{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 12.0,
  "eval_steps": 500,
  "global_step": 504,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "aux_loss": 8.045117378234863,
      "entropy": 0.05226727016270161,
      "epoch": 0.024096385542168676,
      "grad_norm": 0.5021907687187195,
      "learning_rate": 0.0,
      "loss": 0.09741637855768204,
      "mean_token_accuracy": 0.9888820052146912,
      "num_tokens": 154362.0,
      "step": 1
    },
    {
      "aux_loss": 8.047614574432373,
      "entropy": 0.05835266411304474,
      "epoch": 0.04819277108433735,
      "grad_norm": 0.551396369934082,
      "learning_rate": 1.4500000000000001e-07,
      "loss": 0.09991507232189178,
      "mean_token_accuracy": 0.9883801937103271,
      "num_tokens": 310702.0,
      "step": 2
    },
    {
      "aux_loss": 8.049727439880371,
      "entropy": 0.052583809942007065,
      "epoch": 0.07228915662650602,
      "grad_norm": 0.4750031530857086,
      "learning_rate": 2.9000000000000003e-07,
      "loss": 0.09806497395038605,
      "mean_token_accuracy": 0.9886206984519958,
      "num_tokens": 493904.0,
      "step": 3
    },
    {
      "aux_loss": 8.047826766967773,
      "entropy": 0.05231792479753494,
      "epoch": 0.0963855421686747,
      "grad_norm": 0.4863564074039459,
      "learning_rate": 4.35e-07,
      "loss": 0.09776315838098526,
      "mean_token_accuracy": 0.9888334274291992,
      "num_tokens": 663136.0,
      "step": 4
    },
    {
      "aux_loss": 8.047198295593262,
      "entropy": 0.04943128302693367,
      "epoch": 0.12048192771084337,
      "grad_norm": 0.5010114908218384,
      "learning_rate": 5.800000000000001e-07,
      "loss": 0.09543517231941223,
      "mean_token_accuracy": 0.9894405901432037,
      "num_tokens": 812532.0,
      "step": 5
    },
    {
      "aux_loss": 8.049106121063232,
      "entropy": 0.063042301684618,
      "epoch": 0.14457831325301204,
      "grad_norm": 0.5165722966194153,
      "learning_rate": 7.25e-07,
      "loss": 0.10346449911594391,
      "mean_token_accuracy": 0.987544983625412,
      "num_tokens": 973071.0,
      "step": 6
    },
    {
      "aux_loss": 8.047893047332764,
      "entropy": 0.0672433041036129,
      "epoch": 0.1686746987951807,
      "grad_norm": 0.5849524736404419,
      "learning_rate": 8.7e-07,
      "loss": 0.10626787692308426,
      "mean_token_accuracy": 0.9862557649612427,
      "num_tokens": 1110326.0,
      "step": 7
    },
    {
      "aux_loss": 8.04862642288208,
      "entropy": 0.06088166497647762,
      "epoch": 0.1927710843373494,
      "grad_norm": 0.48005586862564087,
      "learning_rate": 1.015e-06,
      "loss": 0.1021210104227066,
      "mean_token_accuracy": 0.9877255260944366,
      "num_tokens": 1259440.0,
      "step": 8
    },
    {
      "aux_loss": 8.047185897827148,
      "entropy": 0.049826065078377724,
      "epoch": 0.21686746987951808,
      "grad_norm": 0.43366947770118713,
      "learning_rate": 1.1600000000000001e-06,
      "loss": 0.09475715458393097,
      "mean_token_accuracy": 0.9898165464401245,
      "num_tokens": 1417756.0,
      "step": 9
    },
    {
      "aux_loss": 8.04772663116455,
      "entropy": 0.054957231506705284,
      "epoch": 0.24096385542168675,
      "grad_norm": 0.4529463052749634,
      "learning_rate": 1.3050000000000002e-06,
      "loss": 0.09896557033061981,
      "mean_token_accuracy": 0.9889253675937653,
      "num_tokens": 1579862.0,
      "step": 10
    },
    {
      "aux_loss": 8.047585487365723,
      "entropy": 0.05796742998063564,
      "epoch": 0.26506024096385544,
      "grad_norm": 0.45284125208854675,
      "learning_rate": 1.45e-06,
      "loss": 0.10001235455274582,
      "mean_token_accuracy": 0.9876790642738342,
      "num_tokens": 1731184.0,
      "step": 11
    },
    {
      "aux_loss": 8.04898738861084,
      "entropy": 0.05215233750641346,
      "epoch": 0.2891566265060241,
      "grad_norm": 0.4420299828052521,
      "learning_rate": 1.4499995904705798e-06,
      "loss": 0.09724222123622894,
      "mean_token_accuracy": 0.9884337782859802,
      "num_tokens": 1893932.0,
      "step": 12
    },
    {
      "aux_loss": 8.04798412322998,
      "entropy": 0.05265074409544468,
      "epoch": 0.3132530120481928,
      "grad_norm": 0.40204742550849915,
      "learning_rate": 1.4499983618832442e-06,
      "loss": 0.0970243290066719,
      "mean_token_accuracy": 0.989496260881424,
      "num_tokens": 2053005.0,
      "step": 13
    },
    {
      "aux_loss": 8.05030107498169,
      "entropy": 0.049178898334503174,
      "epoch": 0.3373493975903614,
      "grad_norm": 0.4108814597129822,
      "learning_rate": 1.4499963142407692e-06,
      "loss": 0.09499870240688324,
      "mean_token_accuracy": 0.9897345900535583,
      "num_tokens": 2217039.0,
      "step": 14
    },
    {
      "aux_loss": 8.046091079711914,
      "entropy": 0.0504351407289505,
      "epoch": 0.3614457831325301,
      "grad_norm": 0.4030505418777466,
      "learning_rate": 1.4499934475477814e-06,
      "loss": 0.09597203135490417,
      "mean_token_accuracy": 0.9898054599761963,
      "num_tokens": 2386769.0,
      "step": 15
    },
    {
      "aux_loss": 8.047934532165527,
      "entropy": 0.04648676887154579,
      "epoch": 0.3855421686746988,
      "grad_norm": 0.4123123586177826,
      "learning_rate": 1.4499897618107582e-06,
      "loss": 0.09345895051956177,
      "mean_token_accuracy": 0.989924967288971,
      "num_tokens": 2551093.0,
      "step": 16
    },
    {
      "aux_loss": 8.04793405532837,
      "entropy": 0.04946734011173248,
      "epoch": 0.40963855421686746,
      "grad_norm": 0.42739129066467285,
      "learning_rate": 1.4499852570380272e-06,
      "loss": 0.09633760154247284,
      "mean_token_accuracy": 0.9889810979366302,
      "num_tokens": 2704860.0,
      "step": 17
    },
    {
      "aux_loss": 8.048253536224365,
      "entropy": 0.0546275470405817,
      "epoch": 0.43373493975903615,
      "grad_norm": 0.44881728291511536,
      "learning_rate": 1.449979933239767e-06,
      "loss": 0.09820500761270523,
      "mean_token_accuracy": 0.9885384142398834,
      "num_tokens": 2850860.0,
      "step": 18
    },
    {
      "aux_loss": 8.050773620605469,
      "entropy": 0.05041976273059845,
      "epoch": 0.4578313253012048,
      "grad_norm": 0.46621593832969666,
      "learning_rate": 1.4499737904280063e-06,
      "loss": 0.09671641141176224,
      "mean_token_accuracy": 0.9887768924236298,
      "num_tokens": 3039323.0,
      "step": 19
    },
    {
      "aux_loss": 8.050614356994629,
      "entropy": 0.050554944202303886,
      "epoch": 0.4819277108433735,
      "grad_norm": 0.47147253155708313,
      "learning_rate": 1.449966828616625e-06,
      "loss": 0.09628961980342865,
      "mean_token_accuracy": 0.9888468682765961,
      "num_tokens": 3195488.0,
      "step": 20
    },
    {
      "aux_loss": 8.050609111785889,
      "entropy": 0.04489951767027378,
      "epoch": 0.5060240963855421,
      "grad_norm": 0.4890425503253937,
      "learning_rate": 1.4499590478213525e-06,
      "loss": 0.09394823014736176,
      "mean_token_accuracy": 0.9896246194839478,
      "num_tokens": 3369627.0,
      "step": 21
    },
    {
      "aux_loss": 8.047994613647461,
      "entropy": 0.0530005544424057,
      "epoch": 0.5301204819277109,
      "grad_norm": 0.5181686878204346,
      "learning_rate": 1.4499504480597702e-06,
      "loss": 0.09846185147762299,
      "mean_token_accuracy": 0.9879846572875977,
      "num_tokens": 3520643.0,
      "step": 22
    },
    {
      "aux_loss": 8.049047946929932,
      "entropy": 0.039870962500572205,
      "epoch": 0.5542168674698795,
      "grad_norm": 0.41603270173072815,
      "learning_rate": 1.4499410293513081e-06,
      "loss": 0.09075028449296951,
      "mean_token_accuracy": 0.9908623099327087,
      "num_tokens": 3702510.0,
      "step": 23
    },
    {
      "aux_loss": 8.050177097320557,
      "entropy": 0.05053388141095638,
      "epoch": 0.5783132530120482,
      "grad_norm": 0.4234795570373535,
      "learning_rate": 1.449930791717248e-06,
      "loss": 0.09689382463693619,
      "mean_token_accuracy": 0.9886369109153748,
      "num_tokens": 3867610.0,
      "step": 24
    },
    {
      "aux_loss": 8.048462867736816,
      "entropy": 0.049235520884394646,
      "epoch": 0.6024096385542169,
      "grad_norm": 0.4375890791416168,
      "learning_rate": 1.4499197351807219e-06,
      "loss": 0.09504671394824982,
      "mean_token_accuracy": 0.9891063868999481,
      "num_tokens": 4021186.0,
      "step": 25
    },
    {
      "aux_loss": 8.050186157226562,
      "entropy": 0.048021236434578896,
      "epoch": 0.6265060240963856,
      "grad_norm": 0.42389488220214844,
      "learning_rate": 1.449907859766711e-06,
      "loss": 0.09462042152881622,
      "mean_token_accuracy": 0.9893973469734192,
      "num_tokens": 4183102.0,
      "step": 26
    },
    {
      "aux_loss": 8.04945182800293,
      "entropy": 0.042689668014645576,
      "epoch": 0.6506024096385542,
      "grad_norm": 0.40038207173347473,
      "learning_rate": 1.4498951655020479e-06,
      "loss": 0.09214234352111816,
      "mean_token_accuracy": 0.9904394745826721,
      "num_tokens": 4348628.0,
      "step": 27
    },
    {
      "aux_loss": 8.050597667694092,
      "entropy": 0.05143065005540848,
      "epoch": 0.6746987951807228,
      "grad_norm": 0.4422961473464966,
      "learning_rate": 1.4498816524154148e-06,
      "loss": 0.09722501039505005,
      "mean_token_accuracy": 0.9888282418251038,
      "num_tokens": 4511217.0,
      "step": 28
    },
    {
      "aux_loss": 8.048704147338867,
      "entropy": 0.044764744117856026,
      "epoch": 0.6987951807228916,
      "grad_norm": 0.4279940128326416,
      "learning_rate": 1.4498673205373443e-06,
      "loss": 0.09300120174884796,
      "mean_token_accuracy": 0.9899511337280273,
      "num_tokens": 4684266.0,
      "step": 29
    },
    {
      "aux_loss": 8.047113418579102,
      "entropy": 0.04995520040392876,
      "epoch": 0.7228915662650602,
      "grad_norm": 0.45497408509254456,
      "learning_rate": 1.4498521699002187e-06,
      "loss": 0.0966445729136467,
      "mean_token_accuracy": 0.9883378744125366,
      "num_tokens": 4858009.0,
      "step": 30
    },
    {
      "aux_loss": 8.050236701965332,
      "entropy": 0.04773502238094807,
      "epoch": 0.7469879518072289,
      "grad_norm": 0.40059739351272583,
      "learning_rate": 1.4498362005382707e-06,
      "loss": 0.09425598382949829,
      "mean_token_accuracy": 0.9895853698253632,
      "num_tokens": 5041560.0,
      "step": 31
    },
    {
      "aux_loss": 8.0483078956604,
      "entropy": 0.04463958367705345,
      "epoch": 0.7710843373493976,
      "grad_norm": 0.421403706073761,
      "learning_rate": 1.4498194124875822e-06,
      "loss": 0.0921817421913147,
      "mean_token_accuracy": 0.9903146624565125,
      "num_tokens": 5207925.0,
      "step": 32
    },
    {
      "aux_loss": 8.051517486572266,
      "entropy": 0.0455746091902256,
      "epoch": 0.7951807228915663,
      "grad_norm": 0.39467114210128784,
      "learning_rate": 1.449801805786086e-06,
      "loss": 0.09311801195144653,
      "mean_token_accuracy": 0.9897293448448181,
      "num_tokens": 5410900.0,
      "step": 33
    },
    {
      "aux_loss": 8.047285079956055,
      "entropy": 0.04475598596036434,
      "epoch": 0.8192771084337349,
      "grad_norm": 0.3885505795478821,
      "learning_rate": 1.4497833804735631e-06,
      "loss": 0.09283807128667831,
      "mean_token_accuracy": 0.9904457926750183,
      "num_tokens": 5579379.0,
      "step": 34
    },
    {
      "aux_loss": 8.049206733703613,
      "entropy": 0.045768845826387405,
      "epoch": 0.8433734939759037,
      "grad_norm": 0.38822969794273376,
      "learning_rate": 1.4497641365916457e-06,
      "loss": 0.0930660218000412,
      "mean_token_accuracy": 0.9898505806922913,
      "num_tokens": 5756181.0,
      "step": 35
    },
    {
      "aux_loss": 8.049548149108887,
      "entropy": 0.0442772526293993,
      "epoch": 0.8674698795180723,
      "grad_norm": 0.42765918374061584,
      "learning_rate": 1.4497440741838147e-06,
      "loss": 0.09213966131210327,
      "mean_token_accuracy": 0.9906859695911407,
      "num_tokens": 5913888.0,
      "step": 36
    },
    {
      "aux_loss": 8.047310829162598,
      "entropy": 0.05137152969837189,
      "epoch": 0.891566265060241,
      "grad_norm": 0.46749433875083923,
      "learning_rate": 1.4497231932954002e-06,
      "loss": 0.09635019302368164,
      "mean_token_accuracy": 0.9894134104251862,
      "num_tokens": 6073822.0,
      "step": 37
    },
    {
      "aux_loss": 8.04575777053833,
      "entropy": 0.04952680319547653,
      "epoch": 0.9156626506024096,
      "grad_norm": 0.40102288126945496,
      "learning_rate": 1.4497014939735825e-06,
      "loss": 0.09520888328552246,
      "mean_token_accuracy": 0.9900149703025818,
      "num_tokens": 6228998.0,
      "step": 38
    },
    {
      "aux_loss": 8.05048656463623,
      "entropy": 0.05141362361609936,
      "epoch": 0.9397590361445783,
      "grad_norm": 0.45891159772872925,
      "learning_rate": 1.4496789762673903e-06,
      "loss": 0.09608915448188782,
      "mean_token_accuracy": 0.9895050525665283,
      "num_tokens": 6380479.0,
      "step": 39
    },
    {
      "aux_loss": 8.049390316009521,
      "entropy": 0.0542642492800951,
      "epoch": 0.963855421686747,
      "grad_norm": 0.4678654968738556,
      "learning_rate": 1.4496556402277018e-06,
      "loss": 0.0986521765589714,
      "mean_token_accuracy": 0.9882148206233978,
      "num_tokens": 6514347.0,
      "step": 40
    },
    {
      "aux_loss": 8.047286033630371,
      "entropy": 0.04129311256110668,
      "epoch": 0.9879518072289156,
      "grad_norm": 0.42529988288879395,
      "learning_rate": 1.4496314859072443e-06,
      "loss": 0.09138549864292145,
      "mean_token_accuracy": 0.9908298254013062,
      "num_tokens": 6684739.0,
      "step": 41
    },
    {
      "aux_loss": 8.050637245178223,
      "entropy": 0.04988313093781471,
      "epoch": 1.0,
      "grad_norm": 0.5628013610839844,
      "learning_rate": 1.4496065133605937e-06,
      "loss": 0.06163128837943077,
      "mean_token_accuracy": 0.9887231588363647,
      "num_tokens": 6741463.0,
      "step": 42
    },
    {
      "aux_loss": 8.047677040100098,
      "entropy": 0.04833286814391613,
      "epoch": 1.0240963855421688,
      "grad_norm": 0.43433839082717896,
      "learning_rate": 1.4495807226441745e-06,
      "loss": 0.09322302043437958,
      "mean_token_accuracy": 0.9907335042953491,
      "num_tokens": 6872162.0,
      "step": 43
    },
    {
      "aux_loss": 8.049845218658447,
      "entropy": 0.04814961925148964,
      "epoch": 1.0481927710843373,
      "grad_norm": 0.38683706521987915,
      "learning_rate": 1.4495541138162606e-06,
      "loss": 0.09274052828550339,
      "mean_token_accuracy": 0.9911822378635406,
      "num_tokens": 7044917.0,
      "step": 44
    },
    {
      "aux_loss": 8.045840740203857,
      "entropy": 0.04862277954816818,
      "epoch": 1.072289156626506,
      "grad_norm": 0.43781232833862305,
      "learning_rate": 1.4495266869369737e-06,
      "loss": 0.09285125136375427,
      "mean_token_accuracy": 0.991097629070282,
      "num_tokens": 7207551.0,
      "step": 45
    },
    {
      "aux_loss": 8.047110080718994,
      "entropy": 0.04816548526287079,
      "epoch": 1.0963855421686748,
      "grad_norm": 0.4117249846458435,
      "learning_rate": 1.4494984420682841e-06,
      "loss": 0.09330926835536957,
      "mean_token_accuracy": 0.990231990814209,
      "num_tokens": 7376391.0,
      "step": 46
    },
    {
      "aux_loss": 8.046239852905273,
      "entropy": 0.04640440456569195,
      "epoch": 1.1204819277108433,
      "grad_norm": 0.3918110728263855,
      "learning_rate": 1.4494693792740101e-06,
      "loss": 0.09266579151153564,
      "mean_token_accuracy": 0.9905861914157867,
      "num_tokens": 7544885.0,
      "step": 47
    },
    {
      "aux_loss": 8.045721054077148,
      "entropy": 0.04254587180912495,
      "epoch": 1.144578313253012,
      "grad_norm": 0.3635147213935852,
      "learning_rate": 1.4494394986198189e-06,
      "loss": 0.08964599668979645,
      "mean_token_accuracy": 0.9919890761375427,
      "num_tokens": 7718825.0,
      "step": 48
    },
    {
      "aux_loss": 8.04783582687378,
      "entropy": 0.0457068495452404,
      "epoch": 1.1686746987951806,
      "grad_norm": 0.37491077184677124,
      "learning_rate": 1.4494088001732242e-06,
      "loss": 0.0912659764289856,
      "mean_token_accuracy": 0.9914940893650055,
      "num_tokens": 7884797.0,
      "step": 49
    },
    {
      "aux_loss": 8.048118591308594,
      "entropy": 0.05809796415269375,
      "epoch": 1.1927710843373494,
      "grad_norm": 0.442947119474411,
      "learning_rate": 1.449377284003589e-06,
      "loss": 0.09830181300640106,
      "mean_token_accuracy": 0.9896205961704254,
      "num_tokens": 8037756.0,
      "step": 50
    },
    {
      "aux_loss": 8.045411109924316,
      "entropy": 0.04734855704009533,
      "epoch": 1.216867469879518,
      "grad_norm": 0.3778594732284546,
      "learning_rate": 1.449344950182123e-06,
      "loss": 0.09220214188098907,
      "mean_token_accuracy": 0.9912082552909851,
      "num_tokens": 8201900.0,
      "step": 51
    },
    {
      "aux_loss": 8.051240921020508,
      "entropy": 0.04616173356771469,
      "epoch": 1.2409638554216866,
      "grad_norm": 0.39906826615333557,
      "learning_rate": 1.4493117987818833e-06,
      "loss": 0.09254124015569687,
      "mean_token_accuracy": 0.9910232424736023,
      "num_tokens": 8368482.0,
      "step": 52
    },
    {
      "aux_loss": 8.048255443572998,
      "entropy": 0.04362104274332523,
      "epoch": 1.2650602409638554,
      "grad_norm": 0.4132522642612457,
      "learning_rate": 1.449277829877775e-06,
      "loss": 0.08969354629516602,
      "mean_token_accuracy": 0.9923582375049591,
      "num_tokens": 8532083.0,
      "step": 53
    },
    {
      "aux_loss": 8.049753189086914,
      "entropy": 0.05157726630568504,
      "epoch": 1.2891566265060241,
      "grad_norm": 0.42263299226760864,
      "learning_rate": 1.4492430435465495e-06,
      "loss": 0.09429101645946503,
      "mean_token_accuracy": 0.9907614588737488,
      "num_tokens": 8707916.0,
      "step": 54
    },
    {
      "aux_loss": 8.04571533203125,
      "entropy": 0.04540949687361717,
      "epoch": 1.3132530120481927,
      "grad_norm": 0.415887713432312,
      "learning_rate": 1.4492074398668059e-06,
      "loss": 0.09169892221689224,
      "mean_token_accuracy": 0.9909594357013702,
      "num_tokens": 8868293.0,
      "step": 55
    },
    {
      "aux_loss": 8.047487258911133,
      "entropy": 0.042821500450372696,
      "epoch": 1.3373493975903614,
      "grad_norm": 0.3941069543361664,
      "learning_rate": 1.4491710189189896e-06,
      "loss": 0.08963996917009354,
      "mean_token_accuracy": 0.9915030598640442,
      "num_tokens": 9040400.0,
      "step": 56
    },
    {
      "aux_loss": 8.046500205993652,
      "entropy": 0.04034057445824146,
      "epoch": 1.3614457831325302,
      "grad_norm": 0.3714058995246887,
      "learning_rate": 1.4491337807853928e-06,
      "loss": 0.08831468969583511,
      "mean_token_accuracy": 0.9924418330192566,
      "num_tokens": 9210005.0,
      "step": 57
    },
    {
      "aux_loss": 8.04520559310913,
      "entropy": 0.04288036748766899,
      "epoch": 1.3855421686746987,
      "grad_norm": 0.4117484986782074,
      "learning_rate": 1.4490957255501541e-06,
      "loss": 0.09062028676271439,
      "mean_token_accuracy": 0.9908726513385773,
      "num_tokens": 9354587.0,
      "step": 58
    },
    {
      "aux_loss": 8.048189163208008,
      "entropy": 0.04117992892861366,
      "epoch": 1.4096385542168675,
      "grad_norm": 0.3880350589752197,
      "learning_rate": 1.449056853299258e-06,
      "loss": 0.08878283947706223,
      "mean_token_accuracy": 0.9918811917304993,
      "num_tokens": 9530098.0,
      "step": 59
    },
    {
      "aux_loss": 8.046634674072266,
      "entropy": 0.046307843178510666,
      "epoch": 1.4337349397590362,
      "grad_norm": 0.41205379366874695,
      "learning_rate": 1.4490171641205356e-06,
      "loss": 0.0928514301776886,
      "mean_token_accuracy": 0.99020916223526,
      "num_tokens": 9696846.0,
      "step": 60
    },
    {
      "aux_loss": 8.04638147354126,
      "entropy": 0.03861385025084019,
      "epoch": 1.4578313253012047,
      "grad_norm": 0.3674647808074951,
      "learning_rate": 1.4489766581036633e-06,
      "loss": 0.08779378235340118,
      "mean_token_accuracy": 0.9924313127994537,
      "num_tokens": 9868180.0,
      "step": 61
    },
    {
      "aux_loss": 8.048831462860107,
      "entropy": 0.04485859349370003,
      "epoch": 1.4819277108433735,
      "grad_norm": 0.4122410714626312,
      "learning_rate": 1.4489353353401633e-06,
      "loss": 0.09213700890541077,
      "mean_token_accuracy": 0.9905197322368622,
      "num_tokens": 10041207.0,
      "step": 62
    },
    {
      "aux_loss": 8.045693397521973,
      "entropy": 0.04181382618844509,
      "epoch": 1.5060240963855422,
      "grad_norm": 0.41252797842025757,
      "learning_rate": 1.4488931959234036e-06,
      "loss": 0.08966240286827087,
      "mean_token_accuracy": 0.9918944835662842,
      "num_tokens": 10216601.0,
      "step": 63
    },
    {
      "aux_loss": 8.045026779174805,
      "entropy": 0.04367237165570259,
      "epoch": 1.5301204819277108,
      "grad_norm": 0.7194992899894714,
      "learning_rate": 1.4488502399485967e-06,
      "loss": 0.09064353257417679,
      "mean_token_accuracy": 0.9913205206394196,
      "num_tokens": 10373157.0,
      "step": 64
    },
    {
      "aux_loss": 8.046749591827393,
      "entropy": 0.049618642777204514,
      "epoch": 1.5542168674698795,
      "grad_norm": 0.4409625828266144,
      "learning_rate": 1.4488064675128008e-06,
      "loss": 0.09418876469135284,
      "mean_token_accuracy": 0.98995640873909,
      "num_tokens": 10513760.0,
      "step": 65
    },
    {
      "aux_loss": 8.04750108718872,
      "entropy": 0.045051392167806625,
      "epoch": 1.5783132530120483,
      "grad_norm": 0.4379954934120178,
      "learning_rate": 1.4487618787149178e-06,
      "loss": 0.09229664504528046,
      "mean_token_accuracy": 0.990358978509903,
      "num_tokens": 10667072.0,
      "step": 66
    },
    {
      "aux_loss": 8.0473952293396,
      "entropy": 0.03963223472237587,
      "epoch": 1.6024096385542168,
      "grad_norm": 0.3720748722553253,
      "learning_rate": 1.4487164736556958e-06,
      "loss": 0.08866938948631287,
      "mean_token_accuracy": 0.9920879304409027,
      "num_tokens": 10839108.0,
      "step": 67
    },
    {
      "aux_loss": 8.047708511352539,
      "entropy": 0.04451095871627331,
      "epoch": 1.6265060240963856,
      "grad_norm": 0.4250375032424927,
      "learning_rate": 1.4486702524377257e-06,
      "loss": 0.09129409492015839,
      "mean_token_accuracy": 0.9904707074165344,
      "num_tokens": 11001747.0,
      "step": 68
    },
    {
      "aux_loss": 8.045770645141602,
      "entropy": 0.041699016466736794,
      "epoch": 1.6506024096385543,
      "grad_norm": 0.5749812722206116,
      "learning_rate": 1.4486232151654433e-06,
      "loss": 0.08955338597297668,
      "mean_token_accuracy": 0.9914253950119019,
      "num_tokens": 11163010.0,
      "step": 69
    },
    {
      "aux_loss": 8.048367977142334,
      "entropy": 0.0460666548460722,
      "epoch": 1.6746987951807228,
      "grad_norm": 0.412992388010025,
      "learning_rate": 1.448575361945128e-06,
      "loss": 0.09151523560285568,
      "mean_token_accuracy": 0.9911079108715057,
      "num_tokens": 11326488.0,
      "step": 70
    },
    {
      "aux_loss": 8.047677040100098,
      "entropy": 0.039530523121356964,
      "epoch": 1.6987951807228916,
      "grad_norm": 0.3757312297821045,
      "learning_rate": 1.4485266928849028e-06,
      "loss": 0.08811481297016144,
      "mean_token_accuracy": 0.9917347729206085,
      "num_tokens": 11501286.0,
      "step": 71
    },
    {
      "aux_loss": 8.046856880187988,
      "entropy": 0.041672926396131516,
      "epoch": 1.7228915662650603,
      "grad_norm": 0.38275983929634094,
      "learning_rate": 1.448477208094734e-06,
      "loss": 0.08906584978103638,
      "mean_token_accuracy": 0.9918383359909058,
      "num_tokens": 11668188.0,
      "step": 72
    },
    {
      "aux_loss": 8.04528522491455,
      "entropy": 0.047908855602145195,
      "epoch": 1.7469879518072289,
      "grad_norm": 0.4008839726448059,
      "learning_rate": 1.4484269076864313e-06,
      "loss": 0.09260185062885284,
      "mean_token_accuracy": 0.9907302260398865,
      "num_tokens": 11834809.0,
      "step": 73
    },
    {
      "aux_loss": 8.047989845275879,
      "entropy": 0.049024928361177444,
      "epoch": 1.7710843373493976,
      "grad_norm": 0.43533122539520264,
      "learning_rate": 1.448375791773647e-06,
      "loss": 0.09409160912036896,
      "mean_token_accuracy": 0.9904188215732574,
      "num_tokens": 12004510.0,
      "step": 74
    },
    {
      "aux_loss": 8.045915126800537,
      "entropy": 0.05250510759651661,
      "epoch": 1.7951807228915664,
      "grad_norm": 0.4690086841583252,
      "learning_rate": 1.4483238604718765e-06,
      "loss": 0.0955488458275795,
      "mean_token_accuracy": 0.9906903803348541,
      "num_tokens": 12137534.0,
      "step": 75
    },
    {
      "aux_loss": 8.044984817504883,
      "entropy": 0.0376634681597352,
      "epoch": 1.819277108433735,
      "grad_norm": 0.35793742537498474,
      "learning_rate": 1.4482711138984564e-06,
      "loss": 0.08688877522945404,
      "mean_token_accuracy": 0.9927193224430084,
      "num_tokens": 12325369.0,
      "step": 76
    },
    {
      "aux_loss": 8.04779052734375,
      "entropy": 0.04571797512471676,
      "epoch": 1.8433734939759037,
      "grad_norm": 0.4300045371055603,
      "learning_rate": 1.4482175521725665e-06,
      "loss": 0.09217096120119095,
      "mean_token_accuracy": 0.9905279278755188,
      "num_tokens": 12488806.0,
      "step": 77
    },
    {
      "aux_loss": 8.044951915740967,
      "entropy": 0.05060774274170399,
      "epoch": 1.8674698795180724,
      "grad_norm": 0.423297643661499,
      "learning_rate": 1.4481631754152286e-06,
      "loss": 0.0944666936993599,
      "mean_token_accuracy": 0.990161120891571,
      "num_tokens": 12640695.0,
      "step": 78
    },
    {
      "aux_loss": 8.047436714172363,
      "entropy": 0.04260541498661041,
      "epoch": 1.891566265060241,
      "grad_norm": 0.39964279532432556,
      "learning_rate": 1.4481079837493048e-06,
      "loss": 0.08988695591688156,
      "mean_token_accuracy": 0.9916346371173859,
      "num_tokens": 12823930.0,
      "step": 79
    },
    {
      "aux_loss": 8.047076225280762,
      "entropy": 0.04447777755558491,
      "epoch": 1.9156626506024095,
      "grad_norm": 0.3796585202217102,
      "learning_rate": 1.4480519772994993e-06,
      "loss": 0.09115425497293472,
      "mean_token_accuracy": 0.9917284250259399,
      "num_tokens": 12998419.0,
      "step": 80
    },
    {
      "aux_loss": 8.04789924621582,
      "entropy": 0.04236619919538498,
      "epoch": 1.9397590361445785,
      "grad_norm": 0.418720543384552,
      "learning_rate": 1.4479951561923573e-06,
      "loss": 0.0903579443693161,
      "mean_token_accuracy": 0.9915480315685272,
      "num_tokens": 13148913.0,
      "step": 81
    },
    {
      "aux_loss": 8.045738697052002,
      "entropy": 0.04441889189183712,
      "epoch": 1.963855421686747,
      "grad_norm": 0.3862975835800171,
      "learning_rate": 1.4479375205562642e-06,
      "loss": 0.09073998779058456,
      "mean_token_accuracy": 0.9912832081317902,
      "num_tokens": 13304132.0,
      "step": 82
    },
    {
      "aux_loss": 8.046059608459473,
      "entropy": 0.05218968167901039,
      "epoch": 1.9879518072289155,
      "grad_norm": 0.4317293167114258,
      "learning_rate": 1.4478790705214464e-06,
      "loss": 0.09526760876178741,
      "mean_token_accuracy": 0.9900495111942291,
      "num_tokens": 13453623.0,
      "step": 83
    },
    {
      "aux_loss": 8.05069351196289,
      "entropy": 0.039391469210386276,
      "epoch": 2.0,
      "grad_norm": 0.6828374266624451,
      "learning_rate": 1.4478198062199706e-06,
      "loss": 0.05737518146634102,
      "mean_token_accuracy": 0.9917781352996826,
      "num_tokens": 13482926.0,
      "step": 84
    },
    {
      "aux_loss": 8.045267105102539,
      "entropy": 0.053030213341116905,
      "epoch": 2.0240963855421685,
      "grad_norm": 0.40254807472229004,
      "learning_rate": 1.4477597277857421e-06,
      "loss": 0.09399621188640594,
      "mean_token_accuracy": 0.9910189509391785,
      "num_tokens": 13616336.0,
      "step": 85
    },
    {
      "aux_loss": 8.04813003540039,
      "entropy": 0.04406110197305679,
      "epoch": 2.0481927710843375,
      "grad_norm": 0.36270391941070557,
      "learning_rate": 1.4476988353545068e-06,
      "loss": 0.08861124515533447,
      "mean_token_accuracy": 0.9931189119815826,
      "num_tokens": 13775230.0,
      "step": 86
    },
    {
      "aux_loss": 8.044662952423096,
      "entropy": 0.03888658061623573,
      "epoch": 2.072289156626506,
      "grad_norm": 0.3652919828891754,
      "learning_rate": 1.4476371290638496e-06,
      "loss": 0.08651480078697205,
      "mean_token_accuracy": 0.9931168258190155,
      "num_tokens": 13935092.0,
      "step": 87
    },
    {
      "aux_loss": 8.048104763031006,
      "entropy": 0.047589803114533424,
      "epoch": 2.0963855421686746,
      "grad_norm": 0.4187248647212982,
      "learning_rate": 1.447574609053194e-06,
      "loss": 0.09169710427522659,
      "mean_token_accuracy": 0.9922028183937073,
      "num_tokens": 14097222.0,
      "step": 88
    },
    {
      "aux_loss": 8.044262409210205,
      "entropy": 0.04133613780140877,
      "epoch": 2.1204819277108435,
      "grad_norm": 0.3797593414783478,
      "learning_rate": 1.4475112754638022e-06,
      "loss": 0.08802482485771179,
      "mean_token_accuracy": 0.9926626980304718,
      "num_tokens": 14255553.0,
      "step": 89
    },
    {
      "aux_loss": 8.046658515930176,
      "entropy": 0.04000333696603775,
      "epoch": 2.144578313253012,
      "grad_norm": 0.4277676045894623,
      "learning_rate": 1.447447128438775e-06,
      "loss": 0.08734406530857086,
      "mean_token_accuracy": 0.993060439825058,
      "num_tokens": 14431608.0,
      "step": 90
    },
    {
      "aux_loss": 8.045912265777588,
      "entropy": 0.041802020743489265,
      "epoch": 2.1686746987951806,
      "grad_norm": 0.37397268414497375,
      "learning_rate": 1.4473821681230506e-06,
      "loss": 0.08865631371736526,
      "mean_token_accuracy": 0.9919825196266174,
      "num_tokens": 14615993.0,
      "step": 91
    },
    {
      "aux_loss": 8.044597625732422,
      "entropy": 0.04039378650486469,
      "epoch": 2.1927710843373496,
      "grad_norm": 0.423145592212677,
      "learning_rate": 1.4473163946634052e-06,
      "loss": 0.0880887508392334,
      "mean_token_accuracy": 0.9921639561653137,
      "num_tokens": 14766575.0,
      "step": 92
    },
    {
      "aux_loss": 8.04832410812378,
      "entropy": 0.04441938176751137,
      "epoch": 2.216867469879518,
      "grad_norm": 0.3814225196838379,
      "learning_rate": 1.4472498082084518e-06,
      "loss": 0.08906400948762894,
      "mean_token_accuracy": 0.9928748905658722,
      "num_tokens": 14931445.0,
      "step": 93
    },
    {
      "aux_loss": 8.042942523956299,
      "entropy": 0.046976158395409584,
      "epoch": 2.2409638554216866,
      "grad_norm": 0.3922480344772339,
      "learning_rate": 1.4471824089086415e-06,
      "loss": 0.09150424599647522,
      "mean_token_accuracy": 0.9916513860225677,
      "num_tokens": 15091210.0,
      "step": 94
    },
    {
      "aux_loss": 8.045757293701172,
      "entropy": 0.035969141870737076,
      "epoch": 2.2650602409638556,
      "grad_norm": 0.3510456681251526,
      "learning_rate": 1.44711419691626e-06,
      "loss": 0.0854613184928894,
      "mean_token_accuracy": 0.9934249818325043,
      "num_tokens": 15260222.0,
      "step": 95
    },
    {
      "aux_loss": 8.044999122619629,
      "entropy": 0.03909716196358204,
      "epoch": 2.289156626506024,
      "grad_norm": 0.3629781901836395,
      "learning_rate": 1.4470451723854314e-06,
      "loss": 0.08637961745262146,
      "mean_token_accuracy": 0.9935233891010284,
      "num_tokens": 15455860.0,
      "step": 96
    },
    {
      "aux_loss": 8.04701280593872,
      "entropy": 0.04057766683399677,
      "epoch": 2.3132530120481927,
      "grad_norm": 0.36879613995552063,
      "learning_rate": 1.4469753354721144e-06,
      "loss": 0.087168388068676,
      "mean_token_accuracy": 0.9932206869125366,
      "num_tokens": 15631021.0,
      "step": 97
    },
    {
      "aux_loss": 8.047924518585205,
      "entropy": 0.04564652778208256,
      "epoch": 2.337349397590361,
      "grad_norm": 0.41153064370155334,
      "learning_rate": 1.4469046863341036e-06,
      "loss": 0.08925162255764008,
      "mean_token_accuracy": 0.9921718537807465,
      "num_tokens": 15789136.0,
      "step": 98
    },
    {
      "aux_loss": 8.0452880859375,
      "entropy": 0.04301636107265949,
      "epoch": 2.36144578313253,
      "grad_norm": 0.376197874546051,
      "learning_rate": 1.4468332251310286e-06,
      "loss": 0.08848991990089417,
      "mean_token_accuracy": 0.9930213391780853,
      "num_tokens": 15938397.0,
      "step": 99
    },
    {
      "aux_loss": 8.043729782104492,
      "entropy": 0.04113657400012016,
      "epoch": 2.3855421686746987,
      "grad_norm": 0.3817802667617798,
      "learning_rate": 1.4467609520243544e-06,
      "loss": 0.08791770040988922,
      "mean_token_accuracy": 0.9926939308643341,
      "num_tokens": 16114803.0,
      "step": 100
    },
    {
      "aux_loss": 8.044564247131348,
      "entropy": 0.04448202811181545,
      "epoch": 2.4096385542168672,
      "grad_norm": 0.41418957710266113,
      "learning_rate": 1.44668786717738e-06,
      "loss": 0.08933436870574951,
      "mean_token_accuracy": 0.9923655092716217,
      "num_tokens": 16267681.0,
      "step": 101
    },
    {
      "aux_loss": 8.045689105987549,
      "entropy": 0.043511977419257164,
      "epoch": 2.433734939759036,
      "grad_norm": 0.3912562429904938,
      "learning_rate": 1.4466139707552382e-06,
      "loss": 0.08852104097604752,
      "mean_token_accuracy": 0.9933134019374847,
      "num_tokens": 16435524.0,
      "step": 102
    },
    {
      "aux_loss": 8.04603910446167,
      "entropy": 0.03945927880704403,
      "epoch": 2.4578313253012047,
      "grad_norm": 0.37696266174316406,
      "learning_rate": 1.4465392629248964e-06,
      "loss": 0.08802925050258636,
      "mean_token_accuracy": 0.9925020635128021,
      "num_tokens": 16601843.0,
      "step": 103
    },
    {
      "aux_loss": 8.041633129119873,
      "entropy": 0.042852023616433144,
      "epoch": 2.4819277108433733,
      "grad_norm": 0.42523500323295593,
      "learning_rate": 1.446463743855155e-06,
      "loss": 0.08881749212741852,
      "mean_token_accuracy": 0.9926598370075226,
      "num_tokens": 16749225.0,
      "step": 104
    },
    {
      "aux_loss": 8.044000625610352,
      "entropy": 0.04509095661342144,
      "epoch": 2.5060240963855422,
      "grad_norm": 0.43666303157806396,
      "learning_rate": 1.4463874137166468e-06,
      "loss": 0.09076739847660065,
      "mean_token_accuracy": 0.9915396869182587,
      "num_tokens": 16897018.0,
      "step": 105
    },
    {
      "aux_loss": 8.044254302978516,
      "entropy": 0.04391391761600971,
      "epoch": 2.5301204819277108,
      "grad_norm": 0.4111824333667755,
      "learning_rate": 1.4463102726818382e-06,
      "loss": 0.08995572477579117,
      "mean_token_accuracy": 0.9917828142642975,
      "num_tokens": 17065753.0,
      "step": 106
    },
    {
      "aux_loss": 8.045661926269531,
      "entropy": 0.04082946851849556,
      "epoch": 2.5542168674698793,
      "grad_norm": 0.3910435438156128,
      "learning_rate": 1.4462323209250268e-06,
      "loss": 0.08806779235601425,
      "mean_token_accuracy": 0.9928287863731384,
      "num_tokens": 17235799.0,
      "step": 107
    },
    {
      "aux_loss": 8.043983459472656,
      "entropy": 0.045222554355859756,
      "epoch": 2.5783132530120483,
      "grad_norm": 0.4386132061481476,
      "learning_rate": 1.4461535586223427e-06,
      "loss": 0.090940460562706,
      "mean_token_accuracy": 0.9915510416030884,
      "num_tokens": 17376992.0,
      "step": 108
    },
    {
      "aux_loss": 8.0429105758667,
      "entropy": 0.04239073768258095,
      "epoch": 2.602409638554217,
      "grad_norm": 0.39754438400268555,
      "learning_rate": 1.4460739859517472e-06,
      "loss": 0.08899098634719849,
      "mean_token_accuracy": 0.9922184944152832,
      "num_tokens": 17537095.0,
      "step": 109
    },
    {
      "aux_loss": 8.046865940093994,
      "entropy": 0.04032203182578087,
      "epoch": 2.6265060240963853,
      "grad_norm": 0.4198264181613922,
      "learning_rate": 1.4459936030930326e-06,
      "loss": 0.08784371614456177,
      "mean_token_accuracy": 0.9927988350391388,
      "num_tokens": 17697756.0,
      "step": 110
    },
    {
      "aux_loss": 8.045022964477539,
      "entropy": 0.04641672037541866,
      "epoch": 2.6506024096385543,
      "grad_norm": 0.41483283042907715,
      "learning_rate": 1.4459124102278213e-06,
      "loss": 0.0906679630279541,
      "mean_token_accuracy": 0.9918768405914307,
      "num_tokens": 17862032.0,
      "step": 111
    },
    {
      "aux_loss": 8.046215057373047,
      "entropy": 0.038604725152254105,
      "epoch": 2.674698795180723,
      "grad_norm": 0.38772159814834595,
      "learning_rate": 1.4458304075395668e-06,
      "loss": 0.08620807528495789,
      "mean_token_accuracy": 0.9930865466594696,
      "num_tokens": 18014648.0,
      "step": 112
    },
    {
      "aux_loss": 8.04534912109375,
      "entropy": 0.047094520181417465,
      "epoch": 2.6987951807228914,
      "grad_norm": 0.4265570342540741,
      "learning_rate": 1.445747595213552e-06,
      "loss": 0.09142982959747314,
      "mean_token_accuracy": 0.9916557669639587,
      "num_tokens": 18170137.0,
      "step": 113
    },
    {
      "aux_loss": 8.045645713806152,
      "entropy": 0.037863900884985924,
      "epoch": 2.7228915662650603,
      "grad_norm": 0.3691053092479706,
      "learning_rate": 1.4456639734368888e-06,
      "loss": 0.08656749129295349,
      "mean_token_accuracy": 0.9931949079036713,
      "num_tokens": 18341196.0,
      "step": 114
    },
    {
      "aux_loss": 8.04396677017212,
      "entropy": 0.04190782643854618,
      "epoch": 2.746987951807229,
      "grad_norm": 0.38775166869163513,
      "learning_rate": 1.4455795423985187e-06,
      "loss": 0.08832530677318573,
      "mean_token_accuracy": 0.9925132691860199,
      "num_tokens": 18499534.0,
      "step": 115
    },
    {
      "aux_loss": 8.046361446380615,
      "entropy": 0.044008735567331314,
      "epoch": 2.7710843373493974,
      "grad_norm": 0.3934473395347595,
      "learning_rate": 1.445494302289211e-06,
      "loss": 0.08926442265510559,
      "mean_token_accuracy": 0.9925075769424438,
      "num_tokens": 18659760.0,
      "step": 116
    },
    {
      "aux_loss": 8.044501304626465,
      "entropy": 0.04429285787045956,
      "epoch": 2.7951807228915664,
      "grad_norm": 0.4424228072166443,
      "learning_rate": 1.4454082533015636e-06,
      "loss": 0.08979617059230804,
      "mean_token_accuracy": 0.9921565651893616,
      "num_tokens": 18831165.0,
      "step": 117
    },
    {
      "aux_loss": 8.042979717254639,
      "entropy": 0.040504178032279015,
      "epoch": 2.819277108433735,
      "grad_norm": 0.4018898904323578,
      "learning_rate": 1.4453213956300015e-06,
      "loss": 0.08786608278751373,
      "mean_token_accuracy": 0.992399662733078,
      "num_tokens": 18984127.0,
      "step": 118
    },
    {
      "aux_loss": 8.045685768127441,
      "entropy": 0.041042402386665344,
      "epoch": 2.8433734939759034,
      "grad_norm": 0.5288944840431213,
      "learning_rate": 1.4452337294707773e-06,
      "loss": 0.08859241008758545,
      "mean_token_accuracy": 0.9921814501285553,
      "num_tokens": 19160419.0,
      "step": 119
    },
    {
      "aux_loss": 8.046215057373047,
      "entropy": 0.04414335638284683,
      "epoch": 2.8674698795180724,
      "grad_norm": 0.4066352844238281,
      "learning_rate": 1.4451452550219706e-06,
      "loss": 0.08978652209043503,
      "mean_token_accuracy": 0.9923945069313049,
      "num_tokens": 19335546.0,
      "step": 120
    },
    {
      "aux_loss": 8.044774055480957,
      "entropy": 0.03940526768565178,
      "epoch": 2.891566265060241,
      "grad_norm": 0.4042900800704956,
      "learning_rate": 1.445055972483487e-06,
      "loss": 0.08746181428432465,
      "mean_token_accuracy": 0.9926989078521729,
      "num_tokens": 19496314.0,
      "step": 121
    },
    {
      "aux_loss": 8.043519020080566,
      "entropy": 0.03671685978770256,
      "epoch": 2.9156626506024095,
      "grad_norm": 0.3954128921031952,
      "learning_rate": 1.4449658820570577e-06,
      "loss": 0.08653290569782257,
      "mean_token_accuracy": 0.9926026463508606,
      "num_tokens": 19661718.0,
      "step": 122
    },
    {
      "aux_loss": 8.045527458190918,
      "entropy": 0.04113632068037987,
      "epoch": 2.9397590361445785,
      "grad_norm": 0.39803871512413025,
      "learning_rate": 1.4448749839462395e-06,
      "loss": 0.08833590149879456,
      "mean_token_accuracy": 0.9923411309719086,
      "num_tokens": 19847373.0,
      "step": 123
    },
    {
      "aux_loss": 8.044186115264893,
      "entropy": 0.03642796538770199,
      "epoch": 2.963855421686747,
      "grad_norm": 0.3793511390686035,
      "learning_rate": 1.4447832783564146e-06,
      "loss": 0.086070716381073,
      "mean_token_accuracy": 0.9930998384952545,
      "num_tokens": 20025875.0,
      "step": 124
    },
    {
      "aux_loss": 8.046002388000488,
      "entropy": 0.047230470925569534,
      "epoch": 2.9879518072289155,
      "grad_norm": 0.4477235972881317,
      "learning_rate": 1.444690765494789e-06,
      "loss": 0.09117428958415985,
      "mean_token_accuracy": 0.9920151233673096,
      "num_tokens": 20182590.0,
      "step": 125
    },
    {
      "aux_loss": 8.042522430419922,
      "entropy": 0.040220800787210464,
      "epoch": 3.0,
      "grad_norm": 0.5846906900405884,
      "learning_rate": 1.4445974455703927e-06,
      "loss": 0.0573921799659729,
      "mean_token_accuracy": 0.9929695129394531,
      "num_tokens": 20224389.0,
      "step": 126
    },
    {
      "aux_loss": 8.044944763183594,
      "entropy": 0.039370276033878326,
      "epoch": 3.0240963855421685,
      "grad_norm": 0.3742491900920868,
      "learning_rate": 1.44450331879408e-06,
      "loss": 0.08556805551052094,
      "mean_token_accuracy": 0.9942886531352997,
      "num_tokens": 20380640.0,
      "step": 127
    },
    {
      "aux_loss": 8.043789863586426,
      "entropy": 0.045225948095321655,
      "epoch": 3.0481927710843375,
      "grad_norm": 0.38408586382865906,
      "learning_rate": 1.4444083853785277e-06,
      "loss": 0.08862867951393127,
      "mean_token_accuracy": 0.9931619465351105,
      "num_tokens": 20532981.0,
      "step": 128
    },
    {
      "aux_loss": 8.042508125305176,
      "entropy": 0.0450060423463583,
      "epoch": 3.072289156626506,
      "grad_norm": 0.37083542346954346,
      "learning_rate": 1.444312645538235e-06,
      "loss": 0.08864334225654602,
      "mean_token_accuracy": 0.9928875267505646,
      "num_tokens": 20683874.0,
      "step": 129
    },
    {
      "aux_loss": 8.041747093200684,
      "entropy": 0.04171221703290939,
      "epoch": 3.0963855421686746,
      "grad_norm": 0.36446937918663025,
      "learning_rate": 1.4442160994895236e-06,
      "loss": 0.08605008572340012,
      "mean_token_accuracy": 0.9937500953674316,
      "num_tokens": 20836727.0,
      "step": 130
    },
    {
      "aux_loss": 8.042478561401367,
      "entropy": 0.04033582843840122,
      "epoch": 3.1204819277108435,
      "grad_norm": 0.37039053440093994,
      "learning_rate": 1.4441187474505368e-06,
      "loss": 0.08619973808526993,
      "mean_token_accuracy": 0.9936582446098328,
      "num_tokens": 20989173.0,
      "step": 131
    },
    {
      "aux_loss": 8.04190444946289,
      "entropy": 0.04400304704904556,
      "epoch": 3.144578313253012,
      "grad_norm": 0.39441367983818054,
      "learning_rate": 1.4440205896412386e-06,
      "loss": 0.08861161768436432,
      "mean_token_accuracy": 0.993509829044342,
      "num_tokens": 21137746.0,
      "step": 132
    },
    {
      "aux_loss": 8.042577743530273,
      "entropy": 0.0362233929336071,
      "epoch": 3.1686746987951806,
      "grad_norm": 0.3513298034667969,
      "learning_rate": 1.4439216262834142e-06,
      "loss": 0.08395665884017944,
      "mean_token_accuracy": 0.9942896068096161,
      "num_tokens": 21303083.0,
      "step": 133
    },
    {
      "aux_loss": 8.045822143554688,
      "entropy": 0.044513555243611336,
      "epoch": 3.1927710843373496,
      "grad_norm": 0.4062028229236603,
      "learning_rate": 1.4438218576006684e-06,
      "loss": 0.08783506602048874,
      "mean_token_accuracy": 0.9932784140110016,
      "num_tokens": 21465720.0,
      "step": 134
    },
    {
      "aux_loss": 8.042749881744385,
      "entropy": 0.03609209880232811,
      "epoch": 3.216867469879518,
      "grad_norm": 0.36185288429260254,
      "learning_rate": 1.4437212838184255e-06,
      "loss": 0.08398795872926712,
      "mean_token_accuracy": 0.9942994117736816,
      "num_tokens": 21634832.0,
      "step": 135
    },
    {
      "aux_loss": 8.046396255493164,
      "entropy": 0.039127200841903687,
      "epoch": 3.2409638554216866,
      "grad_norm": 0.3764151334762573,
      "learning_rate": 1.44361990516393e-06,
      "loss": 0.08616587519645691,
      "mean_token_accuracy": 0.993415892124176,
      "num_tokens": 21801355.0,
      "step": 136
    },
    {
      "aux_loss": 8.043558597564697,
      "entropy": 0.03838344104588032,
      "epoch": 3.2650602409638556,
      "grad_norm": 0.3657795786857605,
      "learning_rate": 1.4435177218662435e-06,
      "loss": 0.08551836758852005,
      "mean_token_accuracy": 0.9938235878944397,
      "num_tokens": 21973015.0,
      "step": 137
    },
    {
      "aux_loss": 8.043213367462158,
      "entropy": 0.04633874632418156,
      "epoch": 3.289156626506024,
      "grad_norm": 0.38201189041137695,
      "learning_rate": 1.4434147341562468e-06,
      "loss": 0.08937641978263855,
      "mean_token_accuracy": 0.9929512739181519,
      "num_tokens": 22134914.0,
      "step": 138
    },
    {
      "aux_loss": 8.04480266571045,
      "entropy": 0.03689238056540489,
      "epoch": 3.3132530120481927,
      "grad_norm": 0.35625335574150085,
      "learning_rate": 1.4433109422666374e-06,
      "loss": 0.08385109901428223,
      "mean_token_accuracy": 0.9944583773612976,
      "num_tokens": 22302240.0,
      "step": 139
    },
    {
      "aux_loss": 8.042608261108398,
      "entropy": 0.04020004719495773,
      "epoch": 3.337349397590361,
      "grad_norm": 0.3919045925140381,
      "learning_rate": 1.4432063464319306e-06,
      "loss": 0.0874231606721878,
      "mean_token_accuracy": 0.9931889474391937,
      "num_tokens": 22471698.0,
      "step": 140
    },
    {
      "aux_loss": 8.042847156524658,
      "entropy": 0.03685566782951355,
      "epoch": 3.36144578313253,
      "grad_norm": 0.3515630066394806,
      "learning_rate": 1.4431009468884578e-06,
      "loss": 0.0849953442811966,
      "mean_token_accuracy": 0.9941838085651398,
      "num_tokens": 22653073.0,
      "step": 141
    },
    {
      "aux_loss": 8.043398380279541,
      "entropy": 0.0423385389149189,
      "epoch": 3.3855421686746987,
      "grad_norm": 0.41736137866973877,
      "learning_rate": 1.442994743874366e-06,
      "loss": 0.08791652321815491,
      "mean_token_accuracy": 0.9927142858505249,
      "num_tokens": 22810578.0,
      "step": 142
    },
    {
      "aux_loss": 8.04459524154663,
      "entropy": 0.04449957422912121,
      "epoch": 3.4096385542168672,
      "grad_norm": 0.4232349693775177,
      "learning_rate": 1.4428877376296185e-06,
      "loss": 0.08924206346273422,
      "mean_token_accuracy": 0.9924839735031128,
      "num_tokens": 22953794.0,
      "step": 143
    },
    {
      "aux_loss": 8.041166305541992,
      "entropy": 0.04270816035568714,
      "epoch": 3.433734939759036,
      "grad_norm": 0.3914850354194641,
      "learning_rate": 1.4427799283959926e-06,
      "loss": 0.08785770833492279,
      "mean_token_accuracy": 0.9931382238864899,
      "num_tokens": 23120357.0,
      "step": 144
    },
    {
      "aux_loss": 8.04544448852539,
      "entropy": 0.03356307931244373,
      "epoch": 3.4578313253012047,
      "grad_norm": 0.3500516712665558,
      "learning_rate": 1.4426713164170808e-06,
      "loss": 0.08334299921989441,
      "mean_token_accuracy": 0.9946193993091583,
      "num_tokens": 23300231.0,
      "step": 145
    },
    {
      "aux_loss": 8.04530382156372,
      "entropy": 0.044576091691851616,
      "epoch": 3.4819277108433733,
      "grad_norm": 0.41992995142936707,
      "learning_rate": 1.4425619019382884e-06,
      "loss": 0.08873672783374786,
      "mean_token_accuracy": 0.9929416477680206,
      "num_tokens": 23453268.0,
      "step": 146
    },
    {
      "aux_loss": 8.044836044311523,
      "entropy": 0.03937925212085247,
      "epoch": 3.5060240963855422,
      "grad_norm": 0.37806037068367004,
      "learning_rate": 1.4424516852068346e-06,
      "loss": 0.08648265898227692,
      "mean_token_accuracy": 0.9935058355331421,
      "num_tokens": 23625257.0,
      "step": 147
    },
    {
      "aux_loss": 8.04463529586792,
      "entropy": 0.0399188157171011,
      "epoch": 3.5301204819277108,
      "grad_norm": 0.37204214930534363,
      "learning_rate": 1.442340666471751e-06,
      "loss": 0.08639834821224213,
      "mean_token_accuracy": 0.9933888912200928,
      "num_tokens": 23820327.0,
      "step": 148
    },
    {
      "aux_loss": 8.044984340667725,
      "entropy": 0.036087172105908394,
      "epoch": 3.5542168674698793,
      "grad_norm": 0.36808425188064575,
      "learning_rate": 1.4422288459838816e-06,
      "loss": 0.08478154242038727,
      "mean_token_accuracy": 0.9945113360881805,
      "num_tokens": 23979995.0,
      "step": 149
    },
    {
      "aux_loss": 8.043323993682861,
      "entropy": 0.04419667460024357,
      "epoch": 3.5783132530120483,
      "grad_norm": 0.41120970249176025,
      "learning_rate": 1.4421162239958818e-06,
      "loss": 0.0894618034362793,
      "mean_token_accuracy": 0.9925802052021027,
      "num_tokens": 24125572.0,
      "step": 150
    },
    {
      "aux_loss": 8.042481422424316,
      "entropy": 0.03813181072473526,
      "epoch": 3.602409638554217,
      "grad_norm": 0.3920314610004425,
      "learning_rate": 1.4420028007622175e-06,
      "loss": 0.08617769181728363,
      "mean_token_accuracy": 0.9939529597759247,
      "num_tokens": 24283705.0,
      "step": 151
    },
    {
      "aux_loss": 8.043877601623535,
      "entropy": 0.03504548221826553,
      "epoch": 3.6265060240963853,
      "grad_norm": 0.38018906116485596,
      "learning_rate": 1.4418885765391659e-06,
      "loss": 0.08467784523963928,
      "mean_token_accuracy": 0.993707150220871,
      "num_tokens": 24472505.0,
      "step": 152
    },
    {
      "aux_loss": 8.042401313781738,
      "entropy": 0.04122787527740002,
      "epoch": 3.6506024096385543,
      "grad_norm": 0.39765456318855286,
      "learning_rate": 1.4417735515848131e-06,
      "loss": 0.08704474568367004,
      "mean_token_accuracy": 0.9933643937110901,
      "num_tokens": 24643360.0,
      "step": 153
    },
    {
      "aux_loss": 8.043309688568115,
      "entropy": 0.04166724532842636,
      "epoch": 3.674698795180723,
      "grad_norm": 0.3905385136604309,
      "learning_rate": 1.4416577261590549e-06,
      "loss": 0.08706511557102203,
      "mean_token_accuracy": 0.9933055341243744,
      "num_tokens": 24802344.0,
      "step": 154
    },
    {
      "aux_loss": 8.040821552276611,
      "entropy": 0.03833147883415222,
      "epoch": 3.6987951807228914,
      "grad_norm": 0.4488506615161896,
      "learning_rate": 1.4415411005235958e-06,
      "loss": 0.08571713417768478,
      "mean_token_accuracy": 0.9934513568878174,
      "num_tokens": 24956387.0,
      "step": 155
    },
    {
      "aux_loss": 8.043232440948486,
      "entropy": 0.03610091283917427,
      "epoch": 3.7228915662650603,
      "grad_norm": 0.3729538023471832,
      "learning_rate": 1.4414236749419483e-06,
      "loss": 0.08453528583049774,
      "mean_token_accuracy": 0.9941688776016235,
      "num_tokens": 25118460.0,
      "step": 156
    },
    {
      "aux_loss": 8.042871475219727,
      "entropy": 0.04386945813894272,
      "epoch": 3.746987951807229,
      "grad_norm": 0.4113171100616455,
      "learning_rate": 1.441305449679432e-06,
      "loss": 0.08860649913549423,
      "mean_token_accuracy": 0.9927978813648224,
      "num_tokens": 25276453.0,
      "step": 157
    },
    {
      "aux_loss": 8.04326343536377,
      "entropy": 0.03775089047849178,
      "epoch": 3.7710843373493974,
      "grad_norm": 0.3969726860523224,
      "learning_rate": 1.441186425003174e-06,
      "loss": 0.08585880696773529,
      "mean_token_accuracy": 0.9934997260570526,
      "num_tokens": 25465189.0,
      "step": 158
    },
    {
      "aux_loss": 8.041276931762695,
      "entropy": 0.03554354049265385,
      "epoch": 3.7951807228915664,
      "grad_norm": 0.37212640047073364,
      "learning_rate": 1.4410666011821071e-06,
      "loss": 0.08398972451686859,
      "mean_token_accuracy": 0.9942491948604584,
      "num_tokens": 25647091.0,
      "step": 159
    },
    {
      "aux_loss": 8.04280948638916,
      "entropy": 0.03362584114074707,
      "epoch": 3.819277108433735,
      "grad_norm": 0.37804487347602844,
      "learning_rate": 1.44094597848697e-06,
      "loss": 0.08355149626731873,
      "mean_token_accuracy": 0.9944404661655426,
      "num_tokens": 25812959.0,
      "step": 160
    },
    {
      "aux_loss": 8.041194915771484,
      "entropy": 0.04549684002995491,
      "epoch": 3.8433734939759034,
      "grad_norm": 0.4044184982776642,
      "learning_rate": 1.4408245571903066e-06,
      "loss": 0.08932945877313614,
      "mean_token_accuracy": 0.9926709532737732,
      "num_tokens": 25975077.0,
      "step": 161
    },
    {
      "aux_loss": 8.040953636169434,
      "entropy": 0.04158872179687023,
      "epoch": 3.8674698795180724,
      "grad_norm": 0.4143630564212799,
      "learning_rate": 1.4407023375664644e-06,
      "loss": 0.08703690767288208,
      "mean_token_accuracy": 0.993314117193222,
      "num_tokens": 26133481.0,
      "step": 162
    },
    {
      "aux_loss": 8.042658805847168,
      "entropy": 0.04513395018875599,
      "epoch": 3.891566265060241,
      "grad_norm": 0.4163261651992798,
      "learning_rate": 1.440579319891596e-06,
      "loss": 0.08887109160423279,
      "mean_token_accuracy": 0.9932331144809723,
      "num_tokens": 26285677.0,
      "step": 163
    },
    {
      "aux_loss": 8.044297218322754,
      "entropy": 0.03768927603960037,
      "epoch": 3.9156626506024095,
      "grad_norm": 0.3762443959712982,
      "learning_rate": 1.440455504443656e-06,
      "loss": 0.08547879010438919,
      "mean_token_accuracy": 0.9941076934337616,
      "num_tokens": 26456101.0,
      "step": 164
    },
    {
      "aux_loss": 8.040573596954346,
      "entropy": 0.040223028510808945,
      "epoch": 3.9397590361445785,
      "grad_norm": 0.44806984066963196,
      "learning_rate": 1.4403308915024023e-06,
      "loss": 0.08660810440778732,
      "mean_token_accuracy": 0.9934520721435547,
      "num_tokens": 26599076.0,
      "step": 165
    },
    {
      "aux_loss": 8.042471408843994,
      "entropy": 0.038684969767928123,
      "epoch": 3.963855421686747,
      "grad_norm": 0.40352219343185425,
      "learning_rate": 1.4402054813493942e-06,
      "loss": 0.08614273369312286,
      "mean_token_accuracy": 0.9937964081764221,
      "num_tokens": 26769500.0,
      "step": 166
    },
    {
      "aux_loss": 8.041973114013672,
      "entropy": 0.037965722382068634,
      "epoch": 3.9879518072289155,
      "grad_norm": 0.38133445382118225,
      "learning_rate": 1.4400792742679925e-06,
      "loss": 0.08504779636859894,
      "mean_token_accuracy": 0.9941068291664124,
      "num_tokens": 26942979.0,
      "step": 167
    },
    {
      "aux_loss": 8.039653778076172,
      "entropy": 0.046194545924663544,
      "epoch": 4.0,
      "grad_norm": 0.5609819889068604,
      "learning_rate": 1.439952270543359e-06,
      "loss": 0.053759872913360596,
      "mean_token_accuracy": 0.9936670064926147,
      "num_tokens": 26965852.0,
      "step": 168
    },
    {
      "aux_loss": 8.043168544769287,
      "entropy": 0.03969798795878887,
      "epoch": 4.024096385542169,
      "grad_norm": 0.5476467609405518,
      "learning_rate": 1.4398244704624548e-06,
      "loss": 0.08540086448192596,
      "mean_token_accuracy": 0.9945910573005676,
      "num_tokens": 27138361.0,
      "step": 169
    },
    {
      "aux_loss": 8.041584968566895,
      "entropy": 0.037986068055033684,
      "epoch": 4.048192771084337,
      "grad_norm": 0.3398607075214386,
      "learning_rate": 1.4396958743140407e-06,
      "loss": 0.08402132987976074,
      "mean_token_accuracy": 0.9951412975788116,
      "num_tokens": 27298070.0,
      "step": 170
    },
    {
      "aux_loss": 8.041838645935059,
      "entropy": 0.03689403831958771,
      "epoch": 4.072289156626506,
      "grad_norm": 0.3282793462276459,
      "learning_rate": 1.4395664823886764e-06,
      "loss": 0.08331362903118134,
      "mean_token_accuracy": 0.9951364398002625,
      "num_tokens": 27477493.0,
      "step": 171
    },
    {
      "aux_loss": 8.042462825775146,
      "entropy": 0.04227207601070404,
      "epoch": 4.096385542168675,
      "grad_norm": 0.3614228367805481,
      "learning_rate": 1.4394362949787193e-06,
      "loss": 0.08635450899600983,
      "mean_token_accuracy": 0.9946377873420715,
      "num_tokens": 27650870.0,
      "step": 172
    },
    {
      "aux_loss": 8.040101051330566,
      "entropy": 0.039225177839398384,
      "epoch": 4.120481927710843,
      "grad_norm": 0.3621145486831665,
      "learning_rate": 1.4393053123783242e-06,
      "loss": 0.08554555475711823,
      "mean_token_accuracy": 0.9942145347595215,
      "num_tokens": 27809889.0,
      "step": 173
    },
    {
      "aux_loss": 8.042874336242676,
      "entropy": 0.04224075563251972,
      "epoch": 4.144578313253012,
      "grad_norm": 0.3778877854347229,
      "learning_rate": 1.439173534883443e-06,
      "loss": 0.08607605844736099,
      "mean_token_accuracy": 0.9945782423019409,
      "num_tokens": 27957420.0,
      "step": 174
    },
    {
      "aux_loss": 8.040241718292236,
      "entropy": 0.03377825953066349,
      "epoch": 4.168674698795181,
      "grad_norm": 0.34376251697540283,
      "learning_rate": 1.439040962791823e-06,
      "loss": 0.08279868215322495,
      "mean_token_accuracy": 0.9951018989086151,
      "num_tokens": 28127629.0,
      "step": 175
    },
    {
      "aux_loss": 8.042171955108643,
      "entropy": 0.03661440312862396,
      "epoch": 4.192771084337349,
      "grad_norm": 0.3872918486595154,
      "learning_rate": 1.4389075964030074e-06,
      "loss": 0.08358034491539001,
      "mean_token_accuracy": 0.9953621327877045,
      "num_tokens": 28293076.0,
      "step": 176
    },
    {
      "aux_loss": 8.040359020233154,
      "entropy": 0.035597002133727074,
      "epoch": 4.216867469879518,
      "grad_norm": 0.357154905796051,
      "learning_rate": 1.438773436018334e-06,
      "loss": 0.0828801617026329,
      "mean_token_accuracy": 0.9952186346054077,
      "num_tokens": 28453952.0,
      "step": 177
    },
    {
      "aux_loss": 8.042491912841797,
      "entropy": 0.04073164612054825,
      "epoch": 4.240963855421687,
      "grad_norm": 0.3982166349887848,
      "learning_rate": 1.438638481940934e-06,
      "loss": 0.08584737777709961,
      "mean_token_accuracy": 0.9942737519741058,
      "num_tokens": 28604320.0,
      "step": 178
    },
    {
      "aux_loss": 8.040094375610352,
      "entropy": 0.03856072202324867,
      "epoch": 4.265060240963855,
      "grad_norm": 0.3902375102043152,
      "learning_rate": 1.4385027344757331e-06,
      "loss": 0.08460918813943863,
      "mean_token_accuracy": 0.994719922542572,
      "num_tokens": 28756892.0,
      "step": 179
    },
    {
      "aux_loss": 8.040781497955322,
      "entropy": 0.035547565668821335,
      "epoch": 4.289156626506024,
      "grad_norm": 0.3302716016769409,
      "learning_rate": 1.438366193929449e-06,
      "loss": 0.0830703154206276,
      "mean_token_accuracy": 0.995084673166275,
      "num_tokens": 28930439.0,
      "step": 180
    },
    {
      "aux_loss": 8.039629459381104,
      "entropy": 0.04048551991581917,
      "epoch": 4.313253012048193,
      "grad_norm": 0.4318244457244873,
      "learning_rate": 1.4382288606105908e-06,
      "loss": 0.08579900115728378,
      "mean_token_accuracy": 0.9941352307796478,
      "num_tokens": 29092853.0,
      "step": 181
    },
    {
      "aux_loss": 8.040911674499512,
      "entropy": 0.03764263913035393,
      "epoch": 4.337349397590361,
      "grad_norm": 0.42847853899002075,
      "learning_rate": 1.4380907348294597e-06,
      "loss": 0.0839407742023468,
      "mean_token_accuracy": 0.9944304823875427,
      "num_tokens": 29226583.0,
      "step": 182
    },
    {
      "aux_loss": 8.039612770080566,
      "entropy": 0.03596244379878044,
      "epoch": 4.36144578313253,
      "grad_norm": 0.36768803000450134,
      "learning_rate": 1.4379518168981469e-06,
      "loss": 0.08371734619140625,
      "mean_token_accuracy": 0.9946404695510864,
      "num_tokens": 29381987.0,
      "step": 183
    },
    {
      "aux_loss": 8.040324211120605,
      "entropy": 0.038307564333081245,
      "epoch": 4.385542168674699,
      "grad_norm": 0.38769692182540894,
      "learning_rate": 1.437812107130534e-06,
      "loss": 0.08458788692951202,
      "mean_token_accuracy": 0.9944848418235779,
      "num_tokens": 29540073.0,
      "step": 184
    },
    {
      "aux_loss": 8.040425777435303,
      "entropy": 0.03475116938352585,
      "epoch": 4.409638554216867,
      "grad_norm": 0.3598844110965729,
      "learning_rate": 1.437671605842291e-06,
      "loss": 0.083564393222332,
      "mean_token_accuracy": 0.9947148263454437,
      "num_tokens": 29701935.0,
      "step": 185
    },
    {
      "aux_loss": 8.041619300842285,
      "entropy": 0.037232447415590286,
      "epoch": 4.433734939759036,
      "grad_norm": 0.4138000011444092,
      "learning_rate": 1.4375303133508774e-06,
      "loss": 0.08460228890180588,
      "mean_token_accuracy": 0.9941424429416656,
      "num_tokens": 29843884.0,
      "step": 186
    },
    {
      "aux_loss": 8.04134750366211,
      "entropy": 0.036806127056479454,
      "epoch": 4.457831325301205,
      "grad_norm": 0.4106656610965729,
      "learning_rate": 1.437388229975539e-06,
      "loss": 0.08441147208213806,
      "mean_token_accuracy": 0.9946633577346802,
      "num_tokens": 29987812.0,
      "step": 187
    },
    {
      "aux_loss": 8.039554119110107,
      "entropy": 0.03547658585011959,
      "epoch": 4.481927710843373,
      "grad_norm": 0.3592974841594696,
      "learning_rate": 1.4372453560373097e-06,
      "loss": 0.0837140679359436,
      "mean_token_accuracy": 0.9946017563343048,
      "num_tokens": 30149270.0,
      "step": 188
    },
    {
      "aux_loss": 8.041821479797363,
      "entropy": 0.035800814628601074,
      "epoch": 4.506024096385542,
      "grad_norm": 0.36003249883651733,
      "learning_rate": 1.437101691859009e-06,
      "loss": 0.08376017212867737,
      "mean_token_accuracy": 0.9945588111877441,
      "num_tokens": 30307956.0,
      "step": 189
    },
    {
      "aux_loss": 8.042970180511475,
      "entropy": 0.040934668853878975,
      "epoch": 4.530120481927711,
      "grad_norm": 0.42162829637527466,
      "learning_rate": 1.4369572377652423e-06,
      "loss": 0.08601701259613037,
      "mean_token_accuracy": 0.9943284690380096,
      "num_tokens": 30494893.0,
      "step": 190
    },
    {
      "aux_loss": 8.041118144989014,
      "entropy": 0.03687039390206337,
      "epoch": 4.554216867469879,
      "grad_norm": 0.370378315448761,
      "learning_rate": 1.4368119940823998e-06,
      "loss": 0.08406563103199005,
      "mean_token_accuracy": 0.9943959414958954,
      "num_tokens": 30672420.0,
      "step": 191
    },
    {
      "aux_loss": 8.040582656860352,
      "entropy": 0.03540595434606075,
      "epoch": 4.578313253012048,
      "grad_norm": 0.38017594814300537,
      "learning_rate": 1.4366659611386557e-06,
      "loss": 0.08305495977401733,
      "mean_token_accuracy": 0.9947693049907684,
      "num_tokens": 30847054.0,
      "step": 192
    },
    {
      "aux_loss": 8.042843341827393,
      "entropy": 0.04187178239226341,
      "epoch": 4.602409638554217,
      "grad_norm": 0.3843596577644348,
      "learning_rate": 1.436519139263967e-06,
      "loss": 0.08659607917070389,
      "mean_token_accuracy": 0.9939823746681213,
      "num_tokens": 31000786.0,
      "step": 193
    },
    {
      "aux_loss": 8.041165351867676,
      "entropy": 0.034371569752693176,
      "epoch": 4.626506024096385,
      "grad_norm": 0.37609967589378357,
      "learning_rate": 1.4363715287900742e-06,
      "loss": 0.08291493356227875,
      "mean_token_accuracy": 0.9949217736721039,
      "num_tokens": 31180147.0,
      "step": 194
    },
    {
      "aux_loss": 8.040238380432129,
      "entropy": 0.03516504354774952,
      "epoch": 4.650602409638554,
      "grad_norm": 0.3637091815471649,
      "learning_rate": 1.4362231300504988e-06,
      "loss": 0.08338074386119843,
      "mean_token_accuracy": 0.9948630630970001,
      "num_tokens": 31343986.0,
      "step": 195
    },
    {
      "aux_loss": 8.041361808776855,
      "entropy": 0.04160070978105068,
      "epoch": 4.674698795180722,
      "grad_norm": 0.4149928390979767,
      "learning_rate": 1.4360739433805439e-06,
      "loss": 0.08658482134342194,
      "mean_token_accuracy": 0.9941949248313904,
      "num_tokens": 31512880.0,
      "step": 196
    },
    {
      "aux_loss": 8.040586948394775,
      "entropy": 0.04432631470263004,
      "epoch": 4.698795180722891,
      "grad_norm": 0.40140870213508606,
      "learning_rate": 1.4359239691172926e-06,
      "loss": 0.08757840096950531,
      "mean_token_accuracy": 0.9936575889587402,
      "num_tokens": 31661985.0,
      "step": 197
    },
    {
      "aux_loss": 8.040112018585205,
      "entropy": 0.037308650091290474,
      "epoch": 4.72289156626506,
      "grad_norm": 0.39514875411987305,
      "learning_rate": 1.4357732075996076e-06,
      "loss": 0.08466262370347977,
      "mean_token_accuracy": 0.9945145845413208,
      "num_tokens": 31842845.0,
      "step": 198
    },
    {
      "aux_loss": 8.041626930236816,
      "entropy": 0.04097534902393818,
      "epoch": 4.746987951807229,
      "grad_norm": 0.405231237411499,
      "learning_rate": 1.4356216591681305e-06,
      "loss": 0.08651918172836304,
      "mean_token_accuracy": 0.9938961267471313,
      "num_tokens": 31994483.0,
      "step": 199
    },
    {
      "aux_loss": 8.038835525512695,
      "entropy": 0.03753245994448662,
      "epoch": 4.771084337349397,
      "grad_norm": 0.3914196789264679,
      "learning_rate": 1.4354693241652809e-06,
      "loss": 0.08507226407527924,
      "mean_token_accuracy": 0.9937683045864105,
      "num_tokens": 32153186.0,
      "step": 200
    },
    {
      "aux_loss": 8.039247989654541,
      "entropy": 0.037701187655329704,
      "epoch": 4.795180722891566,
      "grad_norm": 0.3929915130138397,
      "learning_rate": 1.4353162029352554e-06,
      "loss": 0.0845346599817276,
      "mean_token_accuracy": 0.9943271279335022,
      "num_tokens": 32309805.0,
      "step": 201
    },
    {
      "aux_loss": 8.04191780090332,
      "entropy": 0.04091975465416908,
      "epoch": 4.8192771084337345,
      "grad_norm": 0.4561111032962799,
      "learning_rate": 1.4351622958240273e-06,
      "loss": 0.08624639362096786,
      "mean_token_accuracy": 0.9937124252319336,
      "num_tokens": 32483341.0,
      "step": 202
    },
    {
      "aux_loss": 8.04048204421997,
      "entropy": 0.034330458380281925,
      "epoch": 4.843373493975903,
      "grad_norm": 0.367446631193161,
      "learning_rate": 1.4350076031793456e-06,
      "loss": 0.08318167924880981,
      "mean_token_accuracy": 0.9946629703044891,
      "num_tokens": 32647509.0,
      "step": 203
    },
    {
      "aux_loss": 8.04125452041626,
      "entropy": 0.03692132607102394,
      "epoch": 4.867469879518072,
      "grad_norm": 0.403191477060318,
      "learning_rate": 1.4348521253507344e-06,
      "loss": 0.08472205698490143,
      "mean_token_accuracy": 0.9941362142562866,
      "num_tokens": 32839256.0,
      "step": 204
    },
    {
      "aux_loss": 8.039531230926514,
      "entropy": 0.038704318925738335,
      "epoch": 4.891566265060241,
      "grad_norm": 0.39772483706474304,
      "learning_rate": 1.434695862689491e-06,
      "loss": 0.08526815474033356,
      "mean_token_accuracy": 0.9942019283771515,
      "num_tokens": 33003150.0,
      "step": 205
    },
    {
      "aux_loss": 8.040374755859375,
      "entropy": 0.03595581278204918,
      "epoch": 4.9156626506024095,
      "grad_norm": 0.4022948741912842,
      "learning_rate": 1.4345388155486866e-06,
      "loss": 0.08391736447811127,
      "mean_token_accuracy": 0.9942536056041718,
      "num_tokens": 33179546.0,
      "step": 206
    },
    {
      "aux_loss": 8.039075374603271,
      "entropy": 0.03800219111144543,
      "epoch": 4.9397590361445785,
      "grad_norm": 0.380649209022522,
      "learning_rate": 1.4343809842831657e-06,
      "loss": 0.08527541905641556,
      "mean_token_accuracy": 0.9943437278270721,
      "num_tokens": 33351836.0,
      "step": 207
    },
    {
      "aux_loss": 8.043510913848877,
      "entropy": 0.04215468093752861,
      "epoch": 4.9638554216867465,
      "grad_norm": 0.41969671845436096,
      "learning_rate": 1.4342223692495428e-06,
      "loss": 0.08674986660480499,
      "mean_token_accuracy": 0.9934200346469879,
      "num_tokens": 33505025.0,
      "step": 208
    },
    {
      "aux_loss": 8.039401054382324,
      "entropy": 0.03880974091589451,
      "epoch": 4.9879518072289155,
      "grad_norm": 0.37678730487823486,
      "learning_rate": 1.4340629708062048e-06,
      "loss": 0.08524405211210251,
      "mean_token_accuracy": 0.9945527911186218,
      "num_tokens": 33660485.0,
      "step": 209
    },
    {
      "aux_loss": 8.04077434539795,
      "entropy": 0.03518826887011528,
      "epoch": 5.0,
      "grad_norm": 0.48134711384773254,
      "learning_rate": 1.433902789313308e-06,
      "loss": 0.050809722393751144,
      "mean_token_accuracy": 0.9939299821853638,
      "num_tokens": 33707315.0,
      "step": 210
    },
    {
      "aux_loss": 8.039217472076416,
      "entropy": 0.03736879117786884,
      "epoch": 5.024096385542169,
      "grad_norm": 0.3484879732131958,
      "learning_rate": 1.433741825132778e-06,
      "loss": 0.08324334025382996,
      "mean_token_accuracy": 0.9955032467842102,
      "num_tokens": 33863394.0,
      "step": 211
    },
    {
      "aux_loss": 8.038911819458008,
      "entropy": 0.03610601648688316,
      "epoch": 5.048192771084337,
      "grad_norm": 0.3639015555381775,
      "learning_rate": 1.4335800786283092e-06,
      "loss": 0.08297591656446457,
      "mean_token_accuracy": 0.9954032003879547,
      "num_tokens": 34014983.0,
      "step": 212
    },
    {
      "aux_loss": 8.038973808288574,
      "entropy": 0.0321112684905529,
      "epoch": 5.072289156626506,
      "grad_norm": 0.3035561144351959,
      "learning_rate": 1.433417550165363e-06,
      "loss": 0.080979123711586,
      "mean_token_accuracy": 0.9960999488830566,
      "num_tokens": 34202386.0,
      "step": 213
    },
    {
      "aux_loss": 8.037017822265625,
      "entropy": 0.03347293473780155,
      "epoch": 5.096385542168675,
      "grad_norm": 0.32014161348342896,
      "learning_rate": 1.4332542401111685e-06,
      "loss": 0.0811787024140358,
      "mean_token_accuracy": 0.9961134791374207,
      "num_tokens": 34369320.0,
      "step": 214
    },
    {
      "aux_loss": 8.04166555404663,
      "entropy": 0.03581018000841141,
      "epoch": 5.120481927710843,
      "grad_norm": 0.3393751084804535,
      "learning_rate": 1.43309014883472e-06,
      "loss": 0.0826868787407875,
      "mean_token_accuracy": 0.9955731332302094,
      "num_tokens": 34545517.0,
      "step": 215
    },
    {
      "aux_loss": 8.0395827293396,
      "entropy": 0.03482176177203655,
      "epoch": 5.144578313253012,
      "grad_norm": 0.37512803077697754,
      "learning_rate": 1.4329252767067773e-06,
      "loss": 0.08189767599105835,
      "mean_token_accuracy": 0.995861828327179,
      "num_tokens": 34730633.0,
      "step": 216
    },
    {
      "aux_loss": 8.040324211120605,
      "entropy": 0.03882727213203907,
      "epoch": 5.168674698795181,
      "grad_norm": 0.34391677379608154,
      "learning_rate": 1.4327596240998645e-06,
      "loss": 0.08408463001251221,
      "mean_token_accuracy": 0.9951691627502441,
      "num_tokens": 34897938.0,
      "step": 217
    },
    {
      "aux_loss": 8.038958549499512,
      "entropy": 0.03700985945761204,
      "epoch": 5.192771084337349,
      "grad_norm": 0.3792520761489868,
      "learning_rate": 1.4325931913882691e-06,
      "loss": 0.08347566425800323,
      "mean_token_accuracy": 0.9954339861869812,
      "num_tokens": 35042500.0,
      "step": 218
    },
    {
      "aux_loss": 8.039316177368164,
      "entropy": 0.04214794747531414,
      "epoch": 5.216867469879518,
      "grad_norm": 0.38034358620643616,
      "learning_rate": 1.4324259789480415e-06,
      "loss": 0.08621414750814438,
      "mean_token_accuracy": 0.9948817193508148,
      "num_tokens": 35199253.0,
      "step": 219
    },
    {
      "aux_loss": 8.040565967559814,
      "entropy": 0.03364016301929951,
      "epoch": 5.240963855421687,
      "grad_norm": 0.33803656697273254,
      "learning_rate": 1.4322579871569933e-06,
      "loss": 0.08156892657279968,
      "mean_token_accuracy": 0.9959030747413635,
      "num_tokens": 35368834.0,
      "step": 220
    },
    {
      "aux_loss": 8.038960933685303,
      "entropy": 0.033280991949141026,
      "epoch": 5.265060240963855,
      "grad_norm": 0.3262208104133606,
      "learning_rate": 1.4320892163946975e-06,
      "loss": 0.08120635151863098,
      "mean_token_accuracy": 0.9958637356758118,
      "num_tokens": 35535515.0,
      "step": 221
    },
    {
      "aux_loss": 8.038169860839844,
      "entropy": 0.031326593831181526,
      "epoch": 5.289156626506024,
      "grad_norm": 0.342752069234848,
      "learning_rate": 1.431919667042487e-06,
      "loss": 0.07946254312992096,
      "mean_token_accuracy": 0.9960636794567108,
      "num_tokens": 35700534.0,
      "step": 222
    },
    {
      "aux_loss": 8.039755821228027,
      "entropy": 0.034905584529042244,
      "epoch": 5.313253012048193,
      "grad_norm": 0.3462710678577423,
      "learning_rate": 1.431749339483454e-06,
      "loss": 0.08242812752723694,
      "mean_token_accuracy": 0.9953078627586365,
      "num_tokens": 35880129.0,
      "step": 223
    },
    {
      "aux_loss": 8.03895616531372,
      "entropy": 0.037569453939795494,
      "epoch": 5.337349397590361,
      "grad_norm": 0.3580102026462555,
      "learning_rate": 1.4315782341024486e-06,
      "loss": 0.08401724696159363,
      "mean_token_accuracy": 0.9949289560317993,
      "num_tokens": 36049636.0,
      "step": 224
    },
    {
      "aux_loss": 8.035909175872803,
      "entropy": 0.03622467443346977,
      "epoch": 5.36144578313253,
      "grad_norm": 0.3741055727005005,
      "learning_rate": 1.4314063512860793e-06,
      "loss": 0.08327938616275787,
      "mean_token_accuracy": 0.994949609041214,
      "num_tokens": 36198666.0,
      "step": 225
    },
    {
      "aux_loss": 8.03954553604126,
      "entropy": 0.03737705014646053,
      "epoch": 5.385542168674699,
      "grad_norm": 0.3866511583328247,
      "learning_rate": 1.4312336914227104e-06,
      "loss": 0.0833113044500351,
      "mean_token_accuracy": 0.9953556656837463,
      "num_tokens": 36338753.0,
      "step": 226
    },
    {
      "aux_loss": 8.040770053863525,
      "entropy": 0.03640630468726158,
      "epoch": 5.409638554216867,
      "grad_norm": 0.37529176473617554,
      "learning_rate": 1.4310602549024621e-06,
      "loss": 0.0830492153763771,
      "mean_token_accuracy": 0.9955044686794281,
      "num_tokens": 36498867.0,
      "step": 227
    },
    {
      "aux_loss": 8.038099765777588,
      "entropy": 0.03481042757630348,
      "epoch": 5.433734939759036,
      "grad_norm": 0.3607235252857208,
      "learning_rate": 1.4308860421172094e-06,
      "loss": 0.08244866877794266,
      "mean_token_accuracy": 0.995325118303299,
      "num_tokens": 36682484.0,
      "step": 228
    },
    {
      "aux_loss": 8.03747844696045,
      "entropy": 0.036855777725577354,
      "epoch": 5.457831325301205,
      "grad_norm": 0.5099928379058838,
      "learning_rate": 1.4307110534605815e-06,
      "loss": 0.08275534957647324,
      "mean_token_accuracy": 0.9952235817909241,
      "num_tokens": 36839247.0,
      "step": 229
    },
    {
      "aux_loss": 8.039116859436035,
      "entropy": 0.044368937611579895,
      "epoch": 5.481927710843373,
      "grad_norm": 0.39856716990470886,
      "learning_rate": 1.4305352893279603e-06,
      "loss": 0.08694414794445038,
      "mean_token_accuracy": 0.9942440986633301,
      "num_tokens": 36984989.0,
      "step": 230
    },
    {
      "aux_loss": 8.04021692276001,
      "entropy": 0.03966960869729519,
      "epoch": 5.506024096385542,
      "grad_norm": 0.3691135048866272,
      "learning_rate": 1.4303587501164801e-06,
      "loss": 0.08465172350406647,
      "mean_token_accuracy": 0.9948685467243195,
      "num_tokens": 37136734.0,
      "step": 231
    },
    {
      "aux_loss": 8.038138389587402,
      "entropy": 0.03791094198822975,
      "epoch": 5.530120481927711,
      "grad_norm": 0.37443944811820984,
      "learning_rate": 1.4301814362250267e-06,
      "loss": 0.08411930501461029,
      "mean_token_accuracy": 0.9945204854011536,
      "num_tokens": 37306594.0,
      "step": 232
    },
    {
      "aux_loss": 8.04017972946167,
      "entropy": 0.03802665323019028,
      "epoch": 5.554216867469879,
      "grad_norm": 0.40194767713546753,
      "learning_rate": 1.4300033480542355e-06,
      "loss": 0.08439681679010391,
      "mean_token_accuracy": 0.9948781728744507,
      "num_tokens": 37482014.0,
      "step": 233
    },
    {
      "aux_loss": 8.037278175354004,
      "entropy": 0.0383224431425333,
      "epoch": 5.578313253012048,
      "grad_norm": 0.4007725715637207,
      "learning_rate": 1.4298244860064922e-06,
      "loss": 0.08417156338691711,
      "mean_token_accuracy": 0.9947126507759094,
      "num_tokens": 37655390.0,
      "step": 234
    },
    {
      "aux_loss": 8.038545608520508,
      "entropy": 0.03560042753815651,
      "epoch": 5.602409638554217,
      "grad_norm": 0.4014568030834198,
      "learning_rate": 1.4296448504859306e-06,
      "loss": 0.08304604887962341,
      "mean_token_accuracy": 0.9949043989181519,
      "num_tokens": 37807861.0,
      "step": 235
    },
    {
      "aux_loss": 8.039902210235596,
      "entropy": 0.03399462066590786,
      "epoch": 5.626506024096385,
      "grad_norm": 0.353241890668869,
      "learning_rate": 1.4294644418984323e-06,
      "loss": 0.08209338784217834,
      "mean_token_accuracy": 0.9952429831027985,
      "num_tokens": 37993700.0,
      "step": 236
    },
    {
      "aux_loss": 8.039174556732178,
      "entropy": 0.033101960085332394,
      "epoch": 5.650602409638554,
      "grad_norm": 0.3676363229751587,
      "learning_rate": 1.4292832606516255e-06,
      "loss": 0.08126258105039597,
      "mean_token_accuracy": 0.9955662786960602,
      "num_tokens": 38168498.0,
      "step": 237
    },
    {
      "aux_loss": 8.038545608520508,
      "entropy": 0.035288846120238304,
      "epoch": 5.674698795180722,
      "grad_norm": 0.36642128229141235,
      "learning_rate": 1.4291013071548844e-06,
      "loss": 0.08272764086723328,
      "mean_token_accuracy": 0.9954147934913635,
      "num_tokens": 38340718.0,
      "step": 238
    },
    {
      "aux_loss": 8.038650035858154,
      "entropy": 0.03259859047830105,
      "epoch": 5.698795180722891,
      "grad_norm": 0.37624770402908325,
      "learning_rate": 1.428918581819328e-06,
      "loss": 0.08136217296123505,
      "mean_token_accuracy": 0.9960896074771881,
      "num_tokens": 38491373.0,
      "step": 239
    },
    {
      "aux_loss": 8.038266658782959,
      "entropy": 0.04310726746916771,
      "epoch": 5.72289156626506,
      "grad_norm": 0.43467146158218384,
      "learning_rate": 1.4287350850578186e-06,
      "loss": 0.08748319000005722,
      "mean_token_accuracy": 0.9937043190002441,
      "num_tokens": 38639028.0,
      "step": 240
    },
    {
      "aux_loss": 8.039052963256836,
      "entropy": 0.038960253819823265,
      "epoch": 5.746987951807229,
      "grad_norm": 0.40311357378959656,
      "learning_rate": 1.4285508172849628e-06,
      "loss": 0.08427684009075165,
      "mean_token_accuracy": 0.9949110746383667,
      "num_tokens": 38800330.0,
      "step": 241
    },
    {
      "aux_loss": 8.038615703582764,
      "entropy": 0.038292521610856056,
      "epoch": 5.771084337349397,
      "grad_norm": 0.37565159797668457,
      "learning_rate": 1.4283657789171085e-06,
      "loss": 0.08385735750198364,
      "mean_token_accuracy": 0.9953287243843079,
      "num_tokens": 38948505.0,
      "step": 242
    },
    {
      "aux_loss": 8.039562702178955,
      "entropy": 0.03734001237899065,
      "epoch": 5.795180722891566,
      "grad_norm": 0.49312564730644226,
      "learning_rate": 1.4281799703723446e-06,
      "loss": 0.08247716724872589,
      "mean_token_accuracy": 0.9953389167785645,
      "num_tokens": 39093422.0,
      "step": 243
    },
    {
      "aux_loss": 8.038256645202637,
      "entropy": 0.03265094570815563,
      "epoch": 5.8192771084337345,
      "grad_norm": 0.3904739022254944,
      "learning_rate": 1.4279933920705006e-06,
      "loss": 0.08162276446819305,
      "mean_token_accuracy": 0.9952907860279083,
      "num_tokens": 39267459.0,
      "step": 244
    },
    {
      "aux_loss": 8.037330627441406,
      "entropy": 0.03682544641196728,
      "epoch": 5.843373493975903,
      "grad_norm": 0.397100031375885,
      "learning_rate": 1.4278060444331451e-06,
      "loss": 0.08333480358123779,
      "mean_token_accuracy": 0.9949441850185394,
      "num_tokens": 39422907.0,
      "step": 245
    },
    {
      "aux_loss": 8.038883686065674,
      "entropy": 0.04045848548412323,
      "epoch": 5.867469879518072,
      "grad_norm": 0.4215233623981476,
      "learning_rate": 1.4276179278835848e-06,
      "loss": 0.08535461127758026,
      "mean_token_accuracy": 0.9943825006484985,
      "num_tokens": 39579636.0,
      "step": 246
    },
    {
      "aux_loss": 8.039095878601074,
      "entropy": 0.033118318766355515,
      "epoch": 5.891566265060241,
      "grad_norm": 0.3765990138053894,
      "learning_rate": 1.427429042846864e-06,
      "loss": 0.08269308507442474,
      "mean_token_accuracy": 0.9951244592666626,
      "num_tokens": 39753106.0,
      "step": 247
    },
    {
      "aux_loss": 8.037528038024902,
      "entropy": 0.03207779489457607,
      "epoch": 5.9156626506024095,
      "grad_norm": 0.3518119156360626,
      "learning_rate": 1.4272393897497632e-06,
      "loss": 0.08054585009813309,
      "mean_token_accuracy": 0.9962049722671509,
      "num_tokens": 39912952.0,
      "step": 248
    },
    {
      "aux_loss": 8.037781715393066,
      "entropy": 0.035104215145111084,
      "epoch": 5.9397590361445785,
      "grad_norm": 0.3629360795021057,
      "learning_rate": 1.4270489690207984e-06,
      "loss": 0.0826537013053894,
      "mean_token_accuracy": 0.9953400492668152,
      "num_tokens": 40083911.0,
      "step": 249
    },
    {
      "aux_loss": 8.03683614730835,
      "entropy": 0.034510133787989616,
      "epoch": 5.9638554216867465,
      "grad_norm": 0.37607911229133606,
      "learning_rate": 1.4268577810902201e-06,
      "loss": 0.08266992121934891,
      "mean_token_accuracy": 0.9955215454101562,
      "num_tokens": 40245447.0,
      "step": 250
    },
    {
      "aux_loss": 8.036605834960938,
      "entropy": 0.03547329269349575,
      "epoch": 5.9879518072289155,
      "grad_norm": 0.3630676567554474,
      "learning_rate": 1.4266658263900122e-06,
      "loss": 0.08295561373233795,
      "mean_token_accuracy": 0.995036780834198,
      "num_tokens": 40406226.0,
      "step": 251
    },
    {
      "aux_loss": 8.03968620300293,
      "entropy": 0.04327882453799248,
      "epoch": 6.0,
      "grad_norm": 0.5565012693405151,
      "learning_rate": 1.426473105353891e-06,
      "loss": 0.05327405780553818,
      "mean_token_accuracy": 0.9942361116409302,
      "num_tokens": 40448778.0,
      "step": 252
    },
    {
      "aux_loss": 8.037991523742676,
      "entropy": 0.03559446148574352,
      "epoch": 6.024096385542169,
      "grad_norm": 0.360224187374115,
      "learning_rate": 1.4262796184173047e-06,
      "loss": 0.08191978931427002,
      "mean_token_accuracy": 0.9960768818855286,
      "num_tokens": 40621399.0,
      "step": 253
    },
    {
      "aux_loss": 8.036860942840576,
      "entropy": 0.03401390090584755,
      "epoch": 6.048192771084337,
      "grad_norm": 0.337534099817276,
      "learning_rate": 1.4260853660174314e-06,
      "loss": 0.08101242035627365,
      "mean_token_accuracy": 0.9961624443531036,
      "num_tokens": 40775180.0,
      "step": 254
    },
    {
      "aux_loss": 8.03651237487793,
      "entropy": 0.03811703994870186,
      "epoch": 6.072289156626506,
      "grad_norm": 0.3574149012565613,
      "learning_rate": 1.4258903485931792e-06,
      "loss": 0.08330834656953812,
      "mean_token_accuracy": 0.9954338371753693,
      "num_tokens": 40924568.0,
      "step": 255
    },
    {
      "aux_loss": 8.036168575286865,
      "entropy": 0.03353409469127655,
      "epoch": 6.096385542168675,
      "grad_norm": 0.3296172618865967,
      "learning_rate": 1.4256945665851847e-06,
      "loss": 0.08137848973274231,
      "mean_token_accuracy": 0.9961557686328888,
      "num_tokens": 41090965.0,
      "step": 256
    },
    {
      "aux_loss": 8.037352561950684,
      "entropy": 0.032407025806605816,
      "epoch": 6.120481927710843,
      "grad_norm": 0.3700825870037079,
      "learning_rate": 1.425498020435812e-06,
      "loss": 0.08028987050056458,
      "mean_token_accuracy": 0.9963049590587616,
      "num_tokens": 41260686.0,
      "step": 257
    },
    {
      "aux_loss": 8.037732124328613,
      "entropy": 0.03817327693104744,
      "epoch": 6.144578313253012,
      "grad_norm": 0.4077751934528351,
      "learning_rate": 1.4253007105891516e-06,
      "loss": 0.08337738364934921,
      "mean_token_accuracy": 0.9955407381057739,
      "num_tokens": 41389101.0,
      "step": 258
    },
    {
      "aux_loss": 8.03888988494873,
      "entropy": 0.033729828894138336,
      "epoch": 6.168674698795181,
      "grad_norm": 0.36580121517181396,
      "learning_rate": 1.4251026374910198e-06,
      "loss": 0.08104156702756882,
      "mean_token_accuracy": 0.9964797794818878,
      "num_tokens": 41559499.0,
      "step": 259
    },
    {
      "aux_loss": 8.036795616149902,
      "entropy": 0.03791739046573639,
      "epoch": 6.192771084337349,
      "grad_norm": 0.37497496604919434,
      "learning_rate": 1.4249038015889575e-06,
      "loss": 0.08325420320034027,
      "mean_token_accuracy": 0.9954448640346527,
      "num_tokens": 41705274.0,
      "step": 260
    },
    {
      "aux_loss": 8.037520408630371,
      "entropy": 0.03363081254065037,
      "epoch": 6.216867469879518,
      "grad_norm": 0.3485323488712311,
      "learning_rate": 1.4247042033322288e-06,
      "loss": 0.08111946284770966,
      "mean_token_accuracy": 0.9960795640945435,
      "num_tokens": 41861920.0,
      "step": 261
    },
    {
      "aux_loss": 8.037950038909912,
      "entropy": 0.032483769580721855,
      "epoch": 6.240963855421687,
      "grad_norm": 0.32457178831100464,
      "learning_rate": 1.4245038431718204e-06,
      "loss": 0.07976509630680084,
      "mean_token_accuracy": 0.9964925646781921,
      "num_tokens": 42030201.0,
      "step": 262
    },
    {
      "aux_loss": 8.037893772125244,
      "entropy": 0.04124067164957523,
      "epoch": 6.265060240963855,
      "grad_norm": 0.4116523563861847,
      "learning_rate": 1.424302721560441e-06,
      "loss": 0.0846301019191742,
      "mean_token_accuracy": 0.9952834844589233,
      "num_tokens": 42177436.0,
      "step": 263
    },
    {
      "aux_loss": 8.037344932556152,
      "entropy": 0.03193550743162632,
      "epoch": 6.289156626506024,
      "grad_norm": 0.34990963339805603,
      "learning_rate": 1.424100838952519e-06,
      "loss": 0.08070550858974457,
      "mean_token_accuracy": 0.9957850873470306,
      "num_tokens": 42351445.0,
      "step": 264
    },
    {
      "aux_loss": 8.036174774169922,
      "entropy": 0.03807485289871693,
      "epoch": 6.313253012048193,
      "grad_norm": 0.39334771037101746,
      "learning_rate": 1.4238981958042027e-06,
      "loss": 0.0833045095205307,
      "mean_token_accuracy": 0.9958217740058899,
      "num_tokens": 42498981.0,
      "step": 265
    },
    {
      "aux_loss": 8.03799819946289,
      "entropy": 0.036790693178772926,
      "epoch": 6.337349397590361,
      "grad_norm": 0.36928167939186096,
      "learning_rate": 1.423694792573359e-06,
      "loss": 0.08284773677587509,
      "mean_token_accuracy": 0.9955715537071228,
      "num_tokens": 42691804.0,
      "step": 266
    },
    {
      "aux_loss": 8.036050796508789,
      "entropy": 0.03653477132320404,
      "epoch": 6.36144578313253,
      "grad_norm": 0.38227298855781555,
      "learning_rate": 1.4234906297195716e-06,
      "loss": 0.08240111172199249,
      "mean_token_accuracy": 0.9956892132759094,
      "num_tokens": 42839877.0,
      "step": 267
    },
    {
      "aux_loss": 8.036399364471436,
      "entropy": 0.03486090246587992,
      "epoch": 6.385542168674699,
      "grad_norm": 0.35526585578918457,
      "learning_rate": 1.4232857077041412e-06,
      "loss": 0.08210219442844391,
      "mean_token_accuracy": 0.9954483807086945,
      "num_tokens": 42997903.0,
      "step": 268
    },
    {
      "aux_loss": 8.03887128829956,
      "entropy": 0.031132452189922333,
      "epoch": 6.409638554216867,
      "grad_norm": 0.32824158668518066,
      "learning_rate": 1.4230800269900832e-06,
      "loss": 0.08031217753887177,
      "mean_token_accuracy": 0.9960026741027832,
      "num_tokens": 43184524.0,
      "step": 269
    },
    {
      "aux_loss": 8.037085056304932,
      "entropy": 0.03395012393593788,
      "epoch": 6.433734939759036,
      "grad_norm": 0.39305827021598816,
      "learning_rate": 1.422873588042128e-06,
      "loss": 0.0817105770111084,
      "mean_token_accuracy": 0.9954641461372375,
      "num_tokens": 43352033.0,
      "step": 270
    },
    {
      "aux_loss": 8.03740119934082,
      "entropy": 0.033392174169421196,
      "epoch": 6.457831325301205,
      "grad_norm": 0.341763436794281,
      "learning_rate": 1.422666391326718e-06,
      "loss": 0.08093781024217606,
      "mean_token_accuracy": 0.9962793290615082,
      "num_tokens": 43528223.0,
      "step": 271
    },
    {
      "aux_loss": 8.035551071166992,
      "entropy": 0.03841528482735157,
      "epoch": 6.481927710843373,
      "grad_norm": 0.40078938007354736,
      "learning_rate": 1.4224584373120093e-06,
      "loss": 0.08335836231708527,
      "mean_token_accuracy": 0.9956542551517487,
      "num_tokens": 43672878.0,
      "step": 272
    },
    {
      "aux_loss": 8.036177158355713,
      "entropy": 0.03540636785328388,
      "epoch": 6.506024096385542,
      "grad_norm": 0.37283870577812195,
      "learning_rate": 1.4222497264678678e-06,
      "loss": 0.08214269578456879,
      "mean_token_accuracy": 0.9954660832881927,
      "num_tokens": 43815578.0,
      "step": 273
    },
    {
      "aux_loss": 8.034242630004883,
      "entropy": 0.03531269542872906,
      "epoch": 6.530120481927711,
      "grad_norm": 0.38215646147727966,
      "learning_rate": 1.4220402592658704e-06,
      "loss": 0.08176004141569138,
      "mean_token_accuracy": 0.996109277009964,
      "num_tokens": 43962507.0,
      "step": 274
    },
    {
      "aux_loss": 8.03614330291748,
      "entropy": 0.03466031700372696,
      "epoch": 6.554216867469879,
      "grad_norm": 0.3749046325683594,
      "learning_rate": 1.421830036179302e-06,
      "loss": 0.08171263337135315,
      "mean_token_accuracy": 0.9958585798740387,
      "num_tokens": 44131399.0,
      "step": 275
    },
    {
      "aux_loss": 8.037889957427979,
      "entropy": 0.03461333829909563,
      "epoch": 6.578313253012048,
      "grad_norm": 0.34714046120643616,
      "learning_rate": 1.4216190576831563e-06,
      "loss": 0.081275574862957,
      "mean_token_accuracy": 0.9961965978145599,
      "num_tokens": 44286358.0,
      "step": 276
    },
    {
      "aux_loss": 8.03689432144165,
      "entropy": 0.034249075688421726,
      "epoch": 6.602409638554217,
      "grad_norm": 0.32196035981178284,
      "learning_rate": 1.4214073242541331e-06,
      "loss": 0.08108819276094437,
      "mean_token_accuracy": 0.9961409270763397,
      "num_tokens": 44456280.0,
      "step": 277
    },
    {
      "aux_loss": 8.039196014404297,
      "entropy": 0.03051491640508175,
      "epoch": 6.626506024096385,
      "grad_norm": 0.3290470838546753,
      "learning_rate": 1.421194836370639e-06,
      "loss": 0.08000116050243378,
      "mean_token_accuracy": 0.996231198310852,
      "num_tokens": 44651684.0,
      "step": 278
    },
    {
      "aux_loss": 8.033888816833496,
      "entropy": 0.03408683463931084,
      "epoch": 6.650602409638554,
      "grad_norm": 0.3726556599140167,
      "learning_rate": 1.4209815945127836e-06,
      "loss": 0.08141613751649857,
      "mean_token_accuracy": 0.9959953725337982,
      "num_tokens": 44794109.0,
      "step": 279
    },
    {
      "aux_loss": 8.035766124725342,
      "entropy": 0.028662268072366714,
      "epoch": 6.674698795180722,
      "grad_norm": 0.34546366333961487,
      "learning_rate": 1.4207675991623816e-06,
      "loss": 0.07919106632471085,
      "mean_token_accuracy": 0.9962852001190186,
      "num_tokens": 44967068.0,
      "step": 280
    },
    {
      "aux_loss": 8.036411762237549,
      "entropy": 0.03132207505404949,
      "epoch": 6.698795180722891,
      "grad_norm": 0.3478906750679016,
      "learning_rate": 1.4205528508029496e-06,
      "loss": 0.07999314367771149,
      "mean_token_accuracy": 0.9962101578712463,
      "num_tokens": 45141741.0,
      "step": 281
    },
    {
      "aux_loss": 8.03869342803955,
      "entropy": 0.03163009416311979,
      "epoch": 6.72289156626506,
      "grad_norm": 0.3555811941623688,
      "learning_rate": 1.4203373499197057e-06,
      "loss": 0.08022312819957733,
      "mean_token_accuracy": 0.9961192905902863,
      "num_tokens": 45317687.0,
      "step": 282
    },
    {
      "aux_loss": 8.036246299743652,
      "entropy": 0.029239365831017494,
      "epoch": 6.746987951807229,
      "grad_norm": 0.34071168303489685,
      "learning_rate": 1.4201210969995682e-06,
      "loss": 0.0793105736374855,
      "mean_token_accuracy": 0.9963167905807495,
      "num_tokens": 45504365.0,
      "step": 283
    },
    {
      "aux_loss": 8.03716516494751,
      "entropy": 0.03523336164653301,
      "epoch": 6.771084337349397,
      "grad_norm": 0.3549567461013794,
      "learning_rate": 1.4199040925311544e-06,
      "loss": 0.08193039894104004,
      "mean_token_accuracy": 0.9957601726055145,
      "num_tokens": 45690787.0,
      "step": 284
    },
    {
      "aux_loss": 8.036113262176514,
      "entropy": 0.03256150148808956,
      "epoch": 6.795180722891566,
      "grad_norm": 0.36457401514053345,
      "learning_rate": 1.4196863370047803e-06,
      "loss": 0.08117170631885529,
      "mean_token_accuracy": 0.9957631528377533,
      "num_tokens": 45850417.0,
      "step": 285
    },
    {
      "aux_loss": 8.035535335540771,
      "entropy": 0.03171788528561592,
      "epoch": 6.8192771084337345,
      "grad_norm": 0.351662278175354,
      "learning_rate": 1.4194678309124583e-06,
      "loss": 0.0804198831319809,
      "mean_token_accuracy": 0.9961467385292053,
      "num_tokens": 46013319.0,
      "step": 286
    },
    {
      "aux_loss": 8.033803462982178,
      "entropy": 0.035000565461814404,
      "epoch": 6.843373493975903,
      "grad_norm": 0.3920998275279999,
      "learning_rate": 1.4192485747478974e-06,
      "loss": 0.08269241452217102,
      "mean_token_accuracy": 0.9949135482311249,
      "num_tokens": 46188144.0,
      "step": 287
    },
    {
      "aux_loss": 8.03599739074707,
      "entropy": 0.03238750249147415,
      "epoch": 6.867469879518072,
      "grad_norm": 0.35797789692878723,
      "learning_rate": 1.4190285690065004e-06,
      "loss": 0.08076779544353485,
      "mean_token_accuracy": 0.9960897862911224,
      "num_tokens": 46341440.0,
      "step": 288
    },
    {
      "aux_loss": 8.035959243774414,
      "entropy": 0.03638170100748539,
      "epoch": 6.891566265060241,
      "grad_norm": 0.3702660799026489,
      "learning_rate": 1.4188078141853645e-06,
      "loss": 0.08285987377166748,
      "mean_token_accuracy": 0.9956148564815521,
      "num_tokens": 46502026.0,
      "step": 289
    },
    {
      "aux_loss": 8.034866333007812,
      "entropy": 0.034105464816093445,
      "epoch": 6.9156626506024095,
      "grad_norm": 0.3803676962852478,
      "learning_rate": 1.418586310783279e-06,
      "loss": 0.08216661214828491,
      "mean_token_accuracy": 0.9953741133213043,
      "num_tokens": 46670197.0,
      "step": 290
    },
    {
      "aux_loss": 8.038596153259277,
      "entropy": 0.03966710343956947,
      "epoch": 6.9397590361445785,
      "grad_norm": 0.3919319808483124,
      "learning_rate": 1.4183640593007253e-06,
      "loss": 0.08459129929542542,
      "mean_token_accuracy": 0.9947242140769958,
      "num_tokens": 46827184.0,
      "step": 291
    },
    {
      "aux_loss": 8.035058975219727,
      "entropy": 0.0370329562574625,
      "epoch": 6.9638554216867465,
      "grad_norm": 0.3695419132709503,
      "learning_rate": 1.418141060239874e-06,
      "loss": 0.08234144747257233,
      "mean_token_accuracy": 0.9958735704421997,
      "num_tokens": 46975123.0,
      "step": 292
    },
    {
      "aux_loss": 8.034956932067871,
      "entropy": 0.03599015064537525,
      "epoch": 6.9879518072289155,
      "grad_norm": 0.372112512588501,
      "learning_rate": 1.4179173141045854e-06,
      "loss": 0.08284641802310944,
      "mean_token_accuracy": 0.9954551458358765,
      "num_tokens": 47143463.0,
      "step": 293
    },
    {
      "aux_loss": 8.03744888305664,
      "entropy": 0.030542269349098206,
      "epoch": 7.0,
      "grad_norm": 0.47627538442611694,
      "learning_rate": 1.417692821400408e-06,
      "loss": 0.04683445766568184,
      "mean_token_accuracy": 0.995657205581665,
      "num_tokens": 47190241.0,
      "step": 294
    },
    {
      "aux_loss": 8.033048629760742,
      "entropy": 0.035450490191578865,
      "epoch": 7.024096385542169,
      "grad_norm": 0.34915679693222046,
      "learning_rate": 1.417467582634577e-06,
      "loss": 0.08114811033010483,
      "mean_token_accuracy": 0.9963439106941223,
      "num_tokens": 47348345.0,
      "step": 295
    },
    {
      "aux_loss": 8.034916877746582,
      "entropy": 0.03222184535115957,
      "epoch": 7.048192771084337,
      "grad_norm": 0.32226625084877014,
      "learning_rate": 1.417241598316013e-06,
      "loss": 0.07955966889858246,
      "mean_token_accuracy": 0.9968876838684082,
      "num_tokens": 47515241.0,
      "step": 296
    },
    {
      "aux_loss": 8.035215377807617,
      "entropy": 0.03598351962864399,
      "epoch": 7.072289156626506,
      "grad_norm": 0.34765782952308655,
      "learning_rate": 1.4170148689553215e-06,
      "loss": 0.08156674355268478,
      "mean_token_accuracy": 0.9967100620269775,
      "num_tokens": 47663653.0,
      "step": 297
    },
    {
      "aux_loss": 8.035448551177979,
      "entropy": 0.032758140936493874,
      "epoch": 7.096385542168675,
      "grad_norm": 0.338187038898468,
      "learning_rate": 1.4167873950647913e-06,
      "loss": 0.07984046638011932,
      "mean_token_accuracy": 0.9966485798358917,
      "num_tokens": 47828224.0,
      "step": 298
    },
    {
      "aux_loss": 8.03564167022705,
      "entropy": 0.0363839715719223,
      "epoch": 7.120481927710843,
      "grad_norm": 0.3636642396450043,
      "learning_rate": 1.4165591771583934e-06,
      "loss": 0.0818038135766983,
      "mean_token_accuracy": 0.9960289299488068,
      "num_tokens": 47965572.0,
      "step": 299
    },
    {
      "aux_loss": 8.037732124328613,
      "entropy": 0.03432103618979454,
      "epoch": 7.144578313253012,
      "grad_norm": 0.36236822605133057,
      "learning_rate": 1.41633021575178e-06,
      "loss": 0.08107459545135498,
      "mean_token_accuracy": 0.996305376291275,
      "num_tokens": 48137642.0,
      "step": 300
    },
    {
      "aux_loss": 8.036556720733643,
      "entropy": 0.03693149797618389,
      "epoch": 7.168674698795181,
      "grad_norm": 0.34610867500305176,
      "learning_rate": 1.416100511362283e-06,
      "loss": 0.0821981430053711,
      "mean_token_accuracy": 0.99636110663414,
      "num_tokens": 48324625.0,
      "step": 301
    },
    {
      "aux_loss": 8.034196376800537,
      "entropy": 0.029057555831968784,
      "epoch": 7.192771084337349,
      "grad_norm": 0.32378068566322327,
      "learning_rate": 1.4158700645089132e-06,
      "loss": 0.07781493663787842,
      "mean_token_accuracy": 0.9974943101406097,
      "num_tokens": 48469198.0,
      "step": 302
    },
    {
      "aux_loss": 8.036748886108398,
      "entropy": 0.03262832574546337,
      "epoch": 7.216867469879518,
      "grad_norm": 0.3372521996498108,
      "learning_rate": 1.415638875712359e-06,
      "loss": 0.08014824986457825,
      "mean_token_accuracy": 0.9964551031589508,
      "num_tokens": 48634260.0,
      "step": 303
    },
    {
      "aux_loss": 8.037166595458984,
      "entropy": 0.032166702672839165,
      "epoch": 7.240963855421687,
      "grad_norm": 0.34929683804512024,
      "learning_rate": 1.4154069454949852e-06,
      "loss": 0.08027002215385437,
      "mean_token_accuracy": 0.996328204870224,
      "num_tokens": 48804133.0,
      "step": 304
    },
    {
      "aux_loss": 8.03482961654663,
      "entropy": 0.03405679203569889,
      "epoch": 7.265060240963855,
      "grad_norm": 0.34674111008644104,
      "learning_rate": 1.415174274380832e-06,
      "loss": 0.08115323632955551,
      "mean_token_accuracy": 0.9964562058448792,
      "num_tokens": 48964195.0,
      "step": 305
    },
    {
      "aux_loss": 8.032870292663574,
      "entropy": 0.030429409816861153,
      "epoch": 7.289156626506024,
      "grad_norm": 0.3174818158149719,
      "learning_rate": 1.4149408628956132e-06,
      "loss": 0.07890744507312775,
      "mean_token_accuracy": 0.9967007339000702,
      "num_tokens": 49129035.0,
      "step": 306
    },
    {
      "aux_loss": 8.035819053649902,
      "entropy": 0.028406686149537563,
      "epoch": 7.313253012048193,
      "grad_norm": 0.3168235123157501,
      "learning_rate": 1.4147067115667155e-06,
      "loss": 0.07822935283184052,
      "mean_token_accuracy": 0.996906578540802,
      "num_tokens": 49300740.0,
      "step": 307
    },
    {
      "aux_loss": 8.03471565246582,
      "entropy": 0.03468763642013073,
      "epoch": 7.337349397590361,
      "grad_norm": 0.3807547092437744,
      "learning_rate": 1.4144718209231978e-06,
      "loss": 0.08181793242692947,
      "mean_token_accuracy": 0.9959447979927063,
      "num_tokens": 49446228.0,
      "step": 308
    },
    {
      "aux_loss": 8.035276889801025,
      "entropy": 0.03498452343046665,
      "epoch": 7.36144578313253,
      "grad_norm": 0.37493276596069336,
      "learning_rate": 1.4142361914957885e-06,
      "loss": 0.0809975117444992,
      "mean_token_accuracy": 0.9967085421085358,
      "num_tokens": 49620124.0,
      "step": 309
    },
    {
      "aux_loss": 8.03686237335205,
      "entropy": 0.03699825517833233,
      "epoch": 7.385542168674699,
      "grad_norm": 0.380402147769928,
      "learning_rate": 1.4139998238168866e-06,
      "loss": 0.08249437808990479,
      "mean_token_accuracy": 0.9959370195865631,
      "num_tokens": 49785119.0,
      "step": 310
    },
    {
      "aux_loss": 8.03669023513794,
      "entropy": 0.03257426247000694,
      "epoch": 7.409638554216867,
      "grad_norm": 0.3667578101158142,
      "learning_rate": 1.413762718420558e-06,
      "loss": 0.08071167767047882,
      "mean_token_accuracy": 0.9963789582252502,
      "num_tokens": 49952981.0,
      "step": 311
    },
    {
      "aux_loss": 8.034005641937256,
      "entropy": 0.02788942214101553,
      "epoch": 7.433734939759036,
      "grad_norm": 0.2974223494529724,
      "learning_rate": 1.4135248758425359e-06,
      "loss": 0.07806290686130524,
      "mean_token_accuracy": 0.99703049659729,
      "num_tokens": 50143595.0,
      "step": 312
    },
    {
      "aux_loss": 8.033695697784424,
      "entropy": 0.03359641693532467,
      "epoch": 7.457831325301205,
      "grad_norm": 0.3415364623069763,
      "learning_rate": 1.413286296620219e-06,
      "loss": 0.0807826817035675,
      "mean_token_accuracy": 0.9961223006248474,
      "num_tokens": 50297111.0,
      "step": 313
    },
    {
      "aux_loss": 8.034516334533691,
      "entropy": 0.03204282093793154,
      "epoch": 7.481927710843373,
      "grad_norm": 0.37344956398010254,
      "learning_rate": 1.4130469812926708e-06,
      "loss": 0.08044802397489548,
      "mean_token_accuracy": 0.9963233172893524,
      "num_tokens": 50448811.0,
      "step": 314
    },
    {
      "aux_loss": 8.033761978149414,
      "entropy": 0.031184443272650242,
      "epoch": 7.506024096385542,
      "grad_norm": 0.34687599539756775,
      "learning_rate": 1.4128069304006176e-06,
      "loss": 0.07920951396226883,
      "mean_token_accuracy": 0.9964482486248016,
      "num_tokens": 50609715.0,
      "step": 315
    },
    {
      "aux_loss": 8.034749031066895,
      "entropy": 0.029387577436864376,
      "epoch": 7.530120481927711,
      "grad_norm": 0.31832030415534973,
      "learning_rate": 1.4125661444864477e-06,
      "loss": 0.07884480059146881,
      "mean_token_accuracy": 0.9967005848884583,
      "num_tokens": 50772691.0,
      "step": 316
    },
    {
      "aux_loss": 8.03644847869873,
      "entropy": 0.03081085067242384,
      "epoch": 7.554216867469879,
      "grad_norm": 0.3318593204021454,
      "learning_rate": 1.4123246240942104e-06,
      "loss": 0.07955613732337952,
      "mean_token_accuracy": 0.996260017156601,
      "num_tokens": 50946461.0,
      "step": 317
    },
    {
      "aux_loss": 8.03480863571167,
      "entropy": 0.029540213756263256,
      "epoch": 7.578313253012048,
      "grad_norm": 0.3266274631023407,
      "learning_rate": 1.4120823697696144e-06,
      "loss": 0.07922594994306564,
      "mean_token_accuracy": 0.9967344701290131,
      "num_tokens": 51132312.0,
      "step": 318
    },
    {
      "aux_loss": 8.034060955047607,
      "entropy": 0.032534703612327576,
      "epoch": 7.602409638554217,
      "grad_norm": 0.35959550738334656,
      "learning_rate": 1.4118393820600268e-06,
      "loss": 0.08021605014801025,
      "mean_token_accuracy": 0.9963286817073822,
      "num_tokens": 51292228.0,
      "step": 319
    },
    {
      "aux_loss": 8.034628868103027,
      "entropy": 0.03200039733201265,
      "epoch": 7.626506024096385,
      "grad_norm": 0.3670026957988739,
      "learning_rate": 1.4115956615144717e-06,
      "loss": 0.0801118016242981,
      "mean_token_accuracy": 0.9965364933013916,
      "num_tokens": 51460338.0,
      "step": 320
    },
    {
      "aux_loss": 8.0361909866333,
      "entropy": 0.029335936531424522,
      "epoch": 7.650602409638554,
      "grad_norm": 0.35437721014022827,
      "learning_rate": 1.4113512086836288e-06,
      "loss": 0.07883332669734955,
      "mean_token_accuracy": 0.9966942071914673,
      "num_tokens": 51619895.0,
      "step": 321
    },
    {
      "aux_loss": 8.035427570343018,
      "entropy": 0.032436246052384377,
      "epoch": 7.674698795180722,
      "grad_norm": 0.34983742237091064,
      "learning_rate": 1.4111060241198329e-06,
      "loss": 0.08025002479553223,
      "mean_token_accuracy": 0.9963445365428925,
      "num_tokens": 51785833.0,
      "step": 322
    },
    {
      "aux_loss": 8.033997058868408,
      "entropy": 0.03502529673278332,
      "epoch": 7.698795180722891,
      "grad_norm": 0.39680731296539307,
      "learning_rate": 1.4108601083770718e-06,
      "loss": 0.08174897730350494,
      "mean_token_accuracy": 0.9958369731903076,
      "num_tokens": 51950251.0,
      "step": 323
    },
    {
      "aux_loss": 8.03468132019043,
      "entropy": 0.039487240836024284,
      "epoch": 7.72289156626506,
      "grad_norm": 0.3813090920448303,
      "learning_rate": 1.4106134620109848e-06,
      "loss": 0.08411987125873566,
      "mean_token_accuracy": 0.9956846833229065,
      "num_tokens": 52104274.0,
      "step": 324
    },
    {
      "aux_loss": 8.034347534179688,
      "entropy": 0.0300306910648942,
      "epoch": 7.746987951807229,
      "grad_norm": 0.3322348892688751,
      "learning_rate": 1.4103660855788636e-06,
      "loss": 0.07942084223031998,
      "mean_token_accuracy": 0.9967765212059021,
      "num_tokens": 52266704.0,
      "step": 325
    },
    {
      "aux_loss": 8.032934665679932,
      "entropy": 0.030973643995821476,
      "epoch": 7.771084337349397,
      "grad_norm": 0.3334362208843231,
      "learning_rate": 1.4101179796396477e-06,
      "loss": 0.07914458960294724,
      "mean_token_accuracy": 0.9966495633125305,
      "num_tokens": 52446376.0,
      "step": 326
    },
    {
      "aux_loss": 8.032021522521973,
      "entropy": 0.030552386306226254,
      "epoch": 7.795180722891566,
      "grad_norm": 0.37135079503059387,
      "learning_rate": 1.409869144753926e-06,
      "loss": 0.07951807975769043,
      "mean_token_accuracy": 0.9962173700332642,
      "num_tokens": 52600001.0,
      "step": 327
    },
    {
      "aux_loss": 8.035574913024902,
      "entropy": 0.03106264304369688,
      "epoch": 7.8192771084337345,
      "grad_norm": 0.3528701663017273,
      "learning_rate": 1.4096195814839339e-06,
      "loss": 0.0800810307264328,
      "mean_token_accuracy": 0.996282309293747,
      "num_tokens": 52768318.0,
      "step": 328
    },
    {
      "aux_loss": 8.035249710083008,
      "entropy": 0.03810280002653599,
      "epoch": 7.843373493975903,
      "grad_norm": 0.38768336176872253,
      "learning_rate": 1.4093692903935533e-06,
      "loss": 0.08325342833995819,
      "mean_token_accuracy": 0.9957176744937897,
      "num_tokens": 52937588.0,
      "step": 329
    },
    {
      "aux_loss": 8.033713340759277,
      "entropy": 0.034183334559202194,
      "epoch": 7.867469879518072,
      "grad_norm": 0.35069799423217773,
      "learning_rate": 1.4091182720483095e-06,
      "loss": 0.08079741895198822,
      "mean_token_accuracy": 0.9960699379444122,
      "num_tokens": 53089655.0,
      "step": 330
    },
    {
      "aux_loss": 8.032924175262451,
      "entropy": 0.03523244708776474,
      "epoch": 7.891566265060241,
      "grad_norm": 0.39111587405204773,
      "learning_rate": 1.4088665270153717e-06,
      "loss": 0.08104196190834045,
      "mean_token_accuracy": 0.9962975680828094,
      "num_tokens": 53248483.0,
      "step": 331
    },
    {
      "aux_loss": 8.033989906311035,
      "entropy": 0.03644202649593353,
      "epoch": 7.9156626506024095,
      "grad_norm": 0.3810213506221771,
      "learning_rate": 1.4086140558635514e-06,
      "loss": 0.08186738193035126,
      "mean_token_accuracy": 0.9956882297992706,
      "num_tokens": 53411522.0,
      "step": 332
    },
    {
      "aux_loss": 8.034379005432129,
      "entropy": 0.03020784631371498,
      "epoch": 7.9397590361445785,
      "grad_norm": 0.3371445834636688,
      "learning_rate": 1.4083608591632996e-06,
      "loss": 0.07884131371974945,
      "mean_token_accuracy": 0.9969273209571838,
      "num_tokens": 53574408.0,
      "step": 333
    },
    {
      "aux_loss": 8.03269910812378,
      "entropy": 0.031987156718969345,
      "epoch": 7.9638554216867465,
      "grad_norm": 0.368277370929718,
      "learning_rate": 1.4081069374867079e-06,
      "loss": 0.07966729998588562,
      "mean_token_accuracy": 0.9962456524372101,
      "num_tokens": 53736277.0,
      "step": 334
    },
    {
      "aux_loss": 8.032955646514893,
      "entropy": 0.034337449818849564,
      "epoch": 7.9879518072289155,
      "grad_norm": 0.34589219093322754,
      "learning_rate": 1.407852291407505e-06,
      "loss": 0.08137495815753937,
      "mean_token_accuracy": 0.9959911406040192,
      "num_tokens": 53893541.0,
      "step": 335
    },
    {
      "aux_loss": 8.037422180175781,
      "entropy": 0.03607993572950363,
      "epoch": 8.0,
      "grad_norm": 0.48692095279693604,
      "learning_rate": 1.4075969215010567e-06,
      "loss": 0.04885231703519821,
      "mean_token_accuracy": 0.9962315559387207,
      "num_tokens": 53931704.0,
      "step": 336
    },
    {
      "aux_loss": 8.03453254699707,
      "entropy": 0.03251357935369015,
      "epoch": 8.024096385542169,
      "grad_norm": 0.31673136353492737,
      "learning_rate": 1.4073408283443648e-06,
      "loss": 0.07995766401290894,
      "mean_token_accuracy": 0.9967318177223206,
      "num_tokens": 54091047.0,
      "step": 337
    },
    {
      "aux_loss": 8.032040119171143,
      "entropy": 0.03007432259619236,
      "epoch": 8.048192771084338,
      "grad_norm": 0.30801841616630554,
      "learning_rate": 1.4070840125160644e-06,
      "loss": 0.07856522500514984,
      "mean_token_accuracy": 0.997085839509964,
      "num_tokens": 54251503.0,
      "step": 338
    },
    {
      "aux_loss": 8.033209323883057,
      "entropy": 0.0301275672391057,
      "epoch": 8.072289156626505,
      "grad_norm": 0.32101917266845703,
      "learning_rate": 1.4068264745964238e-06,
      "loss": 0.07812288403511047,
      "mean_token_accuracy": 0.9971620738506317,
      "num_tokens": 54431460.0,
      "step": 339
    },
    {
      "aux_loss": 8.03450059890747,
      "entropy": 0.037498150020837784,
      "epoch": 8.096385542168674,
      "grad_norm": 0.3406485617160797,
      "learning_rate": 1.4065682151673432e-06,
      "loss": 0.08174111694097519,
      "mean_token_accuracy": 0.9965010583400726,
      "num_tokens": 54590360.0,
      "step": 340
    },
    {
      "aux_loss": 8.035029888153076,
      "entropy": 0.031323011964559555,
      "epoch": 8.120481927710843,
      "grad_norm": 0.48250916600227356,
      "learning_rate": 1.4063092348123523e-06,
      "loss": 0.07874792069196701,
      "mean_token_accuracy": 0.9969989359378815,
      "num_tokens": 54752146.0,
      "step": 341
    },
    {
      "aux_loss": 8.034884929656982,
      "entropy": 0.030849049799144268,
      "epoch": 8.144578313253012,
      "grad_norm": 0.3505721688270569,
      "learning_rate": 1.4060495341166107e-06,
      "loss": 0.07901881635189056,
      "mean_token_accuracy": 0.9968648850917816,
      "num_tokens": 54910384.0,
      "step": 342
    },
    {
      "aux_loss": 8.031527996063232,
      "entropy": 0.03027996141463518,
      "epoch": 8.168674698795181,
      "grad_norm": 0.3279053568840027,
      "learning_rate": 1.4057891136669046e-06,
      "loss": 0.0783233493566513,
      "mean_token_accuracy": 0.9972791373729706,
      "num_tokens": 55074547.0,
      "step": 343
    },
    {
      "aux_loss": 8.032787322998047,
      "entropy": 0.03463793359696865,
      "epoch": 8.19277108433735,
      "grad_norm": 0.37317872047424316,
      "learning_rate": 1.4055279740516467e-06,
      "loss": 0.08077745139598846,
      "mean_token_accuracy": 0.9964223802089691,
      "num_tokens": 55223107.0,
      "step": 344
    },
    {
      "aux_loss": 8.032590866088867,
      "entropy": 0.03586240112781525,
      "epoch": 8.216867469879517,
      "grad_norm": 0.3791426420211792,
      "learning_rate": 1.4052661158608755e-06,
      "loss": 0.08125555515289307,
      "mean_token_accuracy": 0.9962078630924225,
      "num_tokens": 55367867.0,
      "step": 345
    },
    {
      "aux_loss": 8.033146858215332,
      "entropy": 0.03275267221033573,
      "epoch": 8.240963855421686,
      "grad_norm": 0.3555714786052704,
      "learning_rate": 1.4050035396862518e-06,
      "loss": 0.0799730122089386,
      "mean_token_accuracy": 0.9965274631977081,
      "num_tokens": 55531222.0,
      "step": 346
    },
    {
      "aux_loss": 8.033248901367188,
      "entropy": 0.029476456344127655,
      "epoch": 8.265060240963855,
      "grad_norm": 0.323041707277298,
      "learning_rate": 1.4047402461210596e-06,
      "loss": 0.07879751920700073,
      "mean_token_accuracy": 0.9967617392539978,
      "num_tokens": 55707018.0,
      "step": 347
    },
    {
      "aux_loss": 8.034295558929443,
      "entropy": 0.034110626205801964,
      "epoch": 8.289156626506024,
      "grad_norm": 0.3517826199531555,
      "learning_rate": 1.4044762357602033e-06,
      "loss": 0.08017922937870026,
      "mean_token_accuracy": 0.9966349303722382,
      "num_tokens": 55859239.0,
      "step": 348
    },
    {
      "aux_loss": 8.031924724578857,
      "entropy": 0.02905238326638937,
      "epoch": 8.313253012048193,
      "grad_norm": 0.3139900863170624,
      "learning_rate": 1.4042115092002074e-06,
      "loss": 0.07803536206483841,
      "mean_token_accuracy": 0.9969979524612427,
      "num_tokens": 56002942.0,
      "step": 349
    },
    {
      "aux_loss": 8.03225040435791,
      "entropy": 0.030559261329472065,
      "epoch": 8.337349397590362,
      "grad_norm": 0.34452271461486816,
      "learning_rate": 1.403946067039214e-06,
      "loss": 0.07925976067781448,
      "mean_token_accuracy": 0.9967726767063141,
      "num_tokens": 56159221.0,
      "step": 350
    },
    {
      "aux_loss": 8.033653736114502,
      "entropy": 0.036190446466207504,
      "epoch": 8.36144578313253,
      "grad_norm": 0.3688271939754486,
      "learning_rate": 1.4036799098769828e-06,
      "loss": 0.08133170008659363,
      "mean_token_accuracy": 0.9966385960578918,
      "num_tokens": 56326889.0,
      "step": 351
    },
    {
      "aux_loss": 8.032990455627441,
      "entropy": 0.03038295917212963,
      "epoch": 8.385542168674698,
      "grad_norm": 0.3074115812778473,
      "learning_rate": 1.4034130383148882e-06,
      "loss": 0.0787685215473175,
      "mean_token_accuracy": 0.9970142543315887,
      "num_tokens": 56498544.0,
      "step": 352
    },
    {
      "aux_loss": 8.03279447555542,
      "entropy": 0.03136066906154156,
      "epoch": 8.409638554216867,
      "grad_norm": 0.3597790002822876,
      "learning_rate": 1.4031454529559193e-06,
      "loss": 0.07899552583694458,
      "mean_token_accuracy": 0.9971287846565247,
      "num_tokens": 56661229.0,
      "step": 353
    },
    {
      "aux_loss": 8.03244400024414,
      "entropy": 0.03564039617776871,
      "epoch": 8.433734939759036,
      "grad_norm": 0.36866894364356995,
      "learning_rate": 1.4028771544046781e-06,
      "loss": 0.08136041462421417,
      "mean_token_accuracy": 0.9964535534381866,
      "num_tokens": 56809009.0,
      "step": 354
    },
    {
      "aux_loss": 8.033175945281982,
      "entropy": 0.03262287098914385,
      "epoch": 8.457831325301205,
      "grad_norm": 0.3500673770904541,
      "learning_rate": 1.4026081432673777e-06,
      "loss": 0.08022792637348175,
      "mean_token_accuracy": 0.9964220225811005,
      "num_tokens": 56974040.0,
      "step": 355
    },
    {
      "aux_loss": 8.034413814544678,
      "entropy": 0.03208298236131668,
      "epoch": 8.481927710843374,
      "grad_norm": 0.3360278308391571,
      "learning_rate": 1.4023384201518416e-06,
      "loss": 0.07970739901065826,
      "mean_token_accuracy": 0.9968045651912689,
      "num_tokens": 57140951.0,
      "step": 356
    },
    {
      "aux_loss": 8.033014297485352,
      "entropy": 0.030674396082758904,
      "epoch": 8.506024096385541,
      "grad_norm": 0.3344880938529968,
      "learning_rate": 1.4020679856675015e-06,
      "loss": 0.07849317789077759,
      "mean_token_accuracy": 0.9969932436943054,
      "num_tokens": 57282214.0,
      "step": 357
    },
    {
      "aux_loss": 8.031963348388672,
      "entropy": 0.03275129199028015,
      "epoch": 8.53012048192771,
      "grad_norm": 0.3430642783641815,
      "learning_rate": 1.401796840425397e-06,
      "loss": 0.07970772683620453,
      "mean_token_accuracy": 0.9970153570175171,
      "num_tokens": 57430073.0,
      "step": 358
    },
    {
      "aux_loss": 8.032482147216797,
      "entropy": 0.02784721739590168,
      "epoch": 8.55421686746988,
      "grad_norm": 0.33207616209983826,
      "learning_rate": 1.401524985038173e-06,
      "loss": 0.0774514451622963,
      "mean_token_accuracy": 0.9972799718379974,
      "num_tokens": 57611033.0,
      "step": 359
    },
    {
      "aux_loss": 8.032303810119629,
      "entropy": 0.02809650544077158,
      "epoch": 8.578313253012048,
      "grad_norm": 0.36203494668006897,
      "learning_rate": 1.4012524201200799e-06,
      "loss": 0.07764801383018494,
      "mean_token_accuracy": 0.9973397850990295,
      "num_tokens": 57775590.0,
      "step": 360
    },
    {
      "aux_loss": 8.03242826461792,
      "entropy": 0.03094232641160488,
      "epoch": 8.602409638554217,
      "grad_norm": 0.3259294033050537,
      "learning_rate": 1.40097914628697e-06,
      "loss": 0.07914900034666061,
      "mean_token_accuracy": 0.9970998764038086,
      "num_tokens": 57943047.0,
      "step": 361
    },
    {
      "aux_loss": 8.03185749053955,
      "entropy": 0.027600662782788277,
      "epoch": 8.626506024096386,
      "grad_norm": 0.30300894379615784,
      "learning_rate": 1.4007051641562986e-06,
      "loss": 0.07772274315357208,
      "mean_token_accuracy": 0.9973132312297821,
      "num_tokens": 58120845.0,
      "step": 362
    },
    {
      "aux_loss": 8.033005237579346,
      "entropy": 0.030716686509549618,
      "epoch": 8.650602409638553,
      "grad_norm": 0.41477662324905396,
      "learning_rate": 1.4004304743471206e-06,
      "loss": 0.07921659201383591,
      "mean_token_accuracy": 0.9964025318622589,
      "num_tokens": 58290435.0,
      "step": 363
    },
    {
      "aux_loss": 8.031022548675537,
      "entropy": 0.032445063814520836,
      "epoch": 8.674698795180722,
      "grad_norm": 0.36098530888557434,
      "learning_rate": 1.4001550774800902e-06,
      "loss": 0.07978940010070801,
      "mean_token_accuracy": 0.9963515102863312,
      "num_tokens": 58446144.0,
      "step": 364
    },
    {
      "aux_loss": 8.03183650970459,
      "entropy": 0.028307861648499966,
      "epoch": 8.698795180722891,
      "grad_norm": 0.37047818303108215,
      "learning_rate": 1.399878974177459e-06,
      "loss": 0.07777097821235657,
      "mean_token_accuracy": 0.997242659330368,
      "num_tokens": 58623005.0,
      "step": 365
    },
    {
      "aux_loss": 8.032839298248291,
      "entropy": 0.03144409507513046,
      "epoch": 8.72289156626506,
      "grad_norm": 0.33840411901474,
      "learning_rate": 1.3996021650630748e-06,
      "loss": 0.07940583676099777,
      "mean_token_accuracy": 0.9966592788696289,
      "num_tokens": 58781686.0,
      "step": 366
    },
    {
      "aux_loss": 8.030786514282227,
      "entropy": 0.027757197618484497,
      "epoch": 8.74698795180723,
      "grad_norm": 0.3691762387752533,
      "learning_rate": 1.3993246507623804e-06,
      "loss": 0.07730717957019806,
      "mean_token_accuracy": 0.9969027042388916,
      "num_tokens": 58924651.0,
      "step": 367
    },
    {
      "aux_loss": 8.032029628753662,
      "entropy": 0.02889639139175415,
      "epoch": 8.771084337349398,
      "grad_norm": 0.3496450185775757,
      "learning_rate": 1.3990464319024116e-06,
      "loss": 0.07893317937850952,
      "mean_token_accuracy": 0.9967939853668213,
      "num_tokens": 59083066.0,
      "step": 368
    },
    {
      "aux_loss": 8.031221866607666,
      "entropy": 0.030891859903931618,
      "epoch": 8.795180722891565,
      "grad_norm": 0.3323741555213928,
      "learning_rate": 1.3987675091117965e-06,
      "loss": 0.07858480513095856,
      "mean_token_accuracy": 0.9969135522842407,
      "num_tokens": 59256385.0,
      "step": 369
    },
    {
      "aux_loss": 8.032551288604736,
      "entropy": 0.03319684974849224,
      "epoch": 8.819277108433734,
      "grad_norm": 0.34501487016677856,
      "learning_rate": 1.3984878830207534e-06,
      "loss": 0.08011599630117416,
      "mean_token_accuracy": 0.996588408946991,
      "num_tokens": 59427852.0,
      "step": 370
    },
    {
      "aux_loss": 8.032721519470215,
      "entropy": 0.03409439139068127,
      "epoch": 8.843373493975903,
      "grad_norm": 0.3768217861652374,
      "learning_rate": 1.39820755426109e-06,
      "loss": 0.08085048198699951,
      "mean_token_accuracy": 0.9963680505752563,
      "num_tokens": 59595387.0,
      "step": 371
    },
    {
      "aux_loss": 8.032885551452637,
      "entropy": 0.03258616290986538,
      "epoch": 8.867469879518072,
      "grad_norm": 0.3588944375514984,
      "learning_rate": 1.3979265234662012e-06,
      "loss": 0.0799306333065033,
      "mean_token_accuracy": 0.9965479075908661,
      "num_tokens": 59779351.0,
      "step": 372
    },
    {
      "aux_loss": 8.030370235443115,
      "entropy": 0.029054300859570503,
      "epoch": 8.891566265060241,
      "grad_norm": 0.33350667357444763,
      "learning_rate": 1.3976447912710685e-06,
      "loss": 0.07802186906337738,
      "mean_token_accuracy": 0.9969941079616547,
      "num_tokens": 59949200.0,
      "step": 373
    },
    {
      "aux_loss": 8.02937126159668,
      "entropy": 0.028985220938920975,
      "epoch": 8.91566265060241,
      "grad_norm": 0.3544178605079651,
      "learning_rate": 1.3973623583122578e-06,
      "loss": 0.07832932472229004,
      "mean_token_accuracy": 0.9969984292984009,
      "num_tokens": 60109064.0,
      "step": 374
    },
    {
      "aux_loss": 8.03177261352539,
      "entropy": 0.029587114229798317,
      "epoch": 8.939759036144578,
      "grad_norm": 0.3520343005657196,
      "learning_rate": 1.3970792252279195e-06,
      "loss": 0.07858137786388397,
      "mean_token_accuracy": 0.996891975402832,
      "num_tokens": 60274692.0,
      "step": 375
    },
    {
      "aux_loss": 8.035192966461182,
      "entropy": 0.03562385682016611,
      "epoch": 8.963855421686747,
      "grad_norm": 0.35002952814102173,
      "learning_rate": 1.3967953926577841e-06,
      "loss": 0.08112400025129318,
      "mean_token_accuracy": 0.996430367231369,
      "num_tokens": 60441996.0,
      "step": 376
    },
    {
      "aux_loss": 8.033750534057617,
      "entropy": 0.029073555022478104,
      "epoch": 8.987951807228916,
      "grad_norm": 0.346101850271225,
      "learning_rate": 1.3965108612431643e-06,
      "loss": 0.07820385694503784,
      "mean_token_accuracy": 0.9969233572483063,
      "num_tokens": 60627910.0,
      "step": 377
    },
    {
      "aux_loss": 8.03187370300293,
      "entropy": 0.02936927229166031,
      "epoch": 9.0,
      "grad_norm": 0.46344229578971863,
      "learning_rate": 1.3962256316269505e-06,
      "loss": 0.04597163945436478,
      "mean_token_accuracy": 0.9973151087760925,
      "num_tokens": 60673167.0,
      "step": 378
    },
    {
      "aux_loss": 8.03305721282959,
      "entropy": 0.028284212574362755,
      "epoch": 9.024096385542169,
      "grad_norm": 0.31423506140708923,
      "learning_rate": 1.3959397044536114e-06,
      "loss": 0.07713969051837921,
      "mean_token_accuracy": 0.9977681040763855,
      "num_tokens": 60836716.0,
      "step": 379
    },
    {
      "aux_loss": 8.032850742340088,
      "entropy": 0.026606920175254345,
      "epoch": 9.048192771084338,
      "grad_norm": 0.2880466878414154,
      "learning_rate": 1.3956530803691917e-06,
      "loss": 0.0765802338719368,
      "mean_token_accuracy": 0.9976675510406494,
      "num_tokens": 61000573.0,
      "step": 380
    },
    {
      "aux_loss": 8.031289100646973,
      "entropy": 0.030671123415231705,
      "epoch": 9.072289156626505,
      "grad_norm": 0.30061787366867065,
      "learning_rate": 1.3953657600213106e-06,
      "loss": 0.07775267213582993,
      "mean_token_accuracy": 0.9975506663322449,
      "num_tokens": 61154706.0,
      "step": 381
    },
    {
      "aux_loss": 8.032238960266113,
      "entropy": 0.027959519997239113,
      "epoch": 9.096385542168674,
      "grad_norm": 0.28902363777160645,
      "learning_rate": 1.3950777440591605e-06,
      "loss": 0.07733355462551117,
      "mean_token_accuracy": 0.9973697662353516,
      "num_tokens": 61333594.0,
      "step": 382
    },
    {
      "aux_loss": 8.030166625976562,
      "entropy": 0.032277317717671394,
      "epoch": 9.120481927710843,
      "grad_norm": 0.34240710735321045,
      "learning_rate": 1.394789033133506e-06,
      "loss": 0.07901263236999512,
      "mean_token_accuracy": 0.9973782598972321,
      "num_tokens": 61490776.0,
      "step": 383
    },
    {
      "aux_loss": 8.032149314880371,
      "entropy": 0.02693926077336073,
      "epoch": 9.144578313253012,
      "grad_norm": 0.29341667890548706,
      "learning_rate": 1.3944996278966811e-06,
      "loss": 0.076640784740448,
      "mean_token_accuracy": 0.9978146553039551,
      "num_tokens": 61669552.0,
      "step": 384
    },
    {
      "aux_loss": 8.031457901000977,
      "entropy": 0.03191028628498316,
      "epoch": 9.168674698795181,
      "grad_norm": 0.3397424519062042,
      "learning_rate": 1.394209529002589e-06,
      "loss": 0.07893210649490356,
      "mean_token_accuracy": 0.9971943795681,
      "num_tokens": 61843274.0,
      "step": 385
    },
    {
      "aux_loss": 8.03022575378418,
      "entropy": 0.028519313782453537,
      "epoch": 9.19277108433735,
      "grad_norm": 0.30692535638809204,
      "learning_rate": 1.3939187371067007e-06,
      "loss": 0.07746940851211548,
      "mean_token_accuracy": 0.9976769387722015,
      "num_tokens": 61998707.0,
      "step": 386
    },
    {
      "aux_loss": 8.031204223632812,
      "entropy": 0.03095682244747877,
      "epoch": 9.216867469879517,
      "grad_norm": 0.3476795554161072,
      "learning_rate": 1.393627252866052e-06,
      "loss": 0.07881052792072296,
      "mean_token_accuracy": 0.9969981908798218,
      "num_tokens": 62173983.0,
      "step": 387
    },
    {
      "aux_loss": 8.027823448181152,
      "entropy": 0.02978881075978279,
      "epoch": 9.240963855421686,
      "grad_norm": 0.3190312683582306,
      "learning_rate": 1.3933350769392442e-06,
      "loss": 0.07773149758577347,
      "mean_token_accuracy": 0.9973650872707367,
      "num_tokens": 62324901.0,
      "step": 388
    },
    {
      "aux_loss": 8.031431674957275,
      "entropy": 0.027737611904740334,
      "epoch": 9.265060240963855,
      "grad_norm": 0.38071316480636597,
      "learning_rate": 1.39304220998644e-06,
      "loss": 0.07697569578886032,
      "mean_token_accuracy": 0.9974884390830994,
      "num_tokens": 62485228.0,
      "step": 389
    },
    {
      "aux_loss": 8.030527114868164,
      "entropy": 0.02737813163548708,
      "epoch": 9.289156626506024,
      "grad_norm": 0.2937212586402893,
      "learning_rate": 1.3927486526693647e-06,
      "loss": 0.0766058936715126,
      "mean_token_accuracy": 0.9978461861610413,
      "num_tokens": 62647707.0,
      "step": 390
    },
    {
      "aux_loss": 8.030436038970947,
      "entropy": 0.030430599115788937,
      "epoch": 9.313253012048193,
      "grad_norm": 0.30997201800346375,
      "learning_rate": 1.3924544056513032e-06,
      "loss": 0.0783412903547287,
      "mean_token_accuracy": 0.9975597560405731,
      "num_tokens": 62816112.0,
      "step": 391
    },
    {
      "aux_loss": 8.03123950958252,
      "entropy": 0.026135905645787716,
      "epoch": 9.337349397590362,
      "grad_norm": 0.3543896973133087,
      "learning_rate": 1.3921594695970985e-06,
      "loss": 0.07653771340847015,
      "mean_token_accuracy": 0.997521847486496,
      "num_tokens": 62950912.0,
      "step": 392
    },
    {
      "aux_loss": 8.02945613861084,
      "entropy": 0.03068769257515669,
      "epoch": 9.36144578313253,
      "grad_norm": 0.3361717462539673,
      "learning_rate": 1.3918638451731505e-06,
      "loss": 0.07847397774457932,
      "mean_token_accuracy": 0.9971339702606201,
      "num_tokens": 63106341.0,
      "step": 393
    },
    {
      "aux_loss": 8.031757354736328,
      "entropy": 0.03277725353837013,
      "epoch": 9.385542168674698,
      "grad_norm": 0.3406221866607666,
      "learning_rate": 1.3915675330474142e-06,
      "loss": 0.0791986957192421,
      "mean_token_accuracy": 0.9971778690814972,
      "num_tokens": 63284372.0,
      "step": 394
    },
    {
      "aux_loss": 8.030030727386475,
      "entropy": 0.026919249445199966,
      "epoch": 9.409638554216867,
      "grad_norm": 0.31440675258636475,
      "learning_rate": 1.3912705338893993e-06,
      "loss": 0.07689384371042252,
      "mean_token_accuracy": 0.9976073503494263,
      "num_tokens": 63455837.0,
      "step": 395
    },
    {
      "aux_loss": 8.031475067138672,
      "entropy": 0.033060451969504356,
      "epoch": 9.433734939759036,
      "grad_norm": 0.3352949023246765,
      "learning_rate": 1.3909728483701666e-06,
      "loss": 0.07927119731903076,
      "mean_token_accuracy": 0.9972446262836456,
      "num_tokens": 63618970.0,
      "step": 396
    },
    {
      "aux_loss": 8.029807567596436,
      "entropy": 0.03108399175107479,
      "epoch": 9.457831325301205,
      "grad_norm": 0.32964977622032166,
      "learning_rate": 1.3906744771623291e-06,
      "loss": 0.07835537195205688,
      "mean_token_accuracy": 0.9973717033863068,
      "num_tokens": 63781997.0,
      "step": 397
    },
    {
      "aux_loss": 8.031740665435791,
      "entropy": 0.03269661497324705,
      "epoch": 9.481927710843374,
      "grad_norm": 0.3676154911518097,
      "learning_rate": 1.3903754209400482e-06,
      "loss": 0.07965146005153656,
      "mean_token_accuracy": 0.9964734315872192,
      "num_tokens": 63932105.0,
      "step": 398
    },
    {
      "aux_loss": 8.033281803131104,
      "entropy": 0.03186710365116596,
      "epoch": 9.506024096385541,
      "grad_norm": 0.35776206851005554,
      "learning_rate": 1.390075680379033e-06,
      "loss": 0.07902336865663528,
      "mean_token_accuracy": 0.9969694018363953,
      "num_tokens": 64093887.0,
      "step": 399
    },
    {
      "aux_loss": 8.032399654388428,
      "entropy": 0.024799546226859093,
      "epoch": 9.53012048192771,
      "grad_norm": 0.3226422369480133,
      "learning_rate": 1.3897752561565394e-06,
      "loss": 0.07591937482357025,
      "mean_token_accuracy": 0.9975130259990692,
      "num_tokens": 64273357.0,
      "step": 400
    },
    {
      "aux_loss": 8.033264636993408,
      "entropy": 0.0295121967792511,
      "epoch": 9.55421686746988,
      "grad_norm": 0.3351059854030609,
      "learning_rate": 1.3894741489513673e-06,
      "loss": 0.07776688039302826,
      "mean_token_accuracy": 0.997167706489563,
      "num_tokens": 64448304.0,
      "step": 401
    },
    {
      "aux_loss": 8.029650211334229,
      "entropy": 0.02971556317061186,
      "epoch": 9.578313253012048,
      "grad_norm": 0.32523661851882935,
      "learning_rate": 1.3891723594438605e-06,
      "loss": 0.07813172787427902,
      "mean_token_accuracy": 0.9973180294036865,
      "num_tokens": 64596938.0,
      "step": 402
    },
    {
      "aux_loss": 8.031736373901367,
      "entropy": 0.02978976722806692,
      "epoch": 9.602409638554217,
      "grad_norm": 0.3332604169845581,
      "learning_rate": 1.3888698883159043e-06,
      "loss": 0.07776395231485367,
      "mean_token_accuracy": 0.9974326193332672,
      "num_tokens": 64769145.0,
      "step": 403
    },
    {
      "aux_loss": 8.033589839935303,
      "entropy": 0.032476719468832016,
      "epoch": 9.626506024096386,
      "grad_norm": 0.3411960005760193,
      "learning_rate": 1.3885667362509237e-06,
      "loss": 0.07948120683431625,
      "mean_token_accuracy": 0.9970299303531647,
      "num_tokens": 64932386.0,
      "step": 404
    },
    {
      "aux_loss": 8.030903816223145,
      "entropy": 0.03362698294222355,
      "epoch": 9.650602409638553,
      "grad_norm": 0.3845645487308502,
      "learning_rate": 1.3882629039338825e-06,
      "loss": 0.08039066195487976,
      "mean_token_accuracy": 0.9965060353279114,
      "num_tokens": 65085807.0,
      "step": 405
    },
    {
      "aux_loss": 8.029520034790039,
      "entropy": 0.0270113842561841,
      "epoch": 9.674698795180722,
      "grad_norm": 0.34778332710266113,
      "learning_rate": 1.3879583920512816e-06,
      "loss": 0.077044278383255,
      "mean_token_accuracy": 0.9973159432411194,
      "num_tokens": 65246006.0,
      "step": 406
    },
    {
      "aux_loss": 8.030782699584961,
      "entropy": 0.03195221349596977,
      "epoch": 9.698795180722891,
      "grad_norm": 0.35124245285987854,
      "learning_rate": 1.3876532012911574e-06,
      "loss": 0.07930511981248856,
      "mean_token_accuracy": 0.9969338178634644,
      "num_tokens": 65411013.0,
      "step": 407
    },
    {
      "aux_loss": 8.029433727264404,
      "entropy": 0.03094009030610323,
      "epoch": 9.72289156626506,
      "grad_norm": 0.4142710268497467,
      "learning_rate": 1.3873473323430797e-06,
      "loss": 0.07911791652441025,
      "mean_token_accuracy": 0.9967486560344696,
      "num_tokens": 65571476.0,
      "step": 408
    },
    {
      "aux_loss": 8.030499935150146,
      "entropy": 0.027457045391201973,
      "epoch": 9.74698795180723,
      "grad_norm": 0.3272239863872528,
      "learning_rate": 1.3870407858981514e-06,
      "loss": 0.07713578641414642,
      "mean_token_accuracy": 0.9972514510154724,
      "num_tokens": 65742609.0,
      "step": 409
    },
    {
      "aux_loss": 8.029773712158203,
      "entropy": 0.030966890044510365,
      "epoch": 9.771084337349398,
      "grad_norm": 0.42443791031837463,
      "learning_rate": 1.3867335626490055e-06,
      "loss": 0.07875967025756836,
      "mean_token_accuracy": 0.9968508183956146,
      "num_tokens": 65891366.0,
      "step": 410
    },
    {
      "aux_loss": 8.029150009155273,
      "entropy": 0.03328642249107361,
      "epoch": 9.795180722891565,
      "grad_norm": 0.3802852928638458,
      "learning_rate": 1.3864256632898051e-06,
      "loss": 0.07969282567501068,
      "mean_token_accuracy": 0.9970137476921082,
      "num_tokens": 66041964.0,
      "step": 411
    },
    {
      "aux_loss": 8.030668258666992,
      "entropy": 0.02882132213562727,
      "epoch": 9.819277108433734,
      "grad_norm": 0.3516421616077423,
      "learning_rate": 1.3861170885162396e-06,
      "loss": 0.07771005481481552,
      "mean_token_accuracy": 0.9975076615810394,
      "num_tokens": 66231004.0,
      "step": 412
    },
    {
      "aux_loss": 8.029856204986572,
      "entropy": 0.028738058172166348,
      "epoch": 9.843373493975903,
      "grad_norm": 0.34768593311309814,
      "learning_rate": 1.3858078390255256e-06,
      "loss": 0.07793055474758148,
      "mean_token_accuracy": 0.9971697330474854,
      "num_tokens": 66399510.0,
      "step": 413
    },
    {
      "aux_loss": 8.030166625976562,
      "entropy": 0.03400828316807747,
      "epoch": 9.867469879518072,
      "grad_norm": 0.37204456329345703,
      "learning_rate": 1.3854979155164043e-06,
      "loss": 0.08015817403793335,
      "mean_token_accuracy": 0.9967809617519379,
      "num_tokens": 66547156.0,
      "step": 414
    },
    {
      "aux_loss": 8.032981872558594,
      "entropy": 0.03100480604916811,
      "epoch": 9.891566265060241,
      "grad_norm": 0.3481873869895935,
      "learning_rate": 1.3851873186891386e-06,
      "loss": 0.07907165586948395,
      "mean_token_accuracy": 0.9964969456195831,
      "num_tokens": 66718160.0,
      "step": 415
    },
    {
      "aux_loss": 8.028274536132812,
      "entropy": 0.030552691780030727,
      "epoch": 9.91566265060241,
      "grad_norm": 0.33890053629875183,
      "learning_rate": 1.3848760492455145e-06,
      "loss": 0.07868947088718414,
      "mean_token_accuracy": 0.9966758191585541,
      "num_tokens": 66873016.0,
      "step": 416
    },
    {
      "aux_loss": 8.02976369857788,
      "entropy": 0.03107893466949463,
      "epoch": 9.939759036144578,
      "grad_norm": 0.3599156141281128,
      "learning_rate": 1.384564107888836e-06,
      "loss": 0.078771211206913,
      "mean_token_accuracy": 0.9969359636306763,
      "num_tokens": 67038673.0,
      "step": 417
    },
    {
      "aux_loss": 8.02978229522705,
      "entropy": 0.028388715349137783,
      "epoch": 9.963855421686747,
      "grad_norm": 0.35516616702079773,
      "learning_rate": 1.384251495323926e-06,
      "loss": 0.0778445154428482,
      "mean_token_accuracy": 0.9971783459186554,
      "num_tokens": 67218897.0,
      "step": 418
    },
    {
      "aux_loss": 8.031276226043701,
      "entropy": 0.03284810110926628,
      "epoch": 9.987951807228916,
      "grad_norm": 0.3593505024909973,
      "learning_rate": 1.3839382122571248e-06,
      "loss": 0.0795031189918518,
      "mean_token_accuracy": 0.9974339604377747,
      "num_tokens": 67376548.0,
      "step": 419
    },
    {
      "aux_loss": 8.033153533935547,
      "entropy": 0.029837898910045624,
      "epoch": 10.0,
      "grad_norm": 0.4149579107761383,
      "learning_rate": 1.3836242593962863e-06,
      "loss": 0.04504719376564026,
      "mean_token_accuracy": 0.997219443321228,
      "num_tokens": 67414630.0,
      "step": 420
    },
    {
      "aux_loss": 8.03052043914795,
      "entropy": 0.026807748712599277,
      "epoch": 10.024096385542169,
      "grad_norm": 0.3050729036331177,
      "learning_rate": 1.3833096374507787e-06,
      "loss": 0.07622632384300232,
      "mean_token_accuracy": 0.9978890419006348,
      "num_tokens": 67588234.0,
      "step": 421
    },
    {
      "aux_loss": 8.031226634979248,
      "entropy": 0.033840239979326725,
      "epoch": 10.048192771084338,
      "grad_norm": 0.3194749653339386,
      "learning_rate": 1.3829943471314814e-06,
      "loss": 0.07884673774242401,
      "mean_token_accuracy": 0.9978678822517395,
      "num_tokens": 67757352.0,
      "step": 422
    },
    {
      "aux_loss": 8.029906272888184,
      "entropy": 0.027255422435700893,
      "epoch": 10.072289156626505,
      "grad_norm": 0.30284735560417175,
      "learning_rate": 1.3826783891507848e-06,
      "loss": 0.07626917958259583,
      "mean_token_accuracy": 0.9980226457118988,
      "num_tokens": 67920553.0,
      "step": 423
    },
    {
      "aux_loss": 8.02976942062378,
      "entropy": 0.027208078652620316,
      "epoch": 10.096385542168674,
      "grad_norm": 0.29062679409980774,
      "learning_rate": 1.3823617642225866e-06,
      "loss": 0.07693935185670853,
      "mean_token_accuracy": 0.9976333379745483,
      "num_tokens": 68081290.0,
      "step": 424
    },
    {
      "aux_loss": 8.029796123504639,
      "entropy": 0.0285436250269413,
      "epoch": 10.120481927710843,
      "grad_norm": 0.2950977683067322,
      "learning_rate": 1.3820444730622926e-06,
      "loss": 0.07694471627473831,
      "mean_token_accuracy": 0.9980784356594086,
      "num_tokens": 68271566.0,
      "step": 425
    },
    {
      "aux_loss": 8.031370639801025,
      "entropy": 0.03286069352179766,
      "epoch": 10.144578313253012,
      "grad_norm": 0.3592938780784607,
      "learning_rate": 1.3817265163868138e-06,
      "loss": 0.07915187627077103,
      "mean_token_accuracy": 0.9973708391189575,
      "num_tokens": 68410627.0,
      "step": 426
    },
    {
      "aux_loss": 8.02876901626587,
      "entropy": 0.033930208534002304,
      "epoch": 10.168674698795181,
      "grad_norm": 0.3631734549999237,
      "learning_rate": 1.381407894914564e-06,
      "loss": 0.07949021458625793,
      "mean_token_accuracy": 0.9971204400062561,
      "num_tokens": 68547059.0,
      "step": 427
    },
    {
      "aux_loss": 8.028857231140137,
      "entropy": 0.03004965092986822,
      "epoch": 10.19277108433735,
      "grad_norm": 0.32348141074180603,
      "learning_rate": 1.3810886093654599e-06,
      "loss": 0.07751224935054779,
      "mean_token_accuracy": 0.997689813375473,
      "num_tokens": 68719122.0,
      "step": 428
    },
    {
      "aux_loss": 8.0299711227417,
      "entropy": 0.025603781454265118,
      "epoch": 10.216867469879517,
      "grad_norm": 0.2847801148891449,
      "learning_rate": 1.3807686604609185e-06,
      "loss": 0.07599260658025742,
      "mean_token_accuracy": 0.9978285729885101,
      "num_tokens": 68903105.0,
      "step": 429
    },
    {
      "aux_loss": 8.02917766571045,
      "entropy": 0.02435719594359398,
      "epoch": 10.240963855421686,
      "grad_norm": 0.31084614992141724,
      "learning_rate": 1.3804480489238556e-06,
      "loss": 0.07491791248321533,
      "mean_token_accuracy": 0.9980163872241974,
      "num_tokens": 69086727.0,
      "step": 430
    },
    {
      "aux_loss": 8.029499530792236,
      "entropy": 0.02250757720321417,
      "epoch": 10.265060240963855,
      "grad_norm": 0.2934803366661072,
      "learning_rate": 1.3801267754786845e-06,
      "loss": 0.07460962235927582,
      "mean_token_accuracy": 0.9981854856014252,
      "num_tokens": 69255826.0,
      "step": 431
    },
    {
      "aux_loss": 8.028985023498535,
      "entropy": 0.026835591532289982,
      "epoch": 10.289156626506024,
      "grad_norm": 0.31175994873046875,
      "learning_rate": 1.3798048408513135e-06,
      "loss": 0.0759553462266922,
      "mean_token_accuracy": 0.9978678524494171,
      "num_tokens": 69429204.0,
      "step": 432
    },
    {
      "aux_loss": 8.029627323150635,
      "entropy": 0.02608797326683998,
      "epoch": 10.313253012048193,
      "grad_norm": 0.28849494457244873,
      "learning_rate": 1.3794822457691448e-06,
      "loss": 0.07554003596305847,
      "mean_token_accuracy": 0.9981241524219513,
      "num_tokens": 69615147.0,
      "step": 433
    },
    {
      "aux_loss": 8.027892589569092,
      "entropy": 0.029431588016450405,
      "epoch": 10.337349397590362,
      "grad_norm": 0.3519872725009918,
      "learning_rate": 1.3791589909610738e-06,
      "loss": 0.07704924046993256,
      "mean_token_accuracy": 0.9977331459522247,
      "num_tokens": 69763662.0,
      "step": 434
    },
    {
      "aux_loss": 8.028160095214844,
      "entropy": 0.030695267021656036,
      "epoch": 10.36144578313253,
      "grad_norm": 0.35359182953834534,
      "learning_rate": 1.3788350771574854e-06,
      "loss": 0.0779634341597557,
      "mean_token_accuracy": 0.9976026713848114,
      "num_tokens": 69923670.0,
      "step": 435
    },
    {
      "aux_loss": 8.028687000274658,
      "entropy": 0.030781088396906853,
      "epoch": 10.385542168674698,
      "grad_norm": 0.3456423878669739,
      "learning_rate": 1.3785105050902542e-06,
      "loss": 0.07804363965988159,
      "mean_token_accuracy": 0.9974770545959473,
      "num_tokens": 70070248.0,
      "step": 436
    },
    {
      "aux_loss": 8.03036642074585,
      "entropy": 0.025892925448715687,
      "epoch": 10.409638554216867,
      "grad_norm": 0.42847299575805664,
      "learning_rate": 1.3781852754927418e-06,
      "loss": 0.07700430601835251,
      "mean_token_accuracy": 0.9975075125694275,
      "num_tokens": 70249380.0,
      "step": 437
    },
    {
      "aux_loss": 8.031606197357178,
      "entropy": 0.0267898328602314,
      "epoch": 10.433734939759036,
      "grad_norm": 0.29486900568008423,
      "learning_rate": 1.3778593890997959e-06,
      "loss": 0.07694413512945175,
      "mean_token_accuracy": 0.9977473318576813,
      "num_tokens": 70413199.0,
      "step": 438
    },
    {
      "aux_loss": 8.029096126556396,
      "entropy": 0.027300704270601273,
      "epoch": 10.457831325301205,
      "grad_norm": 0.3215588331222534,
      "learning_rate": 1.3775328466477474e-06,
      "loss": 0.07631190121173859,
      "mean_token_accuracy": 0.997810572385788,
      "num_tokens": 70586808.0,
      "step": 439
    },
    {
      "aux_loss": 8.030426025390625,
      "entropy": 0.02695034071803093,
      "epoch": 10.481927710843374,
      "grad_norm": 0.309374064207077,
      "learning_rate": 1.3772056488744102e-06,
      "loss": 0.07675328850746155,
      "mean_token_accuracy": 0.9976034462451935,
      "num_tokens": 70763440.0,
      "step": 440
    },
    {
      "aux_loss": 8.028189659118652,
      "entropy": 0.028853912837803364,
      "epoch": 10.506024096385541,
      "grad_norm": 0.335868239402771,
      "learning_rate": 1.3768777965190792e-06,
      "loss": 0.07752006500959396,
      "mean_token_accuracy": 0.9972420930862427,
      "num_tokens": 70910457.0,
      "step": 441
    },
    {
      "aux_loss": 8.029743194580078,
      "entropy": 0.025770947337150574,
      "epoch": 10.53012048192771,
      "grad_norm": 0.3330857455730438,
      "learning_rate": 1.3765492903225272e-06,
      "loss": 0.07615037262439728,
      "mean_token_accuracy": 0.9975591003894806,
      "num_tokens": 71067271.0,
      "step": 442
    },
    {
      "aux_loss": 8.027770519256592,
      "entropy": 0.02995127160102129,
      "epoch": 10.55421686746988,
      "grad_norm": 0.3459339737892151,
      "learning_rate": 1.3762201310270052e-06,
      "loss": 0.0779542475938797,
      "mean_token_accuracy": 0.9972722232341766,
      "num_tokens": 71225176.0,
      "step": 443
    },
    {
      "aux_loss": 8.02858829498291,
      "entropy": 0.024707495234906673,
      "epoch": 10.578313253012048,
      "grad_norm": 0.32370665669441223,
      "learning_rate": 1.3758903193762397e-06,
      "loss": 0.0753905326128006,
      "mean_token_accuracy": 0.9975995123386383,
      "num_tokens": 71391719.0,
      "step": 444
    },
    {
      "aux_loss": 8.029046535491943,
      "entropy": 0.02720453217625618,
      "epoch": 10.602409638554217,
      "grad_norm": 0.31780052185058594,
      "learning_rate": 1.3755598561154312e-06,
      "loss": 0.077033132314682,
      "mean_token_accuracy": 0.9974277913570404,
      "num_tokens": 71561690.0,
      "step": 445
    },
    {
      "aux_loss": 8.030248165130615,
      "entropy": 0.029147695749998093,
      "epoch": 10.626506024096386,
      "grad_norm": 0.3248000741004944,
      "learning_rate": 1.3752287419912521e-06,
      "loss": 0.07753097265958786,
      "mean_token_accuracy": 0.9971960484981537,
      "num_tokens": 71710468.0,
      "step": 446
    },
    {
      "aux_loss": 8.028928279876709,
      "entropy": 0.029994006268680096,
      "epoch": 10.650602409638553,
      "grad_norm": 0.33071792125701904,
      "learning_rate": 1.3748969777518458e-06,
      "loss": 0.07808565348386765,
      "mean_token_accuracy": 0.9973722398281097,
      "num_tokens": 71875559.0,
      "step": 447
    },
    {
      "aux_loss": 8.030104637145996,
      "entropy": 0.025511790066957474,
      "epoch": 10.674698795180722,
      "grad_norm": 0.32101917266845703,
      "learning_rate": 1.3745645641468248e-06,
      "loss": 0.07576514780521393,
      "mean_token_accuracy": 0.9978866875171661,
      "num_tokens": 72045798.0,
      "step": 448
    },
    {
      "aux_loss": 8.028583526611328,
      "entropy": 0.028393504209816456,
      "epoch": 10.698795180722891,
      "grad_norm": 0.31367799639701843,
      "learning_rate": 1.3742315019272683e-06,
      "loss": 0.07735928893089294,
      "mean_token_accuracy": 0.9973789155483246,
      "num_tokens": 72218153.0,
      "step": 449
    },
    {
      "aux_loss": 8.02676010131836,
      "entropy": 0.03315955400466919,
      "epoch": 10.72289156626506,
      "grad_norm": 0.33943045139312744,
      "learning_rate": 1.3738977918457213e-06,
      "loss": 0.07943802326917648,
      "mean_token_accuracy": 0.9972104430198669,
      "num_tokens": 72360308.0,
      "step": 450
    },
    {
      "aux_loss": 8.028342247009277,
      "entropy": 0.037015119567513466,
      "epoch": 10.74698795180723,
      "grad_norm": 0.35264134407043457,
      "learning_rate": 1.3735634346561928e-06,
      "loss": 0.08016332238912582,
      "mean_token_accuracy": 0.9973889291286469,
      "num_tokens": 72523679.0,
      "step": 451
    },
    {
      "aux_loss": 8.028265953063965,
      "entropy": 0.030274370685219765,
      "epoch": 10.771084337349398,
      "grad_norm": 0.31121376156806946,
      "learning_rate": 1.3732284311141535e-06,
      "loss": 0.0778329148888588,
      "mean_token_accuracy": 0.997382640838623,
      "num_tokens": 72676307.0,
      "step": 452
    },
    {
      "aux_loss": 8.028914451599121,
      "entropy": 0.02892636600881815,
      "epoch": 10.795180722891565,
      "grad_norm": 0.32904553413391113,
      "learning_rate": 1.3728927819765347e-06,
      "loss": 0.07730700075626373,
      "mean_token_accuracy": 0.9975307881832123,
      "num_tokens": 72827705.0,
      "step": 453
    },
    {
      "aux_loss": 8.028653144836426,
      "entropy": 0.026827383786439896,
      "epoch": 10.819277108433734,
      "grad_norm": 0.3565026521682739,
      "learning_rate": 1.3725564880017268e-06,
      "loss": 0.07710254937410355,
      "mean_token_accuracy": 0.9978848695755005,
      "num_tokens": 72984162.0,
      "step": 454
    },
    {
      "aux_loss": 8.027618408203125,
      "entropy": 0.03206733614206314,
      "epoch": 10.843373493975903,
      "grad_norm": 0.33514919877052307,
      "learning_rate": 1.3722195499495765e-06,
      "loss": 0.0789363905787468,
      "mean_token_accuracy": 0.9969374537467957,
      "num_tokens": 73141852.0,
      "step": 455
    },
    {
      "aux_loss": 8.030162811279297,
      "entropy": 0.027127815410494804,
      "epoch": 10.867469879518072,
      "grad_norm": 0.32852309942245483,
      "learning_rate": 1.3718819685813862e-06,
      "loss": 0.07681447267532349,
      "mean_token_accuracy": 0.9976330995559692,
      "num_tokens": 73318634.0,
      "step": 456
    },
    {
      "aux_loss": 8.028889179229736,
      "entropy": 0.02808957453817129,
      "epoch": 10.891566265060241,
      "grad_norm": 0.428763747215271,
      "learning_rate": 1.371543744659912e-06,
      "loss": 0.07722076028585434,
      "mean_token_accuracy": 0.9971847534179688,
      "num_tokens": 73467288.0,
      "step": 457
    },
    {
      "aux_loss": 8.028523445129395,
      "entropy": 0.033010026440024376,
      "epoch": 10.91566265060241,
      "grad_norm": 0.332423597574234,
      "learning_rate": 1.371204878949361e-06,
      "loss": 0.07916770875453949,
      "mean_token_accuracy": 0.9972398281097412,
      "num_tokens": 73624957.0,
      "step": 458
    },
    {
      "aux_loss": 8.027756214141846,
      "entropy": 0.03013605810701847,
      "epoch": 10.939759036144578,
      "grad_norm": 0.4300965666770935,
      "learning_rate": 1.3708653722153918e-06,
      "loss": 0.07825049012899399,
      "mean_token_accuracy": 0.9971310496330261,
      "num_tokens": 73772800.0,
      "step": 459
    },
    {
      "aux_loss": 8.030034065246582,
      "entropy": 0.027245910838246346,
      "epoch": 10.963855421686747,
      "grad_norm": 0.32759150862693787,
      "learning_rate": 1.3705252252251102e-06,
      "loss": 0.076337531208992,
      "mean_token_accuracy": 0.997997373342514,
      "num_tokens": 73971847.0,
      "step": 460
    },
    {
      "aux_loss": 8.029091358184814,
      "entropy": 0.03214616421610117,
      "epoch": 10.987951807228916,
      "grad_norm": 0.36098340153694153,
      "learning_rate": 1.370184438747069e-06,
      "loss": 0.07857733964920044,
      "mean_token_accuracy": 0.9973133206367493,
      "num_tokens": 74122193.0,
      "step": 461
    },
    {
      "aux_loss": 8.02614974975586,
      "entropy": 0.035153578966856,
      "epoch": 11.0,
      "grad_norm": 0.46866685152053833,
      "learning_rate": 1.3698430135512658e-06,
      "loss": 0.04593968391418457,
      "mean_token_accuracy": 0.9968705773353577,
      "num_tokens": 74156093.0,
      "step": 462
    },
    {
      "aux_loss": 8.027415752410889,
      "entropy": 0.028158558532595634,
      "epoch": 11.024096385542169,
      "grad_norm": 0.30607613921165466,
      "learning_rate": 1.369500950409142e-06,
      "loss": 0.07637624442577362,
      "mean_token_accuracy": 0.9978108406066895,
      "num_tokens": 74313575.0,
      "step": 463
    },
    {
      "aux_loss": 8.027459621429443,
      "entropy": 0.029599015600979328,
      "epoch": 11.048192771084338,
      "grad_norm": 0.357259064912796,
      "learning_rate": 1.3691582500935793e-06,
      "loss": 0.07699000835418701,
      "mean_token_accuracy": 0.9981439709663391,
      "num_tokens": 74470458.0,
      "step": 464
    },
    {
      "aux_loss": 8.026456356048584,
      "entropy": 0.028655465692281723,
      "epoch": 11.072289156626505,
      "grad_norm": 0.29174235463142395,
      "learning_rate": 1.3688149133789e-06,
      "loss": 0.07684057950973511,
      "mean_token_accuracy": 0.9977833330631256,
      "num_tokens": 74620787.0,
      "step": 465
    },
    {
      "aux_loss": 8.029138088226318,
      "entropy": 0.022806944325566292,
      "epoch": 11.096385542168674,
      "grad_norm": 0.2944920063018799,
      "learning_rate": 1.368470941040864e-06,
      "loss": 0.07433664798736572,
      "mean_token_accuracy": 0.9983370304107666,
      "num_tokens": 74777566.0,
      "step": 466
    },
    {
      "aux_loss": 8.026599407196045,
      "entropy": 0.025663558393716812,
      "epoch": 11.120481927710843,
      "grad_norm": 0.2916056513786316,
      "learning_rate": 1.3681263338566675e-06,
      "loss": 0.07559385895729065,
      "mean_token_accuracy": 0.9982210993766785,
      "num_tokens": 74948804.0,
      "step": 467
    },
    {
      "aux_loss": 8.026897430419922,
      "entropy": 0.03057689778506756,
      "epoch": 11.144578313253012,
      "grad_norm": 0.31129932403564453,
      "learning_rate": 1.3677810926049408e-06,
      "loss": 0.0773526206612587,
      "mean_token_accuracy": 0.99770787358284,
      "num_tokens": 75097224.0,
      "step": 468
    },
    {
      "aux_loss": 8.024751663208008,
      "entropy": 0.028811393305659294,
      "epoch": 11.168674698795181,
      "grad_norm": 0.3139060437679291,
      "learning_rate": 1.3674352180657472e-06,
      "loss": 0.07690249383449554,
      "mean_token_accuracy": 0.997710257768631,
      "num_tokens": 75237910.0,
      "step": 469
    },
    {
      "aux_loss": 8.026570796966553,
      "entropy": 0.027774027548730373,
      "epoch": 11.19277108433735,
      "grad_norm": 0.30671951174736023,
      "learning_rate": 1.367088711020581e-06,
      "loss": 0.07627267390489578,
      "mean_token_accuracy": 0.9979667663574219,
      "num_tokens": 75384961.0,
      "step": 470
    },
    {
      "aux_loss": 8.029362201690674,
      "entropy": 0.028830256313085556,
      "epoch": 11.216867469879517,
      "grad_norm": 0.3399089574813843,
      "learning_rate": 1.3667415722523648e-06,
      "loss": 0.07696400582790375,
      "mean_token_accuracy": 0.9977167546749115,
      "num_tokens": 75547412.0,
      "step": 471
    },
    {
      "aux_loss": 8.029022693634033,
      "entropy": 0.026009133085608482,
      "epoch": 11.240963855421686,
      "grad_norm": 0.2968512177467346,
      "learning_rate": 1.3663938025454495e-06,
      "loss": 0.07592800259590149,
      "mean_token_accuracy": 0.997904509305954,
      "num_tokens": 75721807.0,
      "step": 472
    },
    {
      "aux_loss": 8.027937889099121,
      "entropy": 0.03286511264741421,
      "epoch": 11.265060240963855,
      "grad_norm": 0.3452129662036896,
      "learning_rate": 1.3660454026856114e-06,
      "loss": 0.0784171000123024,
      "mean_token_accuracy": 0.9975327253341675,
      "num_tokens": 75877734.0,
      "step": 473
    },
    {
      "aux_loss": 8.028258323669434,
      "entropy": 0.025823739357292652,
      "epoch": 11.289156626506024,
      "grad_norm": 0.3015165328979492,
      "learning_rate": 1.3656963734600503e-06,
      "loss": 0.07567039132118225,
      "mean_token_accuracy": 0.9982956647872925,
      "num_tokens": 76034122.0,
      "step": 474
    },
    {
      "aux_loss": 8.026177406311035,
      "entropy": 0.03207191079854965,
      "epoch": 11.313253012048193,
      "grad_norm": 0.3374171853065491,
      "learning_rate": 1.3653467156573882e-06,
      "loss": 0.07849003374576569,
      "mean_token_accuracy": 0.9976714849472046,
      "num_tokens": 76192441.0,
      "step": 475
    },
    {
      "aux_loss": 8.027956008911133,
      "entropy": 0.026211952790617943,
      "epoch": 11.337349397590362,
      "grad_norm": 0.35711753368377686,
      "learning_rate": 1.3649964300676674e-06,
      "loss": 0.07585370540618896,
      "mean_token_accuracy": 0.9977020621299744,
      "num_tokens": 76356381.0,
      "step": 476
    },
    {
      "aux_loss": 8.02691650390625,
      "entropy": 0.03026073705404997,
      "epoch": 11.36144578313253,
      "grad_norm": 0.3285326361656189,
      "learning_rate": 1.3646455174823483e-06,
      "loss": 0.07800611853599548,
      "mean_token_accuracy": 0.9974254965782166,
      "num_tokens": 76515766.0,
      "step": 477
    },
    {
      "aux_loss": 8.028995513916016,
      "entropy": 0.02668727468699217,
      "epoch": 11.385542168674698,
      "grad_norm": 0.3030661642551422,
      "learning_rate": 1.3642939786943089e-06,
      "loss": 0.07588149607181549,
      "mean_token_accuracy": 0.9981980621814728,
      "num_tokens": 76672811.0,
      "step": 478
    },
    {
      "aux_loss": 8.028472423553467,
      "entropy": 0.02798299677670002,
      "epoch": 11.409638554216867,
      "grad_norm": 0.30959540605545044,
      "learning_rate": 1.363941814497841e-06,
      "loss": 0.07691517472267151,
      "mean_token_accuracy": 0.9976131021976471,
      "num_tokens": 76846962.0,
      "step": 479
    },
    {
      "aux_loss": 8.027761459350586,
      "entropy": 0.023294473066926003,
      "epoch": 11.433734939759036,
      "grad_norm": 0.2890002429485321,
      "learning_rate": 1.3635890256886501e-06,
      "loss": 0.07428885996341705,
      "mean_token_accuracy": 0.9982883036136627,
      "num_tokens": 77020669.0,
      "step": 480
    },
    {
      "aux_loss": 8.027469635009766,
      "entropy": 0.028372498229146004,
      "epoch": 11.457831325301205,
      "grad_norm": 0.3321754038333893,
      "learning_rate": 1.3632356130638528e-06,
      "loss": 0.07671969383955002,
      "mean_token_accuracy": 0.9977000951766968,
      "num_tokens": 77181204.0,
      "step": 481
    },
    {
      "aux_loss": 8.027999877929688,
      "entropy": 0.024836488999426365,
      "epoch": 11.481927710843374,
      "grad_norm": 0.2997661232948303,
      "learning_rate": 1.3628815774219754e-06,
      "loss": 0.07534407079219818,
      "mean_token_accuracy": 0.9979956150054932,
      "num_tokens": 77365065.0,
      "step": 482
    },
    {
      "aux_loss": 8.027578830718994,
      "entropy": 0.025402785278856754,
      "epoch": 11.506024096385541,
      "grad_norm": 0.3166537880897522,
      "learning_rate": 1.3625269195629517e-06,
      "loss": 0.07534520328044891,
      "mean_token_accuracy": 0.9980738162994385,
      "num_tokens": 77534153.0,
      "step": 483
    },
    {
      "aux_loss": 8.028955936431885,
      "entropy": 0.027941882610321045,
      "epoch": 11.53012048192771,
      "grad_norm": 0.3131483495235443,
      "learning_rate": 1.3621716402881215e-06,
      "loss": 0.0763813927769661,
      "mean_token_accuracy": 0.9981589615345001,
      "num_tokens": 77697477.0,
      "step": 484
    },
    {
      "aux_loss": 8.028358459472656,
      "entropy": 0.023258808068931103,
      "epoch": 11.55421686746988,
      "grad_norm": 0.28977909684181213,
      "learning_rate": 1.3618157404002283e-06,
      "loss": 0.07519376277923584,
      "mean_token_accuracy": 0.9978441596031189,
      "num_tokens": 77884159.0,
      "step": 485
    },
    {
      "aux_loss": 8.027586460113525,
      "entropy": 0.028671267442405224,
      "epoch": 11.578313253012048,
      "grad_norm": 0.3253847658634186,
      "learning_rate": 1.3614592207034186e-06,
      "loss": 0.07729551196098328,
      "mean_token_accuracy": 0.9978532791137695,
      "num_tokens": 78051531.0,
      "step": 486
    },
    {
      "aux_loss": 8.025540828704834,
      "entropy": 0.02651071920990944,
      "epoch": 11.602409638554217,
      "grad_norm": 0.31019577383995056,
      "learning_rate": 1.3611020820032388e-06,
      "loss": 0.07641060650348663,
      "mean_token_accuracy": 0.9979585409164429,
      "num_tokens": 78225093.0,
      "step": 487
    },
    {
      "aux_loss": 8.02729320526123,
      "entropy": 0.029389088042080402,
      "epoch": 11.626506024096386,
      "grad_norm": 0.3609655499458313,
      "learning_rate": 1.360744325106634e-06,
      "loss": 0.07722356915473938,
      "mean_token_accuracy": 0.9975888133049011,
      "num_tokens": 78382804.0,
      "step": 488
    },
    {
      "aux_loss": 8.027955532073975,
      "entropy": 0.029173430055379868,
      "epoch": 11.650602409638553,
      "grad_norm": 0.34442952275276184,
      "learning_rate": 1.360385950821946e-06,
      "loss": 0.07752852886915207,
      "mean_token_accuracy": 0.9973673522472382,
      "num_tokens": 78541939.0,
      "step": 489
    },
    {
      "aux_loss": 8.026641368865967,
      "entropy": 0.02798728458583355,
      "epoch": 11.674698795180722,
      "grad_norm": 0.34241095185279846,
      "learning_rate": 1.3600269599589116e-06,
      "loss": 0.07710593938827515,
      "mean_token_accuracy": 0.997597724199295,
      "num_tokens": 78699528.0,
      "step": 490
    },
    {
      "aux_loss": 8.02653169631958,
      "entropy": 0.022601360455155373,
      "epoch": 11.698795180722891,
      "grad_norm": 0.35648760199546814,
      "learning_rate": 1.3596673533286612e-06,
      "loss": 0.07467810064554214,
      "mean_token_accuracy": 0.9981017708778381,
      "num_tokens": 78883699.0,
      "step": 491
    },
    {
      "aux_loss": 8.02663803100586,
      "entropy": 0.024978790432214737,
      "epoch": 11.72289156626506,
      "grad_norm": 0.3188750445842743,
      "learning_rate": 1.3593071317437159e-06,
      "loss": 0.07513359189033508,
      "mean_token_accuracy": 0.9980481863021851,
      "num_tokens": 79053318.0,
      "step": 492
    },
    {
      "aux_loss": 8.027148246765137,
      "entropy": 0.025895356200635433,
      "epoch": 11.74698795180723,
      "grad_norm": 0.3157801628112793,
      "learning_rate": 1.3589462960179868e-06,
      "loss": 0.0758017897605896,
      "mean_token_accuracy": 0.9976556599140167,
      "num_tokens": 79210267.0,
      "step": 493
    },
    {
      "aux_loss": 8.026412010192871,
      "entropy": 0.026703675277531147,
      "epoch": 11.771084337349398,
      "grad_norm": 0.29264572262763977,
      "learning_rate": 1.358584846966772e-06,
      "loss": 0.07661619037389755,
      "mean_token_accuracy": 0.9978971779346466,
      "num_tokens": 79382574.0,
      "step": 494
    },
    {
      "aux_loss": 8.02318811416626,
      "entropy": 0.029082953929901123,
      "epoch": 11.795180722891565,
      "grad_norm": 0.3202921748161316,
      "learning_rate": 1.358222785406756e-06,
      "loss": 0.07708336412906647,
      "mean_token_accuracy": 0.9975403249263763,
      "num_tokens": 79539742.0,
      "step": 495
    },
    {
      "aux_loss": 8.028156757354736,
      "entropy": 0.03219949174672365,
      "epoch": 11.819277108433734,
      "grad_norm": 0.347214013338089,
      "learning_rate": 1.3578601121560066e-06,
      "loss": 0.07849562168121338,
      "mean_token_accuracy": 0.9975992739200592,
      "num_tokens": 79715225.0,
      "step": 496
    },
    {
      "aux_loss": 8.02785062789917,
      "entropy": 0.03248244524002075,
      "epoch": 11.843373493975903,
      "grad_norm": 0.36930665373802185,
      "learning_rate": 1.3574968280339748e-06,
      "loss": 0.07867126166820526,
      "mean_token_accuracy": 0.9975761771202087,
      "num_tokens": 79863950.0,
      "step": 497
    },
    {
      "aux_loss": 8.026073455810547,
      "entropy": 0.02557132299989462,
      "epoch": 11.867469879518072,
      "grad_norm": 0.3051406145095825,
      "learning_rate": 1.3571329338614906e-06,
      "loss": 0.07562849670648575,
      "mean_token_accuracy": 0.9980701208114624,
      "num_tokens": 80030537.0,
      "step": 498
    },
    {
      "aux_loss": 8.02685546875,
      "entropy": 0.02724768966436386,
      "epoch": 11.891566265060241,
      "grad_norm": 0.32025477290153503,
      "learning_rate": 1.356768430460763e-06,
      "loss": 0.07616618275642395,
      "mean_token_accuracy": 0.9980640113353729,
      "num_tokens": 80201317.0,
      "step": 499
    },
    {
      "aux_loss": 8.026135921478271,
      "entropy": 0.03000323288142681,
      "epoch": 11.91566265060241,
      "grad_norm": 0.3271973133087158,
      "learning_rate": 1.3564033186553776e-06,
      "loss": 0.07745539397001266,
      "mean_token_accuracy": 0.9974850416183472,
      "num_tokens": 80348643.0,
      "step": 500
    },
    {
      "aux_loss": 8.02552318572998,
      "entropy": 0.027183019556105137,
      "epoch": 11.939759036144578,
      "grad_norm": 0.33340510725975037,
      "learning_rate": 1.3560375992702945e-06,
      "loss": 0.07644230127334595,
      "mean_token_accuracy": 0.9975429177284241,
      "num_tokens": 80511188.0,
      "step": 501
    },
    {
      "aux_loss": 8.02660322189331,
      "entropy": 0.025949115864932537,
      "epoch": 11.963855421686747,
      "grad_norm": 0.30572640895843506,
      "learning_rate": 1.3556712731318467e-06,
      "loss": 0.0758323073387146,
      "mean_token_accuracy": 0.9978821277618408,
      "num_tokens": 80685850.0,
      "step": 502
    },
    {
      "aux_loss": 8.02736520767212,
      "entropy": 0.026593544520437717,
      "epoch": 11.987951807228916,
      "grad_norm": 0.3319381773471832,
      "learning_rate": 1.3553043410677378e-06,
      "loss": 0.07625584304332733,
      "mean_token_accuracy": 0.9979047179222107,
      "num_tokens": 80858074.0,
      "step": 503
    },
    {
      "aux_loss": 8.02574348449707,
      "entropy": 0.026809921488165855,
      "epoch": 12.0,
      "grad_norm": 0.43141719698905945,
      "learning_rate": 1.3549368039070411e-06,
      "loss": 0.04325544089078903,
      "mean_token_accuracy": 0.9976990818977356,
      "num_tokens": 80897556.0,
      "step": 504
    }
  ],
  "logging_steps": 1,
  "max_steps": 2100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.6774226325878604e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}