.0,
      "step": 403
    },
    {
      "epoch": 3.9643296432964332,
      "grad_norm": 0.05481446314958498,
      "learning_rate": 2.0022429276360256e-06,
      "loss": 0.0083,
      "num_tokens": 180829894.0,
      "step": 404
    },
    {
      "epoch": 3.974169741697417,
      "grad_norm": 0.05579653515238233,
      "learning_rate": 1.9843087623737097e-06,
      "loss": 0.0084,
      "num_tokens": 181285317.0,
      "step": 405
    },
    {
      "epoch": 3.984009840098401,
      "grad_norm": 0.08234263765839406,
      "learning_rate": 1.966516782546199e-06,
      "loss": 0.0147,
      "num_tokens": 181726465.0,
      "step": 406
    },
    {
      "epoch": 3.993849938499385,
      "grad_norm": 0.061989197697927254,
      "learning_rate": 1.94886770771623e-06,
      "loss": 0.021,
      "num_tokens": 182137592.0,
      "step": 407
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.06561678793095134,
      "learning_rate": 1.931362251667008e-06,
      "loss": 0.01,
      "num_tokens": 182350141.0,
      "step": 408
    },
    {
      "epoch": 4.009840098400984,
      "grad_norm": 0.08249843123562729,
      "learning_rate": 1.9140011223733576e-06,
      "loss": 0.0077,
      "num_tokens": 182808332.0,
      "step": 409
    },
    {
      "epoch": 4.019680196801968,
      "grad_norm": 0.051015416231373206,
      "learning_rate": 1.8967850219730799e-06,
      "loss": 0.0081,
      "num_tokens": 183234541.0,
      "step": 410
    },
    {
      "epoch": 4.029520295202952,
      "grad_norm": 0.05624434271690936,
      "learning_rate": 1.8797146467385604e-06,
      "loss": 0.008,
      "num_tokens": 183657861.0,
      "step": 411
    },
    {
      "epoch": 4.039360393603936,
      "grad_norm": 0.055439304665308706,
      "learning_rate": 1.8627906870486063e-06,
      "loss": 0.0076,
      "num_tokens": 184114899.0,
      "step": 412
    },
    {
      "epoch": 4.04920049200492,
      "grad_norm": 0.053200752263272366,
      "learning_rate": 1.8460138273605265e-06,
      "loss": 0.0075,
      "num_tokens": 184540917.0,
      "step": 413
    },
    {
      "epoch": 4.059040590405904,
      "grad_norm": 0.05493004000264093,
      "learning_rate": 1.8293847461824538e-06,
      "loss": 0.0182,
      "num_tokens": 184993388.0,
      "step": 414
    },
    {
      "epoch": 4.068880688806888,
      "grad_norm": 0.08372361370144696,
      "learning_rate": 1.8129041160458966e-06,
      "loss": 0.0136,
      "num_tokens": 185443466.0,
      "step": 415
    },
    {
      "epoch": 4.078720787207872,
      "grad_norm": 0.05966232822391394,
      "learning_rate": 1.7965726034785466e-06,
      "loss": 0.0081,
      "num_tokens": 185884446.0,
      "step": 416
    },
    {
      "epoch": 4.088560885608856,
      "grad_norm": 0.05964634351715403,
      "learning_rate": 1.780390868977318e-06,
      "loss": 0.0109,
      "num_tokens": 186324584.0,
      "step": 417
    },
    {
      "epoch": 4.0984009840098405,
      "grad_norm": 0.077920426753446,
      "learning_rate": 1.7643595669816378e-06,
      "loss": 0.0125,
      "num_tokens": 186805249.0,
      "step": 418
    },
    {
      "epoch": 4.108241082410824,
      "grad_norm": 0.052268159950947905,
      "learning_rate": 1.7484793458469745e-06,
      "loss": 0.0075,
      "num_tokens": 187242236.0,
      "step": 419
    },
    {
      "epoch": 4.118081180811808,
      "grad_norm": 0.055826923432114865,
      "learning_rate": 1.7327508478186216e-06,
      "loss": 0.0072,
      "num_tokens": 187710452.0,
      "step": 420
    },
    {
      "epoch": 4.127921279212792,
      "grad_norm": 0.05075579193844761,
      "learning_rate": 1.7171747090057201e-06,
      "loss": 0.0073,
      "num_tokens": 188166192.0,
      "step": 421
    },
    {
      "epoch": 4.137761377613776,
      "grad_norm": 0.05526562542826341,
      "learning_rate": 1.7017515593555295e-06,
      "loss": 0.0075,
      "num_tokens": 188609825.0,
      "step": 422
    },
    {
      "epoch": 4.14760147601476,
      "grad_norm": 0.05657081985879626,
      "learning_rate": 1.6864820226279607e-06,
      "loss": 0.0074,
      "num_tokens": 189068636.0,
      "step": 423
    },
    {
      "epoch": 4.157441574415744,
      "grad_norm": 0.05470927216680097,
      "learning_rate": 1.6713667163703348e-06,
      "loss": 0.008,
      "num_tokens": 189501962.0,
      "step": 424
    },
    {
      "epoch": 4.167281672816729,
      "grad_norm": 0.06396161695980077,
      "learning_rate": 1.6564062518924202e-06,
      "loss": 0.0134,
      "num_tokens": 189933564.0,
      "step": 425
    },
    {
      "epoch": 4.177121771217712,
      "grad_norm": 0.06708045538135683,
      "learning_rate": 1.6416012342417056e-06,
      "loss": 0.013,
      "num_tokens": 190376512.0,
      "step": 426
    },
    {
      "epoch": 4.186961869618696,
      "grad_norm": 0.06902938106259363,
      "learning_rate": 1.6269522621789246e-06,
      "loss": 0.0185,
      "num_tokens": 190824286.0,
      "step": 427
    },
    {
      "epoch": 4.19680196801968,
      "grad_norm": 0.06764173670337073,
      "learning_rate": 1.6124599281538452e-06,
      "loss": 0.023,
      "num_tokens": 191293501.0,
      "step": 428
    },
    {
      "epoch": 4.206642066420664,
      "grad_norm": 0.055114986132894145,
      "learning_rate": 1.5981248182813136e-06,
      "loss": 0.0073,
      "num_tokens": 191738984.0,
      "step": 429
    },
    {
      "epoch": 4.216482164821648,
      "grad_norm": 0.051261855942856496,
      "learning_rate": 1.583947512317537e-06,
      "loss": 0.0075,
      "num_tokens": 192188472.0,
      "step": 430
    },
    {
      "epoch": 4.226322263222632,
      "grad_norm": 0.05658108779272827,
      "learning_rate": 1.5699285836366488e-06,
      "loss": 0.0077,
      "num_tokens": 192610188.0,
      "step": 431
    },
    {
      "epoch": 4.236162361623617,
      "grad_norm": 0.05505791232716452,
      "learning_rate": 1.5560685992075141e-06,
      "loss": 0.0074,
      "num_tokens": 193063967.0,
      "step": 432
    },
    {
      "epoch": 4.2460024600246005,
      "grad_norm": 0.054672853412556996,
      "learning_rate": 1.5423681195707997e-06,
      "loss": 0.0072,
      "num_tokens": 193546556.0,
      "step": 433
    },
    {
      "epoch": 4.255842558425584,
      "grad_norm": 0.05581920906854521,
      "learning_rate": 1.528827698816306e-06,
      "loss": 0.0077,
      "num_tokens": 193976678.0,
      "step": 434
    },
    {
      "epoch": 4.265682656826568,
      "grad_norm": 0.05717281375700824,
      "learning_rate": 1.515447884560556e-06,
      "loss": 0.0074,
      "num_tokens": 194451959.0,
      "step": 435
    },
    {
      "epoch": 4.275522755227552,
      "grad_norm": 0.05353729617839062,
      "learning_rate": 1.502229217924649e-06,
      "loss": 0.0075,
      "num_tokens": 194919769.0,
      "step": 436
    },
    {
      "epoch": 4.285362853628536,
      "grad_norm": 0.05747591456155956,
      "learning_rate": 1.489172233512376e-06,
      "loss": 0.0073,
      "num_tokens": 195385061.0,
      "step": 437
    },
    {
      "epoch": 4.29520295202952,
      "grad_norm": 0.0605171649231424,
      "learning_rate": 1.4762774593885986e-06,
      "loss": 0.008,
      "num_tokens": 195829302.0,
      "step": 438
    },
    {
      "epoch": 4.305043050430505,
      "grad_norm": 0.057478821326489744,
      "learning_rate": 1.4635454170578917e-06,
      "loss": 0.0082,
      "num_tokens": 196272086.0,
      "step": 439
    },
    {
      "epoch": 4.3148831488314885,
      "grad_norm": 0.05677684124604302,
      "learning_rate": 1.4509766214434535e-06,
      "loss": 0.0074,
      "num_tokens": 196705376.0,
      "step": 440
    },
    {
      "epoch": 4.324723247232472,
      "grad_norm": 0.05308761925199772,
      "learning_rate": 1.4385715808662787e-06,
      "loss": 0.0077,
      "num_tokens": 197144202.0,
      "step": 441
    },
    {
      "epoch": 4.334563345633456,
      "grad_norm": 0.0537126730210605,
      "learning_rate": 1.4263307970246027e-06,
      "loss": 0.0072,
      "num_tokens": 197583484.0,
      "step": 442
    },
    {
      "epoch": 4.34440344403444,
      "grad_norm": 0.0581676639275268,
      "learning_rate": 1.41425476497361e-06,
      "loss": 0.01,
      "num_tokens": 198038237.0,
      "step": 443
    },
    {
      "epoch": 4.354243542435424,
      "grad_norm": 0.05692004762795615,
      "learning_rate": 1.4023439731054112e-06,
      "loss": 0.0076,
      "num_tokens": 198478749.0,
      "step": 444
    },
    {
      "epoch": 4.364083640836409,
      "grad_norm": 0.05533945838827064,
      "learning_rate": 1.390598903129296e-06,
      "loss": 0.0075,
      "num_tokens": 198919863.0,
      "step": 445
    },
    {
      "epoch": 4.373923739237393,
      "grad_norm": 0.08520786010753936,
      "learning_rate": 1.3790200300522413e-06,
      "loss": 0.021,
      "num_tokens": 199382929.0,
      "step": 446
    },
    {
      "epoch": 4.3837638376383765,
      "grad_norm": 0.05139822843194828,
      "learning_rate": 1.3676078221597157e-06,
      "loss": 0.0075,
      "num_tokens": 199843263.0,
      "step": 447
    },
    {
      "epoch": 4.39360393603936,
      "grad_norm": 0.31624336801190883,
      "learning_rate": 1.3563627409967257e-06,
      "loss": 0.0554,
      "num_tokens": 200287378.0,
      "step": 448
    },
    {
      "epoch": 4.403444034440344,
      "grad_norm": 0.05264132661612972,
      "learning_rate": 1.3452852413491563e-06,
      "loss": 0.0065,
      "num_tokens": 200793892.0,
      "step": 449
    },
    {
      "epoch": 4.413284132841328,
      "grad_norm": 0.05900297532446137,
      "learning_rate": 1.3343757712253804e-06,
      "loss": 0.0093,
      "num_tokens": 201237298.0,
      "step": 450
    },
    {
      "epoch": 4.423124231242312,
      "grad_norm": 0.28018993976510653,
      "learning_rate": 1.3236347718381338e-06,
      "loss": 0.1846,
      "num_tokens": 201713169.0,
      "step": 451
    },
    {
      "epoch": 4.432964329643297,
      "grad_norm": 0.056807303328480156,
      "learning_rate": 1.3130626775866743e-06,
      "loss": 0.0075,
      "num_tokens": 202186125.0,
      "step": 452
    },
    {
      "epoch": 4.442804428044281,
      "grad_norm": 0.050176964423526066,
      "learning_rate": 1.3026599160392173e-06,
      "loss": 0.0069,
      "num_tokens": 202642625.0,
      "step": 453
    },
    {
      "epoch": 4.4526445264452645,
      "grad_norm": 0.09261733517736204,
      "learning_rate": 1.292426907915634e-06,
      "loss": 0.0116,
      "num_tokens": 203084429.0,
      "step": 454
    },
    {
      "epoch": 4.462484624846248,
      "grad_norm": 0.05582692526106092,
      "learning_rate": 1.2823640670704443e-06,
      "loss": 0.0074,
      "num_tokens": 203531332.0,
      "step": 455
    },
    {
      "epoch": 4.472324723247232,
      "grad_norm": 0.0544386302493359,
      "learning_rate": 1.2724718004760794e-06,
      "loss": 0.007,
      "num_tokens": 203977923.0,
      "step": 456
    },
    {
      "epoch": 4.482164821648216,
      "grad_norm": 0.0485924079874789,
      "learning_rate": 1.2627505082064144e-06,
      "loss": 0.0075,
      "num_tokens": 204447835.0,
      "step": 457
    },
    {
      "epoch": 4.492004920049201,
      "grad_norm": 0.05314912137402164,
      "learning_rate": 1.2532005834205976e-06,
      "loss": 0.0077,
      "num_tokens": 204889787.0,
      "step": 458
    },
    {
      "epoch": 4.501845018450185,
      "grad_norm": 0.059487474238797314,
      "learning_rate": 1.2438224123471442e-06,
      "loss": 0.0081,
      "num_tokens": 205319753.0,
      "step": 459
    },
    {
      "epoch": 4.511685116851169,
      "grad_norm": 0.053210906319883816,
      "learning_rate": 1.2346163742683185e-06,
      "loss": 0.0075,
      "num_tokens": 205765375.0,
      "step": 460
    },
    {
      "epoch": 4.521525215252153,
      "grad_norm": 0.052123105650746834,
      "learning_rate": 1.2255828415047932e-06,
      "loss": 0.0215,
      "num_tokens": 206223886.0,
      "step": 461
    },
    {
      "epoch": 4.531365313653136,
      "grad_norm": 0.058022484093730155,
      "learning_rate": 1.216722179400592e-06,
      "loss": 0.0079,
      "num_tokens": 206657932.0,
      "step": 462
    },
    {
      "epoch": 4.54120541205412,
      "grad_norm": 0.06046936274806646,
      "learning_rate": 1.208034746308315e-06,
      "loss": 0.0119,
      "num_tokens": 207113562.0,
      "step": 463
    },
    {
      "epoch": 4.551045510455104,
      "grad_norm": 0.05202796714609154,
      "learning_rate": 1.1995208935746437e-06,
      "loss": 0.0073,
      "num_tokens": 207577701.0,
      "step": 464
    },
    {
      "epoch": 4.560885608856088,
      "grad_norm": 0.053010352876086804,
      "learning_rate": 1.1911809655261333e-06,
      "loss": 0.0071,
      "num_tokens": 208021678.0,
      "step": 465
    },
    {
      "epoch": 4.570725707257073,
      "grad_norm": 0.05123561999347148,
      "learning_rate": 1.1830152994552866e-06,
      "loss": 0.0072,
      "num_tokens": 208467211.0,
      "step": 466
    },
    {
      "epoch": 4.580565805658057,
      "grad_norm": 0.06003956537674327,
      "learning_rate": 1.175024225606912e-06,
      "loss": 0.0075,
      "num_tokens": 208909702.0,
      "step": 467
    },
    {
      "epoch": 4.590405904059041,
      "grad_norm": 0.056238323571208206,
      "learning_rate": 1.1672080671647695e-06,
      "loss": 0.0076,
      "num_tokens": 209342529.0,
      "step": 468
    },
    {
      "epoch": 4.6002460024600245,
      "grad_norm": 0.0552115239217723,
      "learning_rate": 1.1595671402384966e-06,
      "loss": 0.0112,
      "num_tokens": 209805083.0,
      "step": 469
    },
    {
      "epoch": 4.610086100861008,
      "grad_norm": 0.07199003274369836,
      "learning_rate": 1.152101753850828e-06,
      "loss": 0.0077,
      "num_tokens": 210267459.0,
      "step": 470
    },
    {
      "epoch": 4.619926199261993,
      "grad_norm": 0.050805856130265635,
      "learning_rate": 1.1448122099250946e-06,
      "loss": 0.0076,
      "num_tokens": 210732019.0,
      "step": 471
    },
    {
      "epoch": 4.629766297662977,
      "grad_norm": 0.054524548819372444,
      "learning_rate": 1.1376988032730135e-06,
      "loss": 0.0082,
      "num_tokens": 211158328.0,
      "step": 472
    },
    {
      "epoch": 4.639606396063961,
      "grad_norm": 0.061253234718759104,
      "learning_rate": 1.130761821582766e-06,
      "loss": 0.0068,
      "num_tokens": 211615451.0,
      "step": 473
    },
    {
      "epoch": 4.649446494464945,
      "grad_norm": 0.054438039734835346,
      "learning_rate": 1.1240015454073622e-06,
      "loss": 0.0073,
      "num_tokens": 212052321.0,
      "step": 474
    },
    {
      "epoch": 4.659286592865929,
      "grad_norm": 0.05428969184748025,
      "learning_rate": 1.1174182481532943e-06,
      "loss": 0.0085,
      "num_tokens": 212511116.0,
      "step": 475
    },
    {
      "epoch": 4.6691266912669125,
      "grad_norm": 0.06405292334946296,
      "learning_rate": 1.1110121960694773e-06,
      "loss": 0.015,
      "num_tokens": 212943028.0,
      "step": 476
    },
    {
      "epoch": 4.678966789667896,
      "grad_norm": 0.05682101787290208,
      "learning_rate": 1.104783648236486e-06,
      "loss": 0.0175,
      "num_tokens": 213395996.0,
      "step": 477
    },
    {
      "epoch": 4.68880688806888,
      "grad_norm": 0.05953289126520201,
      "learning_rate": 1.0987328565560711e-06,
      "loss": 0.007,
      "num_tokens": 213829686.0,
      "step": 478
    },
    {
      "epoch": 4.698646986469865,
      "grad_norm": 0.05168917084897981,
      "learning_rate": 1.0928600657409751e-06,
      "loss": 0.007,
      "num_tokens": 214277322.0,
      "step": 479
    },
    {
      "epoch": 4.708487084870849,
      "grad_norm": 0.1608212490455815,
      "learning_rate": 1.0871655133050372e-06,
      "loss": 0.1771,
      "num_tokens": 214763908.0,
      "step": 480
    },
    {
      "epoch": 4.718327183271833,
      "grad_norm": 0.06395186276533307,
      "learning_rate": 1.081649429553581e-06,
      "loss": 0.0096,
      "num_tokens": 215194438.0,
      "step": 481
    },
    {
      "epoch": 4.728167281672817,
      "grad_norm": 0.05084061327021202,
      "learning_rate": 1.076312037574106e-06,
      "loss": 0.0075,
      "num_tokens": 215654297.0,
      "step": 482
    },
    {
      "epoch": 4.7380073800738005,
      "grad_norm": 0.05497766832460128,
      "learning_rate": 1.0711535532272632e-06,
      "loss": 0.0075,
      "num_tokens": 216086613.0,
      "step": 483
    },
    {
      "epoch": 4.747847478474784,
      "grad_norm": 0.055956248844709046,
      "learning_rate": 1.0661741851381256e-06,
      "loss": 0.0076,
      "num_tokens": 216536848.0,
      "step": 484
    },
    {
      "epoch": 4.757687576875769,
      "grad_norm": 0.05660105878886696,
      "learning_rate": 1.0613741346877498e-06,
      "loss": 0.0074,
      "num_tokens": 216997743.0,
      "step": 485
    },
    {
      "epoch": 4.767527675276753,
      "grad_norm": 0.06132538554048896,
      "learning_rate": 1.056753596005032e-06,
      "loss": 0.0072,
      "num_tokens": 217458807.0,
      "step": 486
    },
    {
      "epoch": 4.777367773677737,
      "grad_norm": 0.055422872986755904,
      "learning_rate": 1.0523127559588579e-06,
      "loss": 0.0077,
      "num_tokens": 217917380.0,
      "step": 487
    },
    {
      "epoch": 4.787207872078721,
      "grad_norm": 0.05522528856299787,
      "learning_rate": 1.0480517941505428e-06,
      "loss": 0.0073,
      "num_tokens": 218379436.0,
      "step": 488
    },
    {
      "epoch": 4.797047970479705,
      "grad_norm": 0.06958848204521575,
      "learning_rate": 1.0439708829065708e-06,
      "loss": 0.0077,
      "num_tokens": 218813896.0,
      "step": 489
    },
    {
      "epoch": 4.8068880688806885,
      "grad_norm": 0.06386162989494891,
      "learning_rate": 1.0400701872716227e-06,
      "loss": 0.0073,
      "num_tokens": 219264589.0,
      "step": 490
    },
    {
      "epoch": 4.816728167281672,
      "grad_norm": 0.05787632062480919,
      "learning_rate": 1.0363498650019023e-06,
      "loss": 0.0076,
      "num_tokens": 219704673.0,
      "step": 491
    },
    {
      "epoch": 4.826568265682657,
      "grad_norm": 0.06292874524087964,
      "learning_rate": 1.0328100665587573e-06,
      "loss": 0.0127,
      "num_tokens": 220132557.0,
      "step": 492
    },
    {
      "epoch": 4.836408364083641,
      "grad_norm": 0.05664304451978608,
      "learning_rate": 1.029450935102592e-06,
      "loss": 0.0074,
      "num_tokens": 220586640.0,
      "step": 493
    },
    {
      "epoch": 4.846248462484625,
      "grad_norm": 0.06939065112948692,
      "learning_rate": 1.0262726064870801e-06,
      "loss": 0.0085,
      "num_tokens": 221016982.0,
      "step": 494
    },
    {
      "epoch": 4.856088560885609,
      "grad_norm": 0.052898898903196025,
      "learning_rate": 1.0232752092536666e-06,
      "loss": 0.0074,
      "num_tokens": 221462667.0,
      "step": 495
    },
    {
      "epoch": 4.865928659286593,
      "grad_norm": 0.05776273104035291,
      "learning_rate": 1.0204588646263731e-06,
      "loss": 0.0071,
      "num_tokens": 221913595.0,
      "step": 496
    },
    {
      "epoch": 4.875768757687577,
      "grad_norm": 0.05461381324686126,
      "learning_rate": 1.0178236865068933e-06,
      "loss": 0.0074,
      "num_tokens": 222326904.0,
      "step": 497
    },
    {
      "epoch": 4.885608856088561,
      "grad_norm": 0.055498699411421645,
      "learning_rate": 1.0153697814699858e-06,
      "loss": 0.0072,
      "num_tokens": 222774557.0,
      "step": 498
    },
    {
      "epoch": 4.895448954489545,
      "grad_norm": 0.053031884611377415,
      "learning_rate": 1.0130972487591658e-06,
      "loss": 0.0068,
      "num_tokens": 223236906.0,
      "step": 499
    },
    {
      "epoch": 4.905289052890529,
      "grad_norm": 0.054483452664515604,
      "learning_rate": 1.0110061802826889e-06,
      "loss": 0.0074,
      "num_tokens": 223673165.0,
      "step": 500
    },
    {
      "epoch": 4.915129151291513,
      "grad_norm": 0.05035207860411025,
      "learning_rate": 1.009096660609837e-06,
      "loss": 0.0068,
      "num_tokens": 224139613.0,
      "step": 501
    },
    {
      "epoch": 4.924969249692497,
      "grad_norm": 0.05491435698714458,
      "learning_rate": 1.0073687669674949e-06,
      "loss": 0.0108,
      "num_tokens": 224623785.0,
      "step": 502
    },
    {
      "epoch": 4.934809348093481,
      "grad_norm": 0.05816593959272977,
      "learning_rate": 1.0058225692370299e-06,
      "loss": 0.0074,
      "num_tokens": 225060586.0,
      "step": 503
    },
    {
      "epoch": 4.944649446494465,
      "grad_norm": 0.05535566978108373,
      "learning_rate": 1.0044581299514638e-06,
      "loss": 0.0073,
      "num_tokens": 225504012.0,
      "step": 504
    },
    {
      "epoch": 4.9544895448954485,
      "grad_norm": 0.053032498372300854,
      "learning_rate": 1.003275504292944e-06,
      "loss": 0.0073,
      "num_tokens": 225944573.0,
      "step": 505
    },
    {
      "epoch": 4.964329643296433,
      "grad_norm": 0.051644596347508585,
      "learning_rate": 1.0022747400905126e-06,
      "loss": 0.0072,
      "num_tokens": 226383598.0,
      "step": 506
    },
    {
      "epoch": 4.974169741697417,
      "grad_norm": 0.05140289748972471,
      "learning_rate": 1.0014558778181714e-06,
      "loss": 0.0067,
      "num_tokens": 226851896.0,
      "step": 507
    },
    {
      "epoch": 4.984009840098401,
      "grad_norm": 0.05762796975847249,
      "learning_rate": 1.0008189505932444e-06,
      "loss": 0.0077,
      "num_tokens": 227269800.0,
      "step": 508
    },
    {
      "epoch": 4.993849938499385,
      "grad_norm": 0.05745101659790941,
      "learning_rate": 1.0003639841750404e-06,
      "loss": 0.007,
      "num_tokens": 227720300.0,
      "step": 509
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0720275822622563,
      "learning_rate": 1.0000909969638097e-06,
      "loss": 0.007,
      "num_tokens": 227931822.0,
      "step": 510
    },
    {
      "epoch": 5.0,
      "step": 510,
      "total_flos": 7.699310209956577e+17,
      "train_loss": 0.04971654405360859,
      "train_runtime": 7469.8311,
      "train_samples_per_second": 8.7,
      "train_steps_per_second": 0.068
    }
  ],
  "logging_steps": 1,
  "max_steps": 510,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.699310209956577e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}