{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.694126339120798,
  "eval_steps": 1000,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03694126339120798,
      "grad_norm": 1.4473156929016113,
      "learning_rate": 6.6e-05,
      "loss": 2.0357,
      "mean_token_accuracy": 0.4905380755662918,
      "num_tokens": 246180.0,
      "step": 100
    },
    {
      "epoch": 0.07388252678241596,
      "grad_norm": 1.4902719259262085,
      "learning_rate": 0.00013266666666666667,
      "loss": 1.3937,
      "mean_token_accuracy": 0.5916463854908943,
      "num_tokens": 492915.0,
      "step": 200
    },
    {
      "epoch": 0.11082379017362394,
      "grad_norm": 1.3383102416992188,
      "learning_rate": 0.00019933333333333334,
      "loss": 1.3205,
      "mean_token_accuracy": 0.6068769115209579,
      "num_tokens": 740398.0,
      "step": 300
    },
    {
      "epoch": 0.1477650535648319,
      "grad_norm": 1.2365469932556152,
      "learning_rate": 0.000173421993904824,
      "loss": 1.3026,
      "mean_token_accuracy": 0.6126913416385651,
      "num_tokens": 988197.0,
      "step": 400
    },
    {
      "epoch": 0.1847063169560399,
      "grad_norm": 1.0308854579925537,
      "learning_rate": 0.0001550744859491231,
      "loss": 1.2537,
      "mean_token_accuracy": 0.6190469121932983,
      "num_tokens": 1237334.0,
      "step": 500
    },
    {
      "epoch": 0.22164758034724788,
      "grad_norm": 1.2729270458221436,
      "learning_rate": 0.00014153935488632152,
      "loss": 1.2353,
      "mean_token_accuracy": 0.6254772353172302,
      "num_tokens": 1483308.0,
      "step": 600
    },
    {
      "epoch": 0.25858884373845586,
      "grad_norm": 0.9841827750205994,
      "learning_rate": 0.00013102435641608367,
      "loss": 1.2182,
      "mean_token_accuracy": 0.6275931853055954,
      "num_tokens": 1730192.0,
      "step": 700
    },
    {
      "epoch": 0.2955301071296638,
      "grad_norm": 0.9242544174194336,
      "learning_rate": 0.00012255110553085002,
      "loss": 1.2049,
      "mean_token_accuracy": 0.6283232820034027,
      "num_tokens": 1980054.0,
      "step": 800
    },
    {
      "epoch": 0.33247137052087183,
      "grad_norm": 0.8931549787521362,
      "learning_rate": 0.00011553425737574005,
      "loss": 1.2017,
      "mean_token_accuracy": 0.6302745240926743,
      "num_tokens": 2228605.0,
      "step": 900
    },
    {
      "epoch": 0.3694126339120798,
      "grad_norm": 1.044004201889038,
      "learning_rate": 0.0001095993248702382,
      "loss": 1.2137,
      "step": 1000
    },
    {
      "epoch": 0.3694126339120798,
      "eval_loss": 1.1836973428726196,
      "eval_mean_token_accuracy": 0.6324496693611145,
      "eval_num_tokens": 2475392.0,
      "eval_runtime": 5.3895,
      "eval_samples_per_second": 185.547,
      "eval_steps_per_second": 23.193,
      "step": 1000
    },
    {
      "epoch": 0.40635389730328775,
      "grad_norm": 0.8837220072746277,
      "learning_rate": 0.00010449410169212441,
      "loss": 1.1854,
      "mean_token_accuracy": 0.6299630090594291,
      "num_tokens": 2724102.0,
      "step": 1100
    },
    {
      "epoch": 0.44329516069449576,
      "grad_norm": 0.8813680410385132,
      "learning_rate": 0.00010004169272643103,
      "loss": 1.1776,
      "mean_token_accuracy": 0.6349628627300262,
      "num_tokens": 2972072.0,
      "step": 1200
    },
    {
      "epoch": 0.4802364240857037,
      "grad_norm": 0.9930716753005981,
      "learning_rate": 9.611386626644256e-05,
      "loss": 1.1683,
      "mean_token_accuracy": 0.6352091038227081,
      "num_tokens": 3217529.0,
      "step": 1300
    },
    {
      "epoch": 0.5171776874769117,
      "grad_norm": 0.8394394516944885,
      "learning_rate": 9.261509270476351e-05,
      "loss": 1.1687,
      "mean_token_accuracy": 0.6363370817899704,
      "num_tokens": 3467819.0,
      "step": 1400
    },
    {
      "epoch": 0.5541189508681197,
      "grad_norm": 1.129971981048584,
      "learning_rate": 8.9472548255098e-05,
      "loss": 1.1541,
      "mean_token_accuracy": 0.6417357540130615,
      "num_tokens": 3712960.0,
      "step": 1500
    },
    {
      "epoch": 0.5910602142593276,
      "grad_norm": 1.100881576538086,
      "learning_rate": 8.662961636484199e-05,
      "loss": 1.1602,
      "mean_token_accuracy": 0.6412180256843567,
      "num_tokens": 3956425.0,
      "step": 1600
    },
    {
      "epoch": 0.6280014776505356,
      "grad_norm": 0.9382134079933167,
      "learning_rate": 8.40415267738742e-05,
      "loss": 1.1604,
      "mean_token_accuracy": 0.6395090478658676,
      "num_tokens": 4203009.0,
      "step": 1700
    },
    {
      "epoch": 0.6649427410417437,
      "grad_norm": 1.084293246269226,
      "learning_rate": 8.167234800792304e-05,
      "loss": 1.1352,
      "mean_token_accuracy": 0.646253719329834,
      "num_tokens": 4449448.0,
      "step": 1800
    },
    {
      "epoch": 0.7018840044329516,
      "grad_norm": 0.9637444019317627,
      "learning_rate": 7.949286335171643e-05,
      "loss": 1.1562,
      "mean_token_accuracy": 0.6406971418857574,
      "num_tokens": 4694282.0,
      "step": 1900
    },
    {
      "epoch": 0.7388252678241596,
      "grad_norm": 0.9237338304519653,
      "learning_rate": 7.747903910575024e-05,
      "loss": 1.1424,
      "step": 2000
    },
    {
      "epoch": 0.7388252678241596,
      "eval_loss": 1.1218078136444092,
      "eval_mean_token_accuracy": 0.6506093912124634,
      "eval_num_tokens": 4940833.0,
      "eval_runtime": 5.3952,
      "eval_samples_per_second": 185.351,
      "eval_steps_per_second": 23.169,
      "step": 2000
    },
    {
      "epoch": 0.7757665312153675,
      "grad_norm": 0.777606725692749,
      "learning_rate": 7.561089934060305e-05,
      "loss": 1.1313,
      "mean_token_accuracy": 0.6467883923649788,
      "num_tokens": 5190358.0,
      "step": 2100
    },
    {
      "epoch": 0.8127077946065755,
      "grad_norm": 0.8855065703392029,
      "learning_rate": 7.387168551531178e-05,
      "loss": 1.1309,
      "mean_token_accuracy": 0.6449691706895828,
      "num_tokens": 5438772.0,
      "step": 2200
    },
    {
      "epoch": 0.8496490579977836,
      "grad_norm": 1.2018849849700928,
      "learning_rate": 7.224721947627513e-05,
      "loss": 1.1246,
      "mean_token_accuracy": 0.6473777782917023,
      "num_tokens": 5682900.0,
      "step": 2300
    },
    {
      "epoch": 0.8865903213889915,
      "grad_norm": 0.9426067471504211,
      "learning_rate": 7.07254141150883e-05,
      "loss": 1.1241,
      "mean_token_accuracy": 0.647763032913208,
      "num_tokens": 5931817.0,
      "step": 2400
    },
    {
      "epoch": 0.9235315847801995,
      "grad_norm": 0.7986142039299011,
      "learning_rate": 6.929589286752371e-05,
      "loss": 1.1258,
      "mean_token_accuracy": 0.6497471231222153,
      "num_tokens": 6179818.0,
      "step": 2500
    },
    {
      "epoch": 0.9604728481714074,
      "grad_norm": 1.1682002544403076,
      "learning_rate": 6.794969055356698e-05,
      "loss": 1.1286,
      "mean_token_accuracy": 0.6473524701595307,
      "num_tokens": 6426250.0,
      "step": 2600
    },
    {
      "epoch": 0.9974141115626154,
      "grad_norm": 0.9333537220954895,
      "learning_rate": 6.667901577609308e-05,
      "loss": 1.1239,
      "mean_token_accuracy": 0.6484623271226883,
      "num_tokens": 6672929.0,
      "step": 2700
    },
    {
      "epoch": 1.0343553749538235,
      "grad_norm": 0.5627142190933228,
      "learning_rate": 6.547706044716512e-05,
      "loss": 1.0508,
      "mean_token_accuracy": 0.6667241591215134,
      "num_tokens": 6925125.0,
      "step": 2800
    },
    {
      "epoch": 1.0712966383450313,
      "grad_norm": 0.5715717077255249,
      "learning_rate": 6.433784577873342e-05,
      "loss": 1.0615,
      "mean_token_accuracy": 0.6637595742940903,
      "num_tokens": 7171997.0,
      "step": 2900
    },
    {
      "epoch": 1.1082379017362394,
      "grad_norm": 0.5389025211334229,
      "learning_rate": 6.325609676486509e-05,
      "loss": 1.0709,
      "step": 3000
    },
    {
      "epoch": 1.1082379017362394,
      "eval_loss": 1.0916837453842163,
      "eval_mean_token_accuracy": 0.6582915830612183,
      "eval_num_tokens": 7419279.0,
      "eval_runtime": 5.3894,
      "eval_samples_per_second": 185.548,
      "eval_steps_per_second": 23.193,
      "step": 3000
    },
    {
      "epoch": 1.1451791651274474,
      "grad_norm": 0.56490159034729,
      "learning_rate": 6.22271391287055e-05,
      "loss": 1.0581,
      "mean_token_accuracy": 0.662086527645588,
      "num_tokens": 7664383.0,
      "step": 3100
    },
    {
      "epoch": 1.1821204285186553,
      "grad_norm": 0.543954074382782,
      "learning_rate": 6.12468141320462e-05,
      "loss": 1.0615,
      "mean_token_accuracy": 0.6631740409135819,
      "num_tokens": 7912840.0,
      "step": 3200
    },
    {
      "epoch": 1.2190616919098634,
      "grad_norm": 0.5596346855163574,
      "learning_rate": 6.03114077000131e-05,
      "loss": 1.0395,
      "mean_token_accuracy": 0.671640704870224,
      "num_tokens": 8160289.0,
      "step": 3300
    },
    {
      "epoch": 1.2560029553010712,
      "grad_norm": 0.5955942869186401,
      "learning_rate": 5.9417591102230663e-05,
      "loss": 1.0567,
      "mean_token_accuracy": 0.6638083755970001,
      "num_tokens": 8407857.0,
      "step": 3400
    },
    {
      "epoch": 1.2929442186922793,
      "grad_norm": 0.5933428406715393,
      "learning_rate": 5.856237102757652e-05,
      "loss": 1.0636,
      "mean_token_accuracy": 0.6627275788784027,
      "num_tokens": 8655624.0,
      "step": 3500
    },
    {
      "epoch": 1.3298854820834873,
      "grad_norm": 0.5572307109832764,
      "learning_rate": 5.7743047343661814e-05,
      "loss": 1.0652,
      "mean_token_accuracy": 0.6624369341135025,
      "num_tokens": 8902821.0,
      "step": 3600
    },
    {
      "epoch": 1.3668267454746952,
      "grad_norm": 0.596443235874176,
      "learning_rate": 5.6957177181117404e-05,
      "loss": 1.0622,
      "mean_token_accuracy": 0.6634978985786438,
      "num_tokens": 9149878.0,
      "step": 3700
    },
    {
      "epoch": 1.4037680088659032,
      "grad_norm": 0.5873645544052124,
      "learning_rate": 5.620254425309578e-05,
      "loss": 1.0493,
      "mean_token_accuracy": 0.6654049742221833,
      "num_tokens": 9394858.0,
      "step": 3800
    },
    {
      "epoch": 1.440709272257111,
      "grad_norm": 0.59478759765625,
      "learning_rate": 5.547713253139649e-05,
      "loss": 1.0456,
      "mean_token_accuracy": 0.6683453869819641,
      "num_tokens": 9641008.0,
      "step": 3900
    },
    {
      "epoch": 1.4776505356483192,
      "grad_norm": 0.5986542701721191,
      "learning_rate": 5.477910356647767e-05,
      "loss": 1.043,
      "step": 4000
    },
    {
      "epoch": 1.4776505356483192,
      "eval_loss": 1.083065152168274,
      "eval_mean_token_accuracy": 0.6582373585700989,
      "eval_num_tokens": 9888284.0,
      "eval_runtime": 5.3783,
      "eval_samples_per_second": 185.933,
      "eval_steps_per_second": 23.242,
      "step": 4000
    },
    {
      "epoch": 1.5145917990395272,
      "grad_norm": 0.5704376697540283,
      "learning_rate": 5.410677686985887e-05,
      "loss": 1.0321,
      "mean_token_accuracy": 0.6689175629615783,
      "num_tokens": 10134094.0,
      "step": 4100
    },
    {
      "epoch": 1.551533062430735,
      "grad_norm": 0.5602062344551086,
      "learning_rate": 5.345861288192786e-05,
      "loss": 1.0441,
      "mean_token_accuracy": 0.6680737626552582,
      "num_tokens": 10381316.0,
      "step": 4200
    },
    {
      "epoch": 1.5884743258219431,
      "grad_norm": 0.5552584528923035,
      "learning_rate": 5.283319813188472e-05,
      "loss": 1.0369,
      "mean_token_accuracy": 0.6698204201459884,
      "num_tokens": 10628471.0,
      "step": 4300
    },
    {
      "epoch": 1.625415589213151,
      "grad_norm": 0.6024323105812073,
      "learning_rate": 5.222923226400155e-05,
      "loss": 1.0403,
      "mean_token_accuracy": 0.6691047704219818,
      "num_tokens": 10874430.0,
      "step": 4400
    },
    {
      "epoch": 1.662356852604359,
      "grad_norm": 0.5967562794685364,
      "learning_rate": 5.164551665900703e-05,
      "loss": 1.0483,
      "mean_token_accuracy": 0.6648873990774155,
      "num_tokens": 11123128.0,
      "step": 4500
    },
    {
      "epoch": 1.6992981159955671,
      "grad_norm": 0.5668358206748962,
      "learning_rate": 5.1080944423879696e-05,
      "loss": 1.0392,
      "mean_token_accuracy": 0.6674597597122193,
      "num_tokens": 11368003.0,
      "step": 4600
    },
    {
      "epoch": 1.736239379386775,
      "grad_norm": 0.6373595595359802,
      "learning_rate": 5.053449155971992e-05,
      "loss": 1.0404,
      "mean_token_accuracy": 0.6689798641204834,
      "num_tokens": 11615194.0,
      "step": 4700
    },
    {
      "epoch": 1.773180642777983,
      "grad_norm": 0.6193637847900391,
      "learning_rate": 5.0005209147276734e-05,
      "loss": 1.0355,
      "mean_token_accuracy": 0.6675721609592438,
      "num_tokens": 11863548.0,
      "step": 4800
    },
    {
      "epoch": 1.8101219061691909,
      "grad_norm": 0.5770505666732788,
      "learning_rate": 4.949221641439499e-05,
      "loss": 1.0316,
      "mean_token_accuracy": 0.6724146312475204,
      "num_tokens": 12111554.0,
      "step": 4900
    },
    {
      "epoch": 1.847063169560399,
      "grad_norm": 0.6453244686126709,
      "learning_rate": 4.899469457011854e-05,
      "loss": 1.0348,
      "step": 5000
    },
    {
      "epoch": 1.847063169560399,
      "eval_loss": 1.0698254108428955,
      "eval_mean_token_accuracy": 0.66366588306427,
      "eval_num_tokens": 12359520.0,
      "eval_runtime": 5.3865,
      "eval_samples_per_second": 185.649,
      "eval_steps_per_second": 23.206,
      "step": 5000
    },
    {
      "epoch": 1.884004432951607,
      "grad_norm": 0.6129611730575562,
      "learning_rate": 4.851188130722481e-05,
      "loss": 1.0374,
      "mean_token_accuracy": 0.6697911691665649,
      "num_tokens": 12607238.0,
      "step": 5100
    },
    {
      "epoch": 1.920945696342815,
      "grad_norm": 0.6002670526504517,
      "learning_rate": 4.804306588920635e-05,
      "loss": 1.035,
      "mean_token_accuracy": 0.6703519684076309,
      "num_tokens": 12855037.0,
      "step": 5200
    },
    {
      "epoch": 1.957886959734023,
      "grad_norm": 0.5875000953674316,
      "learning_rate": 4.758758474966023e-05,
      "loss": 1.0348,
      "mean_token_accuracy": 0.6686445927619934,
      "num_tokens": 13100596.0,
      "step": 5300
    },
    {
      "epoch": 1.9948282231252308,
      "grad_norm": 0.587979257106781,
      "learning_rate": 4.7144817542100825e-05,
      "loss": 1.0375,
      "mean_token_accuracy": 0.6698001223802567,
      "num_tokens": 13349667.0,
      "step": 5400
    },
    {
      "epoch": 2.031769486516439,
      "grad_norm": 0.5400444269180298,
      "learning_rate": 4.671418358670517e-05,
      "loss": 1.0064,
      "mean_token_accuracy": 0.6771922719478607,
      "num_tokens": 13599712.0,
      "step": 5500
    },
    {
      "epoch": 2.068710749907647,
      "grad_norm": 0.5831236839294434,
      "learning_rate": 4.6295138667698956e-05,
      "loss": 0.9874,
      "mean_token_accuracy": 0.6801465088129044,
      "num_tokens": 13845238.0,
      "step": 5600
    },
    {
      "epoch": 2.105652013298855,
      "grad_norm": 0.559648334980011,
      "learning_rate": 4.5887172141209994e-05,
      "loss": 0.9824,
      "mean_token_accuracy": 0.6835214233398438,
      "num_tokens": 14094078.0,
      "step": 5700
    },
    {
      "epoch": 2.1425932766900626,
      "grad_norm": 0.5524799227714539,
      "learning_rate": 4.548980431863551e-05,
      "loss": 0.9852,
      "mean_token_accuracy": 0.6822834074497223,
      "num_tokens": 14342112.0,
      "step": 5800
    },
    {
      "epoch": 2.1795345400812707,
      "grad_norm": 0.586271345615387,
      "learning_rate": 4.510258409503273e-05,
      "loss": 0.9807,
      "mean_token_accuracy": 0.6823082774877548,
      "num_tokens": 14587875.0,
      "step": 5900
    },
    {
      "epoch": 2.2164758034724787,
      "grad_norm": 0.599862277507782,
      "learning_rate": 4.472508679587051e-05,
      "loss": 0.9741,
      "step": 6000
    },
    {
      "epoch": 2.2164758034724787,
      "eval_loss": 1.071539282798767,
      "eval_mean_token_accuracy": 0.6636041073799134,
      "eval_num_tokens": 14833253.0,
      "eval_runtime": 5.3837,
      "eval_samples_per_second": 185.745,
      "eval_steps_per_second": 23.218,
      "step": 6000
    },
    {
      "epoch": 2.253417066863687,
      "grad_norm": 0.5849953293800354,
      "learning_rate": 4.435691221877225e-05,
      "loss": 0.9784,
      "mean_token_accuracy": 0.6845134419202804,
      "num_tokens": 15079437.0,
      "step": 6100
    },
    {
      "epoch": 2.290358330254895,
      "grad_norm": 0.5756722688674927,
      "learning_rate": 4.399768284971994e-05,
      "loss": 0.9843,
      "mean_token_accuracy": 0.6839412766695022,
      "num_tokens": 15326743.0,
      "step": 6200
    },
    {
      "epoch": 2.3272995936461025,
      "grad_norm": 0.5707868933677673,
      "learning_rate": 4.364704223564281e-05,
      "loss": 0.9901,
      "mean_token_accuracy": 0.6811071854829788,
      "num_tokens": 15572363.0,
      "step": 6300
    },
    {
      "epoch": 2.3642408570373106,
      "grad_norm": 0.6192522048950195,
      "learning_rate": 4.330465349744206e-05,
      "loss": 0.9762,
      "mean_token_accuracy": 0.6834132850170136,
      "num_tokens": 15818229.0,
      "step": 6400
    },
    {
      "epoch": 2.4011821204285186,
      "grad_norm": 0.5887159109115601,
      "learning_rate": 4.2970197969350315e-05,
      "loss": 0.9797,
      "mean_token_accuracy": 0.6834959721565247,
      "num_tokens": 16066092.0,
      "step": 6500
    },
    {
      "epoch": 2.4381233838197267,
      "grad_norm": 0.6107765436172485,
      "learning_rate": 4.264337395213374e-05,
      "loss": 0.97,
      "mean_token_accuracy": 0.6855223393440246,
      "num_tokens": 16314390.0,
      "step": 6600
    },
    {
      "epoch": 2.4750646472109348,
      "grad_norm": 0.5866128206253052,
      "learning_rate": 4.232389556904849e-05,
      "loss": 0.9794,
      "mean_token_accuracy": 0.6824937015771866,
      "num_tokens": 16560430.0,
      "step": 6700
    },
    {
      "epoch": 2.5120059106021424,
      "grad_norm": 0.5723136067390442,
      "learning_rate": 4.201149171469091e-05,
      "loss": 0.9805,
      "mean_token_accuracy": 0.6831120592355728,
      "num_tokens": 16807172.0,
      "step": 6800
    },
    {
      "epoch": 2.5489471739933505,
      "grad_norm": 0.5534746050834656,
      "learning_rate": 4.170590508795705e-05,
      "loss": 0.9853,
      "mean_token_accuracy": 0.679190359711647,
      "num_tokens": 17054725.0,
      "step": 6900
    },
    {
      "epoch": 2.5858884373845585,
      "grad_norm": 0.6190086007118225,
      "learning_rate": 4.1406891301271574e-05,
      "loss": 0.979,
      "step": 7000
    },
    {
      "epoch": 2.5858884373845585,
      "eval_loss": 1.0672271251678467,
      "eval_mean_token_accuracy": 0.665556743144989,
      "eval_num_tokens": 17301619.0,
      "eval_runtime": 5.3864,
      "eval_samples_per_second": 185.654,
      "eval_steps_per_second": 23.207,
      "step": 7000
    },
    {
      "epoch": 2.6228297007757666,
      "grad_norm": 0.604921281337738,
      "learning_rate": 4.111421805907759e-05,
      "loss": 0.9844,
      "mean_token_accuracy": 0.6844497114419937,
      "num_tokens": 17546632.0,
      "step": 7100
    },
    {
      "epoch": 2.6597709641669747,
      "grad_norm": 0.6087679862976074,
      "learning_rate": 4.082766439931165e-05,
      "loss": 0.9871,
      "mean_token_accuracy": 0.6810534721612931,
      "num_tokens": 17797482.0,
      "step": 7200
    },
    {
      "epoch": 2.6967122275581827,
      "grad_norm": 0.6020961999893188,
      "learning_rate": 4.054701999223518e-05,
      "loss": 0.9839,
      "mean_token_accuracy": 0.6829093122482299,
      "num_tokens": 18043599.0,
      "step": 7300
    },
    {
      "epoch": 2.7336534909493904,
      "grad_norm": 0.6339052319526672,
      "learning_rate": 4.0272084491566247e-05,
      "loss": 0.9863,
      "mean_token_accuracy": 0.6820144325494766,
      "num_tokens": 18287793.0,
      "step": 7400
    },
    {
      "epoch": 2.7705947543405984,
      "grad_norm": 0.6206592321395874,
      "learning_rate": 4.000266693336297e-05,
      "loss": 0.9709,
      "mean_token_accuracy": 0.6874477595090867,
      "num_tokens": 18535420.0,
      "step": 7500
    },
    {
      "epoch": 2.8075360177318065,
      "grad_norm": 0.6235191822052002,
      "learning_rate": 3.973858517856019e-05,
      "loss": 0.9734,
      "mean_token_accuracy": 0.6847814846038819,
      "num_tokens": 18784286.0,
      "step": 7600
    },
    {
      "epoch": 2.8444772811230146,
      "grad_norm": 0.6308836340904236,
      "learning_rate": 3.947966539546186e-05,
      "loss": 0.9813,
      "mean_token_accuracy": 0.6831617254018784,
      "num_tokens": 19035813.0,
      "step": 7700
    },
    {
      "epoch": 2.881418544514222,
      "grad_norm": 0.6001960039138794,
      "learning_rate": 3.922574157884801e-05,
      "loss": 0.987,
      "mean_token_accuracy": 0.6807804244756699,
      "num_tokens": 19282122.0,
      "step": 7800
    },
    {
      "epoch": 2.9183598079054303,
      "grad_norm": 0.6059972643852234,
      "learning_rate": 3.8976655102673755e-05,
      "loss": 0.9859,
      "mean_token_accuracy": 0.6820109623670578,
      "num_tokens": 19529782.0,
      "step": 7900
    },
    {
      "epoch": 2.9553010712966383,
      "grad_norm": 0.640872061252594,
      "learning_rate": 3.873225430362181e-05,
      "loss": 0.9761,
      "step": 8000
    },
    {
      "epoch": 2.9553010712966383,
      "eval_loss": 1.0601364374160767,
      "eval_mean_token_accuracy": 0.6661491613388062,
      "eval_num_tokens": 19777841.0,
      "eval_runtime": 5.4066,
      "eval_samples_per_second": 184.959,
      "eval_steps_per_second": 23.12,
      "step": 8000
    },
    {
      "epoch": 2.9922423346878464,
      "grad_norm": 0.6132466793060303,
      "learning_rate": 3.8492394093024636e-05,
      "loss": 0.9711,
      "mean_token_accuracy": 0.6846305218338966,
      "num_tokens": 20023858.0,
      "step": 8100
    },
    {
      "epoch": 3.0291835980790545,
      "grad_norm": 0.5972515940666199,
      "learning_rate": 3.825693559490006e-05,
      "loss": 0.9376,
      "mean_token_accuracy": 0.692512179017067,
      "num_tokens": 20270581.0,
      "step": 8200
    },
    {
      "epoch": 3.066124861470262,
      "grad_norm": 0.6202205419540405,
      "learning_rate": 3.8025745808048846e-05,
      "loss": 0.9307,
      "mean_token_accuracy": 0.6948988193273544,
      "num_tokens": 20516819.0,
      "step": 8300
    },
    {
      "epoch": 3.10306612486147,
      "grad_norm": 0.585482120513916,
      "learning_rate": 3.779869729034645e-05,
      "loss": 0.935,
      "mean_token_accuracy": 0.6954682809114456,
      "num_tokens": 20763559.0,
      "step": 8400
    },
    {
      "epoch": 3.1400073882526782,
      "grad_norm": 0.6358840465545654,
      "learning_rate": 3.7575667863526335e-05,
      "loss": 0.9292,
      "mean_token_accuracy": 0.6980463570356369,
      "num_tokens": 21011357.0,
      "step": 8500
    },
    {
      "epoch": 3.1769486516438863,
      "grad_norm": 0.6181186437606812,
      "learning_rate": 3.735654033690154e-05,
      "loss": 0.9229,
      "mean_token_accuracy": 0.696455385684967,
      "num_tokens": 21259566.0,
      "step": 8600
    },
    {
      "epoch": 3.2138899150350944,
      "grad_norm": 0.6604560017585754,
      "learning_rate": 3.7141202248604964e-05,
      "loss": 0.9285,
      "mean_token_accuracy": 0.6962138444185257,
      "num_tokens": 21506213.0,
      "step": 8700
    },
    {
      "epoch": 3.250831178426302,
      "grad_norm": 0.5987362265586853,
      "learning_rate": 3.6929545623050815e-05,
      "loss": 0.929,
      "mean_token_accuracy": 0.6957518076896667,
      "num_tokens": 21754908.0,
      "step": 8800
    },
    {
      "epoch": 3.28777244181751,
      "grad_norm": 0.5980191230773926,
      "learning_rate": 3.6721466743428706e-05,
      "loss": 0.938,
      "mean_token_accuracy": 0.6953296983242034,
      "num_tokens": 22001522.0,
      "step": 8900
    },
    {
      "epoch": 3.324713705208718,
      "grad_norm": 0.6171393990516663,
      "learning_rate": 3.6516865938141736e-05,
      "loss": 0.9364,
      "step": 9000
    },
    {
      "epoch": 3.324713705208718,
      "eval_loss": 1.0706262588500977,
      "eval_mean_token_accuracy": 0.6642319107055664,
      "eval_num_tokens": 22248126.0,
      "eval_runtime": 5.3921,
      "eval_samples_per_second": 185.455,
      "eval_steps_per_second": 23.182,
      "step": 9000
    },
    {
      "epoch": 3.361654968599926,
      "grad_norm": 0.6272869110107422,
      "learning_rate": 3.6315647380189556e-05,
      "loss": 0.919,
      "mean_token_accuracy": 0.6962304222583771,
      "num_tokens": 22494098.0,
      "step": 9100
    },
    {
      "epoch": 3.3985962319911343,
      "grad_norm": 0.6461876034736633,
      "learning_rate": 3.611771889857922e-05,
      "loss": 0.9331,
      "mean_token_accuracy": 0.6959864324331284,
      "num_tokens": 22741819.0,
      "step": 9200
    },
    {
      "epoch": 3.4355374953823423,
      "grad_norm": 0.5855452418327332,
      "learning_rate": 3.592299180092082e-05,
      "loss": 0.9283,
      "mean_token_accuracy": 0.6968681657314301,
      "num_tokens": 22988621.0,
      "step": 9300
    },
    {
      "epoch": 3.47247875877355,
      "grad_norm": 0.6252483129501343,
      "learning_rate": 3.573138070643225e-05,
      "loss": 0.9271,
      "mean_token_accuracy": 0.6972170048952102,
      "num_tokens": 23235818.0,
      "step": 9400
    },
    {
      "epoch": 3.509420022164758,
      "grad_norm": 0.6152076125144958,
      "learning_rate": 3.554280338863896e-05,
      "loss": 0.9244,
      "mean_token_accuracy": 0.699801824092865,
      "num_tokens": 23483243.0,
      "step": 9500
    },
    {
      "epoch": 3.546361285555966,
      "grad_norm": 0.6049486398696899,
      "learning_rate": 3.535718062711045e-05,
      "loss": 0.9365,
      "mean_token_accuracy": 0.6951554995775223,
      "num_tokens": 23730501.0,
      "step": 9600
    },
    {
      "epoch": 3.583302548947174,
      "grad_norm": 0.6222932934761047,
      "learning_rate": 3.517443606762636e-05,
      "loss": 0.9374,
      "mean_token_accuracy": 0.6915060871839523,
      "num_tokens": 23978205.0,
      "step": 9700
    },
    {
      "epoch": 3.6202438123383818,
      "grad_norm": 0.6344577670097351,
      "learning_rate": 3.499449609021135e-05,
      "loss": 0.9252,
      "mean_token_accuracy": 0.6969369679689408,
      "num_tokens": 24224944.0,
      "step": 9800
    },
    {
      "epoch": 3.65718507572959,
      "grad_norm": 0.600951611995697,
      "learning_rate": 3.4817289684521056e-05,
      "loss": 0.9226,
      "mean_token_accuracy": 0.6991156005859375,
      "num_tokens": 24470936.0,
      "step": 9900
    },
    {
      "epoch": 3.694126339120798,
      "grad_norm": 0.6917585134506226,
      "learning_rate": 3.4642748332099756e-05,
      "loss": 0.935,
      "step": 10000
    },
    {
      "epoch": 3.694126339120798,
      "eval_loss": 1.0667781829833984,
      "eval_mean_token_accuracy": 0.6657101097106933,
      "eval_num_tokens": 24717605.0,
      "eval_runtime": 5.4034,
      "eval_samples_per_second": 185.07,
      "eval_steps_per_second": 23.134,
      "step": 10000
    }
  ],
  "logging_steps": 100,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.9870782814052352e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}