{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 22.0,
  "eval_steps": 500,
  "global_step": 924,
  "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
    },
    {
      "aux_loss": 8.027431011199951,
      "entropy": 0.028303110972046852,
      "epoch": 12.024096385542169,
      "grad_norm": 0.29158341884613037,
      "learning_rate": 1.354568662480197e-06,
      "loss": 0.07610264420509338,
      "mean_token_accuracy": 0.9983315169811249,
      "num_tokens": 81058588.0,
      "step": 505
    },
    {
      "aux_loss": 8.027827262878418,
      "entropy": 0.029750879853963852,
      "epoch": 12.048192771084338,
      "grad_norm": 0.34390127658843994,
      "learning_rate": 1.3541999176190103e-06,
      "loss": 0.07679297775030136,
      "mean_token_accuracy": 0.9981173276901245,
      "num_tokens": 81218943.0,
      "step": 506
    },
    {
      "aux_loss": 8.025716781616211,
      "entropy": 0.028487034142017365,
      "epoch": 12.072289156626505,
      "grad_norm": 0.32288870215415955,
      "learning_rate": 1.3538305701566505e-06,
      "loss": 0.07639636099338531,
      "mean_token_accuracy": 0.9979084432125092,
      "num_tokens": 81391960.0,
      "step": 507
    },
    {
      "aux_loss": 8.025598049163818,
      "entropy": 0.02569352090358734,
      "epoch": 12.096385542168674,
      "grad_norm": 0.3089020252227783,
      "learning_rate": 1.3534606209276477e-06,
      "loss": 0.07509694248437881,
      "mean_token_accuracy": 0.9983225762844086,
      "num_tokens": 81548788.0,
      "step": 508
    },
    {
      "aux_loss": 8.026646137237549,
      "entropy": 0.023537397384643555,
      "epoch": 12.120481927710843,
      "grad_norm": 0.296945720911026,
      "learning_rate": 1.3530900707678925e-06,
      "loss": 0.07455359399318695,
      "mean_token_accuracy": 0.9982962906360626,
      "num_tokens": 81708198.0,
      "step": 509
    },
    {
      "aux_loss": 8.022956848144531,
      "entropy": 0.028022294864058495,
      "epoch": 12.144578313253012,
      "grad_norm": 0.32837480306625366,
      "learning_rate": 1.3527189205146324e-06,
      "loss": 0.07651332765817642,
      "mean_token_accuracy": 0.997875303030014,
      "num_tokens": 81848561.0,
      "step": 510
    },
    {
      "aux_loss": 8.025831699371338,
      "entropy": 0.026299542747437954,
      "epoch": 12.168674698795181,
      "grad_norm": 0.30512234568595886,
      "learning_rate": 1.3523471710064716e-06,
      "loss": 0.07586384564638138,
      "mean_token_accuracy": 0.9978923797607422,
      "num_tokens": 81994863.0,
      "step": 511
    },
    {
      "aux_loss": 8.024884223937988,
      "entropy": 0.02601549867540598,
      "epoch": 12.19277108433735,
      "grad_norm": 0.3252599537372589,
      "learning_rate": 1.351974823083368e-06,
      "loss": 0.07537113130092621,
      "mean_token_accuracy": 0.998057097196579,
      "num_tokens": 82147211.0,
      "step": 512
    },
    {
      "aux_loss": 8.027286052703857,
      "entropy": 0.02541358768939972,
      "epoch": 12.216867469879517,
      "grad_norm": 0.28130462765693665,
      "learning_rate": 1.3516018775866311e-06,
      "loss": 0.07533696293830872,
      "mean_token_accuracy": 0.9985062777996063,
      "num_tokens": 82296539.0,
      "step": 513
    },
    {
      "aux_loss": 8.023170471191406,
      "entropy": 0.025747119449079037,
      "epoch": 12.240963855421686,
      "grad_norm": 0.3796326518058777,
      "learning_rate": 1.3512283353589214e-06,
      "loss": 0.07526561617851257,
      "mean_token_accuracy": 0.9982715845108032,
      "num_tokens": 82454079.0,
      "step": 514
    },
    {
      "aux_loss": 8.026607990264893,
      "entropy": 0.025363615714013577,
      "epoch": 12.265060240963855,
      "grad_norm": 0.30517780780792236,
      "learning_rate": 1.3508541972442473e-06,
      "loss": 0.07509754598140717,
      "mean_token_accuracy": 0.9983097016811371,
      "num_tokens": 82630292.0,
      "step": 515
    },
    {
      "aux_loss": 8.02615737915039,
      "entropy": 0.02330672275274992,
      "epoch": 12.289156626506024,
      "grad_norm": 0.27616629004478455,
      "learning_rate": 1.3504794640879638e-06,
      "loss": 0.07439641654491425,
      "mean_token_accuracy": 0.9981828331947327,
      "num_tokens": 82818619.0,
      "step": 516
    },
    {
      "aux_loss": 8.027260303497314,
      "entropy": 0.03045511431992054,
      "epoch": 12.313253012048193,
      "grad_norm": 0.36653822660446167,
      "learning_rate": 1.35010413673677e-06,
      "loss": 0.07767168432474136,
      "mean_token_accuracy": 0.9976451992988586,
      "num_tokens": 82986008.0,
      "step": 517
    },
    {
      "aux_loss": 8.025866031646729,
      "entropy": 0.027308608405292034,
      "epoch": 12.337349397590362,
      "grad_norm": 0.47573405504226685,
      "learning_rate": 1.3497282160387075e-06,
      "loss": 0.07590079307556152,
      "mean_token_accuracy": 0.9980502426624298,
      "num_tokens": 83161459.0,
      "step": 518
    },
    {
      "aux_loss": 8.027106285095215,
      "entropy": 0.02594301663339138,
      "epoch": 12.36144578313253,
      "grad_norm": 0.31834688782691956,
      "learning_rate": 1.3493517028431595e-06,
      "loss": 0.07548630237579346,
      "mean_token_accuracy": 0.9980054497718811,
      "num_tokens": 83320022.0,
      "step": 519
    },
    {
      "aux_loss": 8.026435375213623,
      "entropy": 0.023103803396224976,
      "epoch": 12.385542168674698,
      "grad_norm": 0.2995982766151428,
      "learning_rate": 1.348974598000847e-06,
      "loss": 0.07430864870548248,
      "mean_token_accuracy": 0.998340368270874,
      "num_tokens": 83504630.0,
      "step": 520
    },
    {
      "aux_loss": 8.025731563568115,
      "entropy": 0.02603835891932249,
      "epoch": 12.409638554216867,
      "grad_norm": 0.2815133333206177,
      "learning_rate": 1.3485969023638278e-06,
      "loss": 0.07602337002754211,
      "mean_token_accuracy": 0.9980197846889496,
      "num_tokens": 83682302.0,
      "step": 521
    },
    {
      "aux_loss": 8.02298355102539,
      "entropy": 0.026776984333992004,
      "epoch": 12.433734939759036,
      "grad_norm": 0.3041905164718628,
      "learning_rate": 1.3482186167854954e-06,
      "loss": 0.07534859329462051,
      "mean_token_accuracy": 0.9982883036136627,
      "num_tokens": 83829851.0,
      "step": 522
    },
    {
      "aux_loss": 8.025999069213867,
      "entropy": 0.02272285521030426,
      "epoch": 12.457831325301205,
      "grad_norm": 0.39276036620140076,
      "learning_rate": 1.3478397421205751e-06,
      "loss": 0.07404108345508575,
      "mean_token_accuracy": 0.9983120262622833,
      "num_tokens": 84004635.0,
      "step": 523
    },
    {
      "aux_loss": 8.026875972747803,
      "entropy": 0.02525839675217867,
      "epoch": 12.481927710843374,
      "grad_norm": 0.286891907453537,
      "learning_rate": 1.3474602792251244e-06,
      "loss": 0.07579140365123749,
      "mean_token_accuracy": 0.9980730414390564,
      "num_tokens": 84191495.0,
      "step": 524
    },
    {
      "aux_loss": 8.025129318237305,
      "entropy": 0.022243279963731766,
      "epoch": 12.506024096385541,
      "grad_norm": 0.30389881134033203,
      "learning_rate": 1.347080228956529e-06,
      "loss": 0.07404506206512451,
      "mean_token_accuracy": 0.9983067512512207,
      "num_tokens": 84371571.0,
      "step": 525
    },
    {
      "aux_loss": 8.025508403778076,
      "entropy": 0.028252240270376205,
      "epoch": 12.53012048192771,
      "grad_norm": 0.369401216506958,
      "learning_rate": 1.3466995921735024e-06,
      "loss": 0.07609507441520691,
      "mean_token_accuracy": 0.9978316426277161,
      "num_tokens": 84540392.0,
      "step": 526
    },
    {
      "aux_loss": 8.02607536315918,
      "entropy": 0.023751961067318916,
      "epoch": 12.55421686746988,
      "grad_norm": 0.2676929235458374,
      "learning_rate": 1.3463183697360826e-06,
      "loss": 0.07453473657369614,
      "mean_token_accuracy": 0.9983112514019012,
      "num_tokens": 84722188.0,
      "step": 527
    },
    {
      "aux_loss": 8.024996757507324,
      "entropy": 0.024897695519030094,
      "epoch": 12.578313253012048,
      "grad_norm": 0.32287949323654175,
      "learning_rate": 1.3459365625056316e-06,
      "loss": 0.07488568872213364,
      "mean_token_accuracy": 0.9982265532016754,
      "num_tokens": 84885749.0,
      "step": 528
    },
    {
      "aux_loss": 8.024873733520508,
      "entropy": 0.03361297491937876,
      "epoch": 12.602409638554217,
      "grad_norm": 0.3236023187637329,
      "learning_rate": 1.3455541713448322e-06,
      "loss": 0.07850486040115356,
      "mean_token_accuracy": 0.9977066218852997,
      "num_tokens": 85038968.0,
      "step": 529
    },
    {
      "aux_loss": 8.025553226470947,
      "entropy": 0.028920285403728485,
      "epoch": 12.626506024096386,
      "grad_norm": 0.3417403995990753,
      "learning_rate": 1.3451711971176868e-06,
      "loss": 0.07649402320384979,
      "mean_token_accuracy": 0.9976703524589539,
      "num_tokens": 85187038.0,
      "step": 530
    },
    {
      "aux_loss": 8.025936126708984,
      "entropy": 0.023510326631367207,
      "epoch": 12.650602409638553,
      "grad_norm": 0.29556459188461304,
      "learning_rate": 1.3447876406895154e-06,
      "loss": 0.0748981386423111,
      "mean_token_accuracy": 0.9980536699295044,
      "num_tokens": 85367566.0,
      "step": 531
    },
    {
      "aux_loss": 8.023914813995361,
      "entropy": 0.0269571915268898,
      "epoch": 12.674698795180722,
      "grad_norm": 0.32661116123199463,
      "learning_rate": 1.3444035029269533e-06,
      "loss": 0.07619081437587738,
      "mean_token_accuracy": 0.9981045424938202,
      "num_tokens": 85530522.0,
      "step": 532
    },
    {
      "aux_loss": 8.025804996490479,
      "entropy": 0.02588724624365568,
      "epoch": 12.698795180722891,
      "grad_norm": 0.3140789270401001,
      "learning_rate": 1.344018784697949e-06,
      "loss": 0.0757066011428833,
      "mean_token_accuracy": 0.9979479014873505,
      "num_tokens": 85691123.0,
      "step": 533
    },
    {
      "aux_loss": 8.027095317840576,
      "entropy": 0.028189819306135178,
      "epoch": 12.72289156626506,
      "grad_norm": 0.34211185574531555,
      "learning_rate": 1.3436334868717626e-06,
      "loss": 0.07652498781681061,
      "mean_token_accuracy": 0.9978649020195007,
      "num_tokens": 85847441.0,
      "step": 534
    },
    {
      "aux_loss": 8.02366828918457,
      "entropy": 0.030385578982532024,
      "epoch": 12.74698795180723,
      "grad_norm": 0.3527638912200928,
      "learning_rate": 1.3432476103189647e-06,
      "loss": 0.0774821788072586,
      "mean_token_accuracy": 0.9980002343654633,
      "num_tokens": 85989139.0,
      "step": 535
    },
    {
      "aux_loss": 8.026358604431152,
      "entropy": 0.024296218529343605,
      "epoch": 12.771084337349398,
      "grad_norm": 0.3142213225364685,
      "learning_rate": 1.3428611559114323e-06,
      "loss": 0.07496938109397888,
      "mean_token_accuracy": 0.9980889558792114,
      "num_tokens": 86151372.0,
      "step": 536
    },
    {
      "aux_loss": 8.025363445281982,
      "entropy": 0.027813317254185677,
      "epoch": 12.795180722891565,
      "grad_norm": 0.32382503151893616,
      "learning_rate": 1.3424741245223487e-06,
      "loss": 0.0763462707400322,
      "mean_token_accuracy": 0.9977690279483795,
      "num_tokens": 86301367.0,
      "step": 537
    },
    {
      "aux_loss": 8.024674415588379,
      "entropy": 0.026332633569836617,
      "epoch": 12.819277108433734,
      "grad_norm": 0.3284594416618347,
      "learning_rate": 1.3420865170262008e-06,
      "loss": 0.07633200287818909,
      "mean_token_accuracy": 0.9977154731750488,
      "num_tokens": 86450704.0,
      "step": 538
    },
    {
      "aux_loss": 8.024837493896484,
      "entropy": 0.026419474743306637,
      "epoch": 12.843373493975903,
      "grad_norm": 0.3180190920829773,
      "learning_rate": 1.3416983342987772e-06,
      "loss": 0.07557108998298645,
      "mean_token_accuracy": 0.9981383979320526,
      "num_tokens": 86603235.0,
      "step": 539
    },
    {
      "aux_loss": 8.024927616119385,
      "entropy": 0.02800104208290577,
      "epoch": 12.867469879518072,
      "grad_norm": 0.32239335775375366,
      "learning_rate": 1.341309577217166e-06,
      "loss": 0.07647630572319031,
      "mean_token_accuracy": 0.997693806886673,
      "num_tokens": 86768482.0,
      "step": 540
    },
    {
      "aux_loss": 8.024649143218994,
      "entropy": 0.02718318998813629,
      "epoch": 12.891566265060241,
      "grad_norm": 0.3123128116130829,
      "learning_rate": 1.3409202466597532e-06,
      "loss": 0.07626186311244965,
      "mean_token_accuracy": 0.9976713359355927,
      "num_tokens": 86941996.0,
      "step": 541
    },
    {
      "aux_loss": 8.025608539581299,
      "entropy": 0.023810462094843388,
      "epoch": 12.91566265060241,
      "grad_norm": 0.3008756637573242,
      "learning_rate": 1.3405303435062207e-06,
      "loss": 0.07481741905212402,
      "mean_token_accuracy": 0.9984180331230164,
      "num_tokens": 87125542.0,
      "step": 542
    },
    {
      "aux_loss": 8.024901390075684,
      "entropy": 0.02601471357047558,
      "epoch": 12.939759036144578,
      "grad_norm": 0.3155852258205414,
      "learning_rate": 1.340139868637544e-06,
      "loss": 0.07538659870624542,
      "mean_token_accuracy": 0.9985229671001434,
      "num_tokens": 87288460.0,
      "step": 543
    },
    {
      "aux_loss": 8.024972438812256,
      "entropy": 0.022537375800311565,
      "epoch": 12.963855421686747,
      "grad_norm": 0.3109420835971832,
      "learning_rate": 1.3397488229359902e-06,
      "loss": 0.07500609755516052,
      "mean_token_accuracy": 0.9979784190654755,
      "num_tokens": 87451648.0,
      "step": 544
    },
    {
      "aux_loss": 8.025615692138672,
      "entropy": 0.025141560472548008,
      "epoch": 12.987951807228916,
      "grad_norm": 0.3224468529224396,
      "learning_rate": 1.3393572072851165e-06,
      "loss": 0.0755264014005661,
      "mean_token_accuracy": 0.9979694187641144,
      "num_tokens": 87594464.0,
      "step": 545
    },
    {
      "aux_loss": 8.027117729187012,
      "entropy": 0.03544420748949051,
      "epoch": 13.0,
      "grad_norm": 0.41236618161201477,
      "learning_rate": 1.338965022569768e-06,
      "loss": 0.04544805735349655,
      "mean_token_accuracy": 0.9975458383560181,
      "num_tokens": 87639019.0,
      "step": 546
    },
    {
      "aux_loss": 8.024989128112793,
      "entropy": 0.026830715127289295,
      "epoch": 13.024096385542169,
      "grad_norm": 0.2847767472267151,
      "learning_rate": 1.3385722696760749e-06,
      "loss": 0.0754060372710228,
      "mean_token_accuracy": 0.9983295202255249,
      "num_tokens": 87805075.0,
      "step": 547
    },
    {
      "aux_loss": 8.024466037750244,
      "entropy": 0.02858618274331093,
      "epoch": 13.048192771084338,
      "grad_norm": 0.3334623873233795,
      "learning_rate": 1.3381789494914518e-06,
      "loss": 0.07619360089302063,
      "mean_token_accuracy": 0.9982382655143738,
      "num_tokens": 87972930.0,
      "step": 548
    },
    {
      "aux_loss": 8.022853374481201,
      "entropy": 0.024458790197968483,
      "epoch": 13.072289156626505,
      "grad_norm": 0.2856697142124176,
      "learning_rate": 1.337785062904595e-06,
      "loss": 0.07412649691104889,
      "mean_token_accuracy": 0.9987647533416748,
      "num_tokens": 88110730.0,
      "step": 549
    },
    {
      "aux_loss": 8.022730827331543,
      "entropy": 0.02606914658099413,
      "epoch": 13.096385542168674,
      "grad_norm": 0.3089272975921631,
      "learning_rate": 1.3373906108054802e-06,
      "loss": 0.07486039400100708,
      "mean_token_accuracy": 0.9983535408973694,
      "num_tokens": 88252615.0,
      "step": 550
    },
    {
      "aux_loss": 8.026148796081543,
      "entropy": 0.02447528950870037,
      "epoch": 13.120481927710843,
      "grad_norm": 0.2822977900505066,
      "learning_rate": 1.3369955940853614e-06,
      "loss": 0.07445919513702393,
      "mean_token_accuracy": 0.9984354674816132,
      "num_tokens": 88410844.0,
      "step": 551
    },
    {
      "aux_loss": 8.02492618560791,
      "entropy": 0.0269709974527359,
      "epoch": 13.144578313253012,
      "grad_norm": 0.2854020595550537,
      "learning_rate": 1.3366000136367682e-06,
      "loss": 0.0757034569978714,
      "mean_token_accuracy": 0.998249739408493,
      "num_tokens": 88564248.0,
      "step": 552
    },
    {
      "aux_loss": 8.024571895599365,
      "entropy": 0.02222419437021017,
      "epoch": 13.168674698795181,
      "grad_norm": 0.25642162561416626,
      "learning_rate": 1.3362038703535036e-06,
      "loss": 0.07327268272638321,
      "mean_token_accuracy": 0.9988367855548859,
      "num_tokens": 88750778.0,
      "step": 553
    },
    {
      "aux_loss": 8.024845600128174,
      "entropy": 0.02474350295960903,
      "epoch": 13.19277108433735,
      "grad_norm": 0.2963204085826874,
      "learning_rate": 1.3358071651306426e-06,
      "loss": 0.0745597630739212,
      "mean_token_accuracy": 0.9984132945537567,
      "num_tokens": 88919735.0,
      "step": 554
    },
    {
      "aux_loss": 8.02199935913086,
      "entropy": 0.02377183921635151,
      "epoch": 13.216867469879517,
      "grad_norm": 0.2865126132965088,
      "learning_rate": 1.3354098988645298e-06,
      "loss": 0.07410858571529388,
      "mean_token_accuracy": 0.9985485970973969,
      "num_tokens": 89098830.0,
      "step": 555
    },
    {
      "aux_loss": 8.024818897247314,
      "entropy": 0.024852984584867954,
      "epoch": 13.240963855421686,
      "grad_norm": 0.2904917895793915,
      "learning_rate": 1.3350120724527771e-06,
      "loss": 0.07489709556102753,
      "mean_token_accuracy": 0.9982765316963196,
      "num_tokens": 89267886.0,
      "step": 556
    },
    {
      "aux_loss": 8.022951126098633,
      "entropy": 0.02369327563792467,
      "epoch": 13.265060240963855,
      "grad_norm": 0.28303927183151245,
      "learning_rate": 1.334613686794263e-06,
      "loss": 0.07419449090957642,
      "mean_token_accuracy": 0.9987742304801941,
      "num_tokens": 89426149.0,
      "step": 557
    },
    {
      "aux_loss": 8.023364067077637,
      "entropy": 0.029238391667604446,
      "epoch": 13.289156626506024,
      "grad_norm": 0.3138507008552551,
      "learning_rate": 1.3342147427891287e-06,
      "loss": 0.07640309631824493,
      "mean_token_accuracy": 0.9982937872409821,
      "num_tokens": 89568963.0,
      "step": 558
    },
    {
      "aux_loss": 8.02321720123291,
      "entropy": 0.022036599926650524,
      "epoch": 13.313253012048193,
      "grad_norm": 0.30322566628456116,
      "learning_rate": 1.3338152413387773e-06,
      "loss": 0.0736384466290474,
      "mean_token_accuracy": 0.9985252618789673,
      "num_tokens": 89717439.0,
      "step": 559
    },
    {
      "aux_loss": 8.024878978729248,
      "entropy": 0.02454978320747614,
      "epoch": 13.337349397590362,
      "grad_norm": 0.31352540850639343,
      "learning_rate": 1.3334151833458714e-06,
      "loss": 0.07487867027521133,
      "mean_token_accuracy": 0.9984146058559418,
      "num_tokens": 89872747.0,
      "step": 560
    },
    {
      "aux_loss": 8.022562980651855,
      "entropy": 0.02784901112318039,
      "epoch": 13.36144578313253,
      "grad_norm": 0.31067028641700745,
      "learning_rate": 1.3330145697143313e-06,
      "loss": 0.0768236368894577,
      "mean_token_accuracy": 0.9979459643363953,
      "num_tokens": 90034928.0,
      "step": 561
    },
    {
      "aux_loss": 8.023937225341797,
      "entropy": 0.027477607131004333,
      "epoch": 13.385542168674698,
      "grad_norm": 0.30565252900123596,
      "learning_rate": 1.332613401349332e-06,
      "loss": 0.07601162046194077,
      "mean_token_accuracy": 0.9979828000068665,
      "num_tokens": 90191929.0,
      "step": 562
    },
    {
      "aux_loss": 8.022271633148193,
      "entropy": 0.026144892908632755,
      "epoch": 13.409638554216867,
      "grad_norm": 0.35552600026130676,
      "learning_rate": 1.3322116791573032e-06,
      "loss": 0.07517768442630768,
      "mean_token_accuracy": 0.9982477128505707,
      "num_tokens": 90351024.0,
      "step": 563
    },
    {
      "aux_loss": 8.02480411529541,
      "entropy": 0.022527651861310005,
      "epoch": 13.433734939759036,
      "grad_norm": 0.2768544852733612,
      "learning_rate": 1.3318094040459246e-06,
      "loss": 0.07377660274505615,
      "mean_token_accuracy": 0.9984523952007294,
      "num_tokens": 90542541.0,
      "step": 564
    },
    {
      "aux_loss": 8.024526596069336,
      "entropy": 0.022174563258886337,
      "epoch": 13.457831325301205,
      "grad_norm": 0.31696975231170654,
      "learning_rate": 1.331406576924126e-06,
      "loss": 0.07396799325942993,
      "mean_token_accuracy": 0.9982556402683258,
      "num_tokens": 90715821.0,
      "step": 565
    },
    {
      "aux_loss": 8.024879932403564,
      "entropy": 0.025845030322670937,
      "epoch": 13.481927710843374,
      "grad_norm": 0.322178453207016,
      "learning_rate": 1.3310031987020846e-06,
      "loss": 0.0750863328576088,
      "mean_token_accuracy": 0.9982399642467499,
      "num_tokens": 90858351.0,
      "step": 566
    },
    {
      "aux_loss": 8.024662017822266,
      "entropy": 0.025781244970858097,
      "epoch": 13.506024096385541,
      "grad_norm": 0.3323957622051239,
      "learning_rate": 1.3305992702912221e-06,
      "loss": 0.07543926686048508,
      "mean_token_accuracy": 0.9979799687862396,
      "num_tokens": 91020129.0,
      "step": 567
    },
    {
      "aux_loss": 8.024908542633057,
      "entropy": 0.028576262295246124,
      "epoch": 13.53012048192771,
      "grad_norm": 0.3114175498485565,
      "learning_rate": 1.3301947926042036e-06,
      "loss": 0.07623940706253052,
      "mean_token_accuracy": 0.9981194138526917,
      "num_tokens": 91178730.0,
      "step": 568
    },
    {
      "aux_loss": 8.024174690246582,
      "entropy": 0.027286341413855553,
      "epoch": 13.55421686746988,
      "grad_norm": 0.2996421456336975,
      "learning_rate": 1.3297897665549356e-06,
      "loss": 0.0759400948882103,
      "mean_token_accuracy": 0.9981012940406799,
      "num_tokens": 91338932.0,
      "step": 569
    },
    {
      "aux_loss": 8.024128913879395,
      "entropy": 0.027576555497944355,
      "epoch": 13.578313253012048,
      "grad_norm": 0.3370552659034729,
      "learning_rate": 1.3293841930585635e-06,
      "loss": 0.07610613107681274,
      "mean_token_accuracy": 0.9980021715164185,
      "num_tokens": 91475686.0,
      "step": 570
    },
    {
      "aux_loss": 8.023448944091797,
      "entropy": 0.02724074013531208,
      "epoch": 13.602409638554217,
      "grad_norm": 0.2822796702384949,
      "learning_rate": 1.3289780730314693e-06,
      "loss": 0.07561826705932617,
      "mean_token_accuracy": 0.9980600476264954,
      "num_tokens": 91659137.0,
      "step": 571
    },
    {
      "aux_loss": 8.024022102355957,
      "entropy": 0.02411498688161373,
      "epoch": 13.626506024096386,
      "grad_norm": 0.32041260600090027,
      "learning_rate": 1.32857140739127e-06,
      "loss": 0.0743638202548027,
      "mean_token_accuracy": 0.9982563257217407,
      "num_tokens": 91827455.0,
      "step": 572
    },
    {
      "aux_loss": 8.022703170776367,
      "entropy": 0.025913319550454617,
      "epoch": 13.650602409638553,
      "grad_norm": 0.31268373131752014,
      "learning_rate": 1.328164197056816e-06,
      "loss": 0.0751735046505928,
      "mean_token_accuracy": 0.9982975423336029,
      "num_tokens": 92003658.0,
      "step": 573
    },
    {
      "aux_loss": 8.02454948425293,
      "entropy": 0.02658990491181612,
      "epoch": 13.674698795180722,
      "grad_norm": 0.31806057691574097,
      "learning_rate": 1.3277564429481873e-06,
      "loss": 0.07578857988119125,
      "mean_token_accuracy": 0.9981650710105896,
      "num_tokens": 92157556.0,
      "step": 574
    },
    {
      "aux_loss": 8.02476167678833,
      "entropy": 0.023106982000172138,
      "epoch": 13.698795180722891,
      "grad_norm": 0.30248817801475525,
      "learning_rate": 1.3273481459866937e-06,
      "loss": 0.07406168431043625,
      "mean_token_accuracy": 0.9986033737659454,
      "num_tokens": 92330208.0,
      "step": 575
    },
    {
      "aux_loss": 8.025537967681885,
      "entropy": 0.025431593880057335,
      "epoch": 13.72289156626506,
      "grad_norm": 0.29318559169769287,
      "learning_rate": 1.3269393070948708e-06,
      "loss": 0.07514628767967224,
      "mean_token_accuracy": 0.9983897507190704,
      "num_tokens": 92504619.0,
      "step": 576
    },
    {
      "aux_loss": 8.02451229095459,
      "entropy": 0.024779404513537884,
      "epoch": 13.74698795180723,
      "grad_norm": 0.30274254083633423,
      "learning_rate": 1.3265299271964786e-06,
      "loss": 0.0745309591293335,
      "mean_token_accuracy": 0.9984515011310577,
      "num_tokens": 92658774.0,
      "step": 577
    },
    {
      "aux_loss": 8.025212287902832,
      "entropy": 0.02221023663878441,
      "epoch": 13.771084337349398,
      "grad_norm": 0.2912888824939728,
      "learning_rate": 1.3261200072165e-06,
      "loss": 0.0740441381931305,
      "mean_token_accuracy": 0.9983481764793396,
      "num_tokens": 92848734.0,
      "step": 578
    },
    {
      "aux_loss": 8.02445650100708,
      "entropy": 0.023341810330748558,
      "epoch": 13.795180722891565,
      "grad_norm": 0.31299862265586853,
      "learning_rate": 1.3257095480811385e-06,
      "loss": 0.07490390539169312,
      "mean_token_accuracy": 0.9981822967529297,
      "num_tokens": 93035663.0,
      "step": 579
    },
    {
      "aux_loss": 8.0231294631958,
      "entropy": 0.02731005661189556,
      "epoch": 13.819277108433734,
      "grad_norm": 0.31801944971084595,
      "learning_rate": 1.3252985507178148e-06,
      "loss": 0.07540342211723328,
      "mean_token_accuracy": 0.9984270334243774,
      "num_tokens": 93190212.0,
      "step": 580
    },
    {
      "aux_loss": 8.022398948669434,
      "entropy": 0.026008632965385914,
      "epoch": 13.843373493975903,
      "grad_norm": 0.33179038763046265,
      "learning_rate": 1.324887016055166e-06,
      "loss": 0.07522394508123398,
      "mean_token_accuracy": 0.9983549416065216,
      "num_tokens": 93341246.0,
      "step": 581
    },
    {
      "aux_loss": 8.023831844329834,
      "entropy": 0.029645448550581932,
      "epoch": 13.867469879518072,
      "grad_norm": 0.3618362247943878,
      "learning_rate": 1.3244749450230438e-06,
      "loss": 0.0771031528711319,
      "mean_token_accuracy": 0.9976988434791565,
      "num_tokens": 93485629.0,
      "step": 582
    },
    {
      "aux_loss": 8.022218704223633,
      "entropy": 0.02165289968252182,
      "epoch": 13.891566265060241,
      "grad_norm": 0.27375370264053345,
      "learning_rate": 1.324062338552511e-06,
      "loss": 0.07368101179599762,
      "mean_token_accuracy": 0.9984147846698761,
      "num_tokens": 93662529.0,
      "step": 583
    },
    {
      "aux_loss": 8.024479866027832,
      "entropy": 0.02351562399417162,
      "epoch": 13.91566265060241,
      "grad_norm": 0.30303192138671875,
      "learning_rate": 1.3236491975758413e-06,
      "loss": 0.07437251508235931,
      "mean_token_accuracy": 0.9984338581562042,
      "num_tokens": 93848011.0,
      "step": 584
    },
    {
      "aux_loss": 8.022239208221436,
      "entropy": 0.02414259221404791,
      "epoch": 13.939759036144578,
      "grad_norm": 0.2712882161140442,
      "learning_rate": 1.3232355230265152e-06,
      "loss": 0.07462576031684875,
      "mean_token_accuracy": 0.9984138906002045,
      "num_tokens": 94023181.0,
      "step": 585
    },
    {
      "aux_loss": 8.02386999130249,
      "entropy": 0.023252902552485466,
      "epoch": 13.963855421686747,
      "grad_norm": 0.32376915216445923,
      "learning_rate": 1.3228213158392188e-06,
      "loss": 0.07477660477161407,
      "mean_token_accuracy": 0.9978744983673096,
      "num_tokens": 94188080.0,
      "step": 586
    },
    {
      "aux_loss": 8.021694660186768,
      "entropy": 0.028471998870372772,
      "epoch": 13.987951807228916,
      "grad_norm": 0.3617095947265625,
      "learning_rate": 1.322406576949842e-06,
      "loss": 0.07622809708118439,
      "mean_token_accuracy": 0.9979861080646515,
      "num_tokens": 94332756.0,
      "step": 587
    },
    {
      "aux_loss": 8.022180557250977,
      "entropy": 0.023031173273921013,
      "epoch": 14.0,
      "grad_norm": 0.39885711669921875,
      "learning_rate": 1.3219913072954761e-06,
      "loss": 0.04186277836561203,
      "mean_token_accuracy": 0.9979294538497925,
      "num_tokens": 94380482.0,
      "step": 588
    },
    {
      "aux_loss": 8.024760246276855,
      "entropy": 0.02022178377956152,
      "epoch": 14.024096385542169,
      "grad_norm": 0.2387811690568924,
      "learning_rate": 1.3215755078144116e-06,
      "loss": 0.07233056426048279,
      "mean_token_accuracy": 0.9986909627914429,
      "num_tokens": 94562098.0,
      "step": 589
    },
    {
      "aux_loss": 8.022474765777588,
      "entropy": 0.023938612081110477,
      "epoch": 14.048192771084338,
      "grad_norm": 0.27819156646728516,
      "learning_rate": 1.3211591794461356e-06,
      "loss": 0.07419903576374054,
      "mean_token_accuracy": 0.9986561834812164,
      "num_tokens": 94718632.0,
      "step": 590
    },
    {
      "aux_loss": 8.024396419525146,
      "entropy": 0.023845979012548923,
      "epoch": 14.072289156626505,
      "grad_norm": 0.2741614282131195,
      "learning_rate": 1.3207423231313312e-06,
      "loss": 0.07409413903951645,
      "mean_token_accuracy": 0.9986681938171387,
      "num_tokens": 94891903.0,
      "step": 591
    },
    {
      "aux_loss": 8.024489402770996,
      "entropy": 0.02289368212223053,
      "epoch": 14.096385542168674,
      "grad_norm": 0.2634838819503784,
      "learning_rate": 1.3203249398118737e-06,
      "loss": 0.07382728904485703,
      "mean_token_accuracy": 0.9985337257385254,
      "num_tokens": 95061769.0,
      "step": 592
    },
    {
      "aux_loss": 8.023909568786621,
      "entropy": 0.023399224504828453,
      "epoch": 14.120481927710843,
      "grad_norm": 0.25872477889060974,
      "learning_rate": 1.319907030430829e-06,
      "loss": 0.07436823844909668,
      "mean_token_accuracy": 0.9986319541931152,
      "num_tokens": 95226736.0,
      "step": 593
    },
    {
      "aux_loss": 8.021496295928955,
      "entropy": 0.023850570432841778,
      "epoch": 14.144578313253012,
      "grad_norm": 0.3044877350330353,
      "learning_rate": 1.3194885959324523e-06,
      "loss": 0.07421186566352844,
      "mean_token_accuracy": 0.998362272977829,
      "num_tokens": 95377366.0,
      "step": 594
    },
    {
      "aux_loss": 8.021922588348389,
      "entropy": 0.024413741193711758,
      "epoch": 14.168674698795181,
      "grad_norm": 0.28136560320854187,
      "learning_rate": 1.3190696372621847e-06,
      "loss": 0.07438543438911438,
      "mean_token_accuracy": 0.9987072944641113,
      "num_tokens": 95534946.0,
      "step": 595
    },
    {
      "aux_loss": 8.022111892700195,
      "entropy": 0.024447415955364704,
      "epoch": 14.19277108433735,
      "grad_norm": 0.2626717984676361,
      "learning_rate": 1.318650155366652e-06,
      "loss": 0.07457850873470306,
      "mean_token_accuracy": 0.9984690248966217,
      "num_tokens": 95697094.0,
      "step": 596
    },
    {
      "aux_loss": 8.023127555847168,
      "entropy": 0.024961436167359352,
      "epoch": 14.216867469879517,
      "grad_norm": 0.27824899554252625,
      "learning_rate": 1.3182301511936622e-06,
      "loss": 0.07419399917125702,
      "mean_token_accuracy": 0.9986103475093842,
      "num_tokens": 95853061.0,
      "step": 597
    },
    {
      "aux_loss": 8.024140357971191,
      "entropy": 0.026958868838846684,
      "epoch": 14.240963855421686,
      "grad_norm": 0.3049028217792511,
      "learning_rate": 1.317809625692203e-06,
      "loss": 0.07510385662317276,
      "mean_token_accuracy": 0.9983646869659424,
      "num_tokens": 96017500.0,
      "step": 598
    },
    {
      "aux_loss": 8.02279281616211,
      "entropy": 0.024663295596837997,
      "epoch": 14.265060240963855,
      "grad_norm": 0.30349597334861755,
      "learning_rate": 1.3173885798124402e-06,
      "loss": 0.07457980513572693,
      "mean_token_accuracy": 0.9985092580318451,
      "num_tokens": 96175235.0,
      "step": 599
    },
    {
      "aux_loss": 8.022304058074951,
      "entropy": 0.02379010058939457,
      "epoch": 14.289156626506024,
      "grad_norm": 0.26444587111473083,
      "learning_rate": 1.3169670145057156e-06,
      "loss": 0.07390819489955902,
      "mean_token_accuracy": 0.9986671507358551,
      "num_tokens": 96344090.0,
      "step": 600
    },
    {
      "aux_loss": 8.022550582885742,
      "entropy": 0.02393107209354639,
      "epoch": 14.313253012048193,
      "grad_norm": 0.28396496176719666,
      "learning_rate": 1.3165449307245445e-06,
      "loss": 0.07395670562982559,
      "mean_token_accuracy": 0.9986652135848999,
      "num_tokens": 96525865.0,
      "step": 601
    },
    {
      "aux_loss": 8.023728370666504,
      "entropy": 0.028622742742300034,
      "epoch": 14.337349397590362,
      "grad_norm": 0.3755778670310974,
      "learning_rate": 1.3161223294226135e-06,
      "loss": 0.07638661563396454,
      "mean_token_accuracy": 0.9980232417583466,
      "num_tokens": 96662516.0,
      "step": 602
    },
    {
      "aux_loss": 8.023515701293945,
      "entropy": 0.02594769559800625,
      "epoch": 14.36144578313253,
      "grad_norm": 0.30059191584587097,
      "learning_rate": 1.3156992115547792e-06,
      "loss": 0.07463668286800385,
      "mean_token_accuracy": 0.9986335933208466,
      "num_tokens": 96828809.0,
      "step": 603
    },
    {
      "aux_loss": 8.022513389587402,
      "entropy": 0.028040869161486626,
      "epoch": 14.385542168674698,
      "grad_norm": 0.3215217590332031,
      "learning_rate": 1.3152755780770644e-06,
      "loss": 0.07540538161993027,
      "mean_token_accuracy": 0.9985578954219818,
      "num_tokens": 96970896.0,
      "step": 604
    },
    {
      "aux_loss": 8.02444839477539,
      "entropy": 0.026185978204011917,
      "epoch": 14.409638554216867,
      "grad_norm": 0.32708242535591125,
      "learning_rate": 1.3148514299466574e-06,
      "loss": 0.07525956630706787,
      "mean_token_accuracy": 0.9983802437782288,
      "num_tokens": 97128433.0,
      "step": 605
    },
    {
      "aux_loss": 8.022536277770996,
      "entropy": 0.023743015713989735,
      "epoch": 14.433734939759036,
      "grad_norm": 0.2937101125717163,
      "learning_rate": 1.3144267681219092e-06,
      "loss": 0.07399174571037292,
      "mean_token_accuracy": 0.9986098110675812,
      "num_tokens": 97279348.0,
      "step": 606
    },
    {
      "aux_loss": 8.022383689880371,
      "entropy": 0.026632227934896946,
      "epoch": 14.457831325301205,
      "grad_norm": 0.30835458636283875,
      "learning_rate": 1.3140015935623318e-06,
      "loss": 0.07527077198028564,
      "mean_token_accuracy": 0.9982325434684753,
      "num_tokens": 97444407.0,
      "step": 607
    },
    {
      "aux_loss": 8.02312183380127,
      "entropy": 0.024252044036984444,
      "epoch": 14.481927710843374,
      "grad_norm": 0.29159989953041077,
      "learning_rate": 1.3135759072285955e-06,
      "loss": 0.07456320524215698,
      "mean_token_accuracy": 0.9984078705310822,
      "num_tokens": 97624818.0,
      "step": 608
    },
    {
      "aux_loss": 8.022176265716553,
      "entropy": 0.025633632205426693,
      "epoch": 14.506024096385541,
      "grad_norm": 0.2863618731498718,
      "learning_rate": 1.313149710082527e-06,
      "loss": 0.07489114999771118,
      "mean_token_accuracy": 0.9985541701316833,
      "num_tokens": 97788363.0,
      "step": 609
    },
    {
      "aux_loss": 8.022820949554443,
      "entropy": 0.021122743375599384,
      "epoch": 14.53012048192771,
      "grad_norm": 0.26490604877471924,
      "learning_rate": 1.3127230030871068e-06,
      "loss": 0.07331493496894836,
      "mean_token_accuracy": 0.9985091984272003,
      "num_tokens": 97974372.0,
      "step": 610
    },
    {
      "aux_loss": 8.022090911865234,
      "entropy": 0.025382555089890957,
      "epoch": 14.55421686746988,
      "grad_norm": 0.2946323752403259,
      "learning_rate": 1.3122957872064679e-06,
      "loss": 0.07488776743412018,
      "mean_token_accuracy": 0.9984503984451294,
      "num_tokens": 98146312.0,
      "step": 611
    },
    {
      "aux_loss": 8.0248384475708,
      "entropy": 0.023470413871109486,
      "epoch": 14.578313253012048,
      "grad_norm": 0.28668469190597534,
      "learning_rate": 1.3118680634058929e-06,
      "loss": 0.0739784687757492,
      "mean_token_accuracy": 0.9984510242938995,
      "num_tokens": 98322279.0,
      "step": 612
    },
    {
      "aux_loss": 8.021702289581299,
      "entropy": 0.021302196197211742,
      "epoch": 14.602409638554217,
      "grad_norm": 0.27184075117111206,
      "learning_rate": 1.3114398326518119e-06,
      "loss": 0.07294765114784241,
      "mean_token_accuracy": 0.9986790120601654,
      "num_tokens": 98495941.0,
      "step": 613
    },
    {
      "aux_loss": 8.021376609802246,
      "entropy": 0.02515318524092436,
      "epoch": 14.626506024096386,
      "grad_norm": 0.36623698472976685,
      "learning_rate": 1.3110110959118008e-06,
      "loss": 0.07469873130321503,
      "mean_token_accuracy": 0.9984108209609985,
      "num_tokens": 98649028.0,
      "step": 614
    },
    {
      "aux_loss": 8.022244453430176,
      "entropy": 0.022386303171515465,
      "epoch": 14.650602409638553,
      "grad_norm": 0.30380144715309143,
      "learning_rate": 1.3105818541545785e-06,
      "loss": 0.07382786273956299,
      "mean_token_accuracy": 0.9984555244445801,
      "num_tokens": 98820655.0,
      "step": 615
    },
    {
      "aux_loss": 8.022692680358887,
      "entropy": 0.024392325431108475,
      "epoch": 14.674698795180722,
      "grad_norm": 0.32163384556770325,
      "learning_rate": 1.3101521083500047e-06,
      "loss": 0.07459438592195511,
      "mean_token_accuracy": 0.9981200695037842,
      "num_tokens": 98985094.0,
      "step": 616
    },
    {
      "aux_loss": 8.022065162658691,
      "entropy": 0.026209702715277672,
      "epoch": 14.698795180722891,
      "grad_norm": 0.3077583611011505,
      "learning_rate": 1.3097218594690787e-06,
      "loss": 0.07581385970115662,
      "mean_token_accuracy": 0.9981196522712708,
      "num_tokens": 99142817.0,
      "step": 617
    },
    {
      "aux_loss": 8.02006721496582,
      "entropy": 0.023902803659439087,
      "epoch": 14.72289156626506,
      "grad_norm": 0.28286728262901306,
      "learning_rate": 1.309291108483936e-06,
      "loss": 0.07453930377960205,
      "mean_token_accuracy": 0.9983839690685272,
      "num_tokens": 99283960.0,
      "step": 618
    },
    {
      "aux_loss": 8.02124834060669,
      "entropy": 0.024052273482084274,
      "epoch": 14.74698795180723,
      "grad_norm": 0.2964589595794678,
      "learning_rate": 1.3088598563678464e-06,
      "loss": 0.07424813508987427,
      "mean_token_accuracy": 0.9983850717544556,
      "num_tokens": 99428973.0,
      "step": 619
    },
    {
      "aux_loss": 8.022575378417969,
      "entropy": 0.021747206337749958,
      "epoch": 14.771084337349398,
      "grad_norm": 0.3091699779033661,
      "learning_rate": 1.3084281040952126e-06,
      "loss": 0.07359092682600021,
      "mean_token_accuracy": 0.9985521137714386,
      "num_tokens": 99599548.0,
      "step": 620
    },
    {
      "aux_loss": 8.022809505462646,
      "entropy": 0.021415174938738346,
      "epoch": 14.795180722891565,
      "grad_norm": 0.2741931974887848,
      "learning_rate": 1.307995852641567e-06,
      "loss": 0.07326743006706238,
      "mean_token_accuracy": 0.9985740482807159,
      "num_tokens": 99761944.0,
      "step": 621
    },
    {
      "aux_loss": 8.021624088287354,
      "entropy": 0.02542014978826046,
      "epoch": 14.819277108433734,
      "grad_norm": 0.3038715124130249,
      "learning_rate": 1.3075631029835698e-06,
      "loss": 0.07464160025119781,
      "mean_token_accuracy": 0.9984971880912781,
      "num_tokens": 99920090.0,
      "step": 622
    },
    {
      "aux_loss": 8.021050453186035,
      "entropy": 0.02499569673091173,
      "epoch": 14.843373493975903,
      "grad_norm": 0.2812490463256836,
      "learning_rate": 1.307129856099007e-06,
      "loss": 0.07466801255941391,
      "mean_token_accuracy": 0.9985790550708771,
      "num_tokens": 100086789.0,
      "step": 623
    },
    {
      "aux_loss": 8.021025657653809,
      "entropy": 0.026820295490324497,
      "epoch": 14.867469879518072,
      "grad_norm": 0.3224012851715088,
      "learning_rate": 1.3066961129667884e-06,
      "loss": 0.07568350434303284,
      "mean_token_accuracy": 0.9980534017086029,
      "num_tokens": 100240901.0,
      "step": 624
    },
    {
      "aux_loss": 8.021604537963867,
      "entropy": 0.02594233863055706,
      "epoch": 14.891566265060241,
      "grad_norm": 0.3286438286304474,
      "learning_rate": 1.3062618745669444e-06,
      "loss": 0.07503099739551544,
      "mean_token_accuracy": 0.9982061386108398,
      "num_tokens": 100397675.0,
      "step": 625
    },
    {
      "aux_loss": 8.021660804748535,
      "entropy": 0.02323277574032545,
      "epoch": 14.91566265060241,
      "grad_norm": 0.2902336120605469,
      "learning_rate": 1.305827141880625e-06,
      "loss": 0.07387363910675049,
      "mean_token_accuracy": 0.998563826084137,
      "num_tokens": 100571805.0,
      "step": 626
    },
    {
      "aux_loss": 8.023479461669922,
      "entropy": 0.02272103913128376,
      "epoch": 14.939759036144578,
      "grad_norm": 0.2875194847583771,
      "learning_rate": 1.3053919158900968e-06,
      "loss": 0.07392413914203644,
      "mean_token_accuracy": 0.9985417127609253,
      "num_tokens": 100749220.0,
      "step": 627
    },
    {
      "aux_loss": 8.021773338317871,
      "entropy": 0.024398663081228733,
      "epoch": 14.963855421686747,
      "grad_norm": 0.3021429777145386,
      "learning_rate": 1.3049561975787408e-06,
      "loss": 0.07461895793676376,
      "mean_token_accuracy": 0.998176246881485,
      "num_tokens": 100909579.0,
      "step": 628
    },
    {
      "aux_loss": 8.021002292633057,
      "entropy": 0.023792341351509094,
      "epoch": 14.987951807228916,
      "grad_norm": 0.3388311266899109,
      "learning_rate": 1.304519987931051e-06,
      "loss": 0.07436676323413849,
      "mean_token_accuracy": 0.9982630014419556,
      "num_tokens": 101073627.0,
      "step": 629
    },
    {
      "aux_loss": 8.0238037109375,
      "entropy": 0.01941307634115219,
      "epoch": 15.0,
      "grad_norm": 0.39163467288017273,
      "learning_rate": 1.3040832879326307e-06,
      "loss": 0.04034464806318283,
      "mean_token_accuracy": 0.9982436895370483,
      "num_tokens": 101121945.0,
      "step": 630
    },
    {
      "aux_loss": 8.020874977111816,
      "entropy": 0.030838350765407085,
      "epoch": 15.024096385542169,
      "grad_norm": 0.31362706422805786,
      "learning_rate": 1.3036460985701921e-06,
      "loss": 0.07680098712444305,
      "mean_token_accuracy": 0.9982480704784393,
      "num_tokens": 101249459.0,
      "step": 631
    },
    {
      "aux_loss": 8.02276611328125,
      "entropy": 0.02502510603517294,
      "epoch": 15.048192771084338,
      "grad_norm": 0.27360862493515015,
      "learning_rate": 1.3032084208315525e-06,
      "loss": 0.07409708201885223,
      "mean_token_accuracy": 0.9989785850048065,
      "num_tokens": 101401821.0,
      "step": 632
    },
    {
      "aux_loss": 8.021870136260986,
      "entropy": 0.021272542886435986,
      "epoch": 15.072289156626505,
      "grad_norm": 0.25828370451927185,
      "learning_rate": 1.3027702557056325e-06,
      "loss": 0.07290402054786682,
      "mean_token_accuracy": 0.9988965690135956,
      "num_tokens": 101579104.0,
      "step": 633
    },
    {
      "aux_loss": 8.020940780639648,
      "entropy": 0.020017596893012524,
      "epoch": 15.096385542168674,
      "grad_norm": 0.2512258291244507,
      "learning_rate": 1.3023316041824544e-06,
      "loss": 0.07227890193462372,
      "mean_token_accuracy": 0.999037891626358,
      "num_tokens": 101745475.0,
      "step": 634
    },
    {
      "aux_loss": 8.021780967712402,
      "entropy": 0.02389626670628786,
      "epoch": 15.120481927710843,
      "grad_norm": 0.27933046221733093,
      "learning_rate": 1.3018924672531393e-06,
      "loss": 0.07385027408599854,
      "mean_token_accuracy": 0.9987798929214478,
      "num_tokens": 101904997.0,
      "step": 635
    },
    {
      "aux_loss": 8.022858142852783,
      "entropy": 0.022255877032876015,
      "epoch": 15.144578313253012,
      "grad_norm": 0.2804347276687622,
      "learning_rate": 1.3014528459099053e-06,
      "loss": 0.07336017489433289,
      "mean_token_accuracy": 0.998691588640213,
      "num_tokens": 102072594.0,
      "step": 636
    },
    {
      "aux_loss": 8.019988059997559,
      "entropy": 0.02136957924813032,
      "epoch": 15.168674698795181,
      "grad_norm": 0.2652220129966736,
      "learning_rate": 1.3010127411460643e-06,
      "loss": 0.07301878929138184,
      "mean_token_accuracy": 0.9986830353736877,
      "num_tokens": 102237845.0,
      "step": 637
    },
    {
      "aux_loss": 8.020222663879395,
      "entropy": 0.023709292523562908,
      "epoch": 15.19277108433735,
      "grad_norm": 0.37665247917175293,
      "learning_rate": 1.3005721539560217e-06,
      "loss": 0.07381770014762878,
      "mean_token_accuracy": 0.9986214935779572,
      "num_tokens": 102402057.0,
      "step": 638
    },
    {
      "aux_loss": 8.022130966186523,
      "entropy": 0.02243464346975088,
      "epoch": 15.216867469879517,
      "grad_norm": 0.2828420102596283,
      "learning_rate": 1.3001310853352716e-06,
      "loss": 0.0734640508890152,
      "mean_token_accuracy": 0.9986587762832642,
      "num_tokens": 102569638.0,
      "step": 639
    },
    {
      "aux_loss": 8.023382663726807,
      "entropy": 0.023820437490940094,
      "epoch": 15.240963855421686,
      "grad_norm": 0.27643638849258423,
      "learning_rate": 1.2996895362803967e-06,
      "loss": 0.07373812049627304,
      "mean_token_accuracy": 0.9989169836044312,
      "num_tokens": 102738864.0,
      "step": 640
    },
    {
      "aux_loss": 8.021101951599121,
      "entropy": 0.02268280368298292,
      "epoch": 15.265060240963855,
      "grad_norm": 0.27888503670692444,
      "learning_rate": 1.299247507789065e-06,
      "loss": 0.07356101274490356,
      "mean_token_accuracy": 0.9985786378383636,
      "num_tokens": 102903859.0,
      "step": 641
    },
    {
      "aux_loss": 8.02034854888916,
      "entropy": 0.023311128839850426,
      "epoch": 15.289156626506024,
      "grad_norm": 0.30982568860054016,
      "learning_rate": 1.2988050008600278e-06,
      "loss": 0.0734785720705986,
      "mean_token_accuracy": 0.9988106191158295,
      "num_tokens": 103069388.0,
      "step": 642
    },
    {
      "aux_loss": 8.020792484283447,
      "entropy": 0.025052642449736595,
      "epoch": 15.313253012048193,
      "grad_norm": 0.3088907301425934,
      "learning_rate": 1.2983620164931176e-06,
      "loss": 0.07444137334823608,
      "mean_token_accuracy": 0.9985477030277252,
      "num_tokens": 103226610.0,
      "step": 643
    },
    {
      "aux_loss": 8.022226810455322,
      "entropy": 0.02244643308222294,
      "epoch": 15.337349397590362,
      "grad_norm": 0.31514278054237366,
      "learning_rate": 1.2979185556892453e-06,
      "loss": 0.07336429506540298,
      "mean_token_accuracy": 0.9984653294086456,
      "num_tokens": 103408648.0,
      "step": 644
    },
    {
      "aux_loss": 8.022058486938477,
      "entropy": 0.022669089026749134,
      "epoch": 15.36144578313253,
      "grad_norm": 0.27879098057746887,
      "learning_rate": 1.2974746194503983e-06,
      "loss": 0.07349519431591034,
      "mean_token_accuracy": 0.9986022412776947,
      "num_tokens": 103566045.0,
      "step": 645
    },
    {
      "aux_loss": 8.021483898162842,
      "entropy": 0.021356522105634212,
      "epoch": 15.385542168674698,
      "grad_norm": 0.2727099061012268,
      "learning_rate": 1.2970302087796387e-06,
      "loss": 0.07301362603902817,
      "mean_token_accuracy": 0.998765617609024,
      "num_tokens": 103724872.0,
      "step": 646
    },
    {
      "aux_loss": 8.021254539489746,
      "entropy": 0.020334189757704735,
      "epoch": 15.409638554216867,
      "grad_norm": 0.26047971844673157,
      "learning_rate": 1.2965853246811002e-06,
      "loss": 0.07252707332372665,
      "mean_token_accuracy": 0.9989950954914093,
      "num_tokens": 103892586.0,
      "step": 647
    },
    {
      "aux_loss": 8.022690773010254,
      "entropy": 0.023002373054623604,
      "epoch": 15.433734939759036,
      "grad_norm": 0.27840209007263184,
      "learning_rate": 1.296139968159986e-06,
      "loss": 0.07410129904747009,
      "mean_token_accuracy": 0.9984548091888428,
      "num_tokens": 104089061.0,
      "step": 648
    },
    {
      "aux_loss": 8.021388053894043,
      "entropy": 0.0260968254879117,
      "epoch": 15.457831325301205,
      "grad_norm": 0.32532691955566406,
      "learning_rate": 1.295694140222567e-06,
      "loss": 0.07457713037729263,
      "mean_token_accuracy": 0.9986681640148163,
      "num_tokens": 104244702.0,
      "step": 649
    },
    {
      "aux_loss": 8.02051067352295,
      "entropy": 0.02532896865159273,
      "epoch": 15.481927710843374,
      "grad_norm": 0.2929506301879883,
      "learning_rate": 1.2952478418761797e-06,
      "loss": 0.0746280699968338,
      "mean_token_accuracy": 0.9985699653625488,
      "num_tokens": 104394461.0,
      "step": 650
    },
    {
      "aux_loss": 8.020220279693604,
      "entropy": 0.02090471889823675,
      "epoch": 15.506024096385541,
      "grad_norm": 0.2575758993625641,
      "learning_rate": 1.2948010741292229e-06,
      "loss": 0.07256072759628296,
      "mean_token_accuracy": 0.9989967942237854,
      "num_tokens": 104549331.0,
      "step": 651
    },
    {
      "aux_loss": 8.020815372467041,
      "entropy": 0.02374210860580206,
      "epoch": 15.53012048192771,
      "grad_norm": 0.27740997076034546,
      "learning_rate": 1.2943538379911556e-06,
      "loss": 0.07387857139110565,
      "mean_token_accuracy": 0.9988603293895721,
      "num_tokens": 104705482.0,
      "step": 652
    },
    {
      "aux_loss": 8.01929759979248,
      "entropy": 0.022214006632566452,
      "epoch": 15.55421686746988,
      "grad_norm": 0.26949021220207214,
      "learning_rate": 1.293906134472496e-06,
      "loss": 0.07310213893651962,
      "mean_token_accuracy": 0.9987643361091614,
      "num_tokens": 104869500.0,
      "step": 653
    },
    {
      "aux_loss": 8.020097255706787,
      "entropy": 0.02291025221347809,
      "epoch": 15.578313253012048,
      "grad_norm": 0.2881735563278198,
      "learning_rate": 1.293457964584818e-06,
      "loss": 0.07341130077838898,
      "mean_token_accuracy": 0.9985943734645844,
      "num_tokens": 105031708.0,
      "step": 654
    },
    {
      "aux_loss": 8.022075653076172,
      "entropy": 0.026461578905582428,
      "epoch": 15.602409638554217,
      "grad_norm": 0.34265780448913574,
      "learning_rate": 1.2930093293407489e-06,
      "loss": 0.07502055913209915,
      "mean_token_accuracy": 0.9984954297542572,
      "num_tokens": 105186130.0,
      "step": 655
    },
    {
      "aux_loss": 8.02108097076416,
      "entropy": 0.02646957617253065,
      "epoch": 15.626506024096386,
      "grad_norm": 0.30914735794067383,
      "learning_rate": 1.2925602297539683e-06,
      "loss": 0.07509751617908478,
      "mean_token_accuracy": 0.9983176290988922,
      "num_tokens": 105352036.0,
      "step": 656
    },
    {
      "aux_loss": 8.020662784576416,
      "entropy": 0.021242394112050533,
      "epoch": 15.650602409638553,
      "grad_norm": 0.3023134171962738,
      "learning_rate": 1.292110666839204e-06,
      "loss": 0.07290812581777573,
      "mean_token_accuracy": 0.9986430406570435,
      "num_tokens": 105524545.0,
      "step": 657
    },
    {
      "aux_loss": 8.01900053024292,
      "entropy": 0.02397344261407852,
      "epoch": 15.674698795180722,
      "grad_norm": 0.3177374601364136,
      "learning_rate": 1.291660641612231e-06,
      "loss": 0.07397983223199844,
      "mean_token_accuracy": 0.998524397611618,
      "num_tokens": 105672387.0,
      "step": 658
    },
    {
      "aux_loss": 8.020395278930664,
      "entropy": 0.025442958809435368,
      "epoch": 15.698795180722891,
      "grad_norm": 0.3742443919181824,
      "learning_rate": 1.2912101550898697e-06,
      "loss": 0.0751451700925827,
      "mean_token_accuracy": 0.9982341527938843,
      "num_tokens": 105833891.0,
      "step": 659
    },
    {
      "aux_loss": 8.02198600769043,
      "entropy": 0.022907257080078125,
      "epoch": 15.72289156626506,
      "grad_norm": 0.29169169068336487,
      "learning_rate": 1.2907592082899816e-06,
      "loss": 0.0738602802157402,
      "mean_token_accuracy": 0.998700886964798,
      "num_tokens": 106020005.0,
      "step": 660
    },
    {
      "aux_loss": 8.02039098739624,
      "entropy": 0.026243245229125023,
      "epoch": 15.74698795180723,
      "grad_norm": 0.33287695050239563,
      "learning_rate": 1.2903078022314686e-06,
      "loss": 0.07507944852113724,
      "mean_token_accuracy": 0.9982471168041229,
      "num_tokens": 106145420.0,
      "step": 661
    },
    {
      "aux_loss": 8.01774787902832,
      "entropy": 0.02395310066640377,
      "epoch": 15.771084337349398,
      "grad_norm": 0.29184141755104065,
      "learning_rate": 1.2898559379342702e-06,
      "loss": 0.07375696301460266,
      "mean_token_accuracy": 0.9986594617366791,
      "num_tokens": 106294891.0,
      "step": 662
    },
    {
      "aux_loss": 8.022370338439941,
      "entropy": 0.021435990929603577,
      "epoch": 15.795180722891565,
      "grad_norm": 0.26756441593170166,
      "learning_rate": 1.2894036164193619e-06,
      "loss": 0.07307396084070206,
      "mean_token_accuracy": 0.9987272918224335,
      "num_tokens": 106481286.0,
      "step": 663
    },
    {
      "aux_loss": 8.020098686218262,
      "entropy": 0.023615891113877296,
      "epoch": 15.819277108433734,
      "grad_norm": 0.2693254351615906,
      "learning_rate": 1.2889508387087517e-06,
      "loss": 0.07437321543693542,
      "mean_token_accuracy": 0.9984356164932251,
      "num_tokens": 106662412.0,
      "step": 664
    },
    {
      "aux_loss": 8.019869804382324,
      "entropy": 0.02427336946129799,
      "epoch": 15.843373493975903,
      "grad_norm": 0.31732895970344543,
      "learning_rate": 1.288497605825478e-06,
      "loss": 0.07438519597053528,
      "mean_token_accuracy": 0.9983821511268616,
      "num_tokens": 106811595.0,
      "step": 665
    },
    {
      "aux_loss": 8.019139289855957,
      "entropy": 0.024068274535238743,
      "epoch": 15.867469879518072,
      "grad_norm": 0.3149513900279999,
      "learning_rate": 1.2880439187936087e-06,
      "loss": 0.0742831826210022,
      "mean_token_accuracy": 0.9983760118484497,
      "num_tokens": 106979051.0,
      "step": 666
    },
    {
      "aux_loss": 8.02074146270752,
      "entropy": 0.021527115255594254,
      "epoch": 15.891566265060241,
      "grad_norm": 0.3242640495300293,
      "learning_rate": 1.287589778638237e-06,
      "loss": 0.07310071587562561,
      "mean_token_accuracy": 0.998621016740799,
      "num_tokens": 107134402.0,
      "step": 667
    },
    {
      "aux_loss": 8.019129753112793,
      "entropy": 0.022771391086280346,
      "epoch": 15.91566265060241,
      "grad_norm": 0.2968021333217621,
      "learning_rate": 1.2871351863854801e-06,
      "loss": 0.07361795008182526,
      "mean_token_accuracy": 0.9986350834369659,
      "num_tokens": 107297595.0,
      "step": 668
    },
    {
      "aux_loss": 8.020462036132812,
      "entropy": 0.02477428875863552,
      "epoch": 15.939759036144578,
      "grad_norm": 0.2882487177848816,
      "learning_rate": 1.286680143062477e-06,
      "loss": 0.07493428885936737,
      "mean_token_accuracy": 0.9982820153236389,
      "num_tokens": 107464336.0,
      "step": 669
    },
    {
      "aux_loss": 8.02011489868164,
      "entropy": 0.020565422251820564,
      "epoch": 15.963855421686747,
      "grad_norm": 0.28168952465057373,
      "learning_rate": 1.2862246496973851e-06,
      "loss": 0.07310523092746735,
      "mean_token_accuracy": 0.9986270666122437,
      "num_tokens": 107638336.0,
      "step": 670
    },
    {
      "aux_loss": 8.020737648010254,
      "entropy": 0.021479875780642033,
      "epoch": 15.987951807228916,
      "grad_norm": 0.29684916138648987,
      "learning_rate": 1.2857687073193798e-06,
      "loss": 0.07347613573074341,
      "mean_token_accuracy": 0.9984760880470276,
      "num_tokens": 107826080.0,
      "step": 671
    },
    {
      "aux_loss": 8.01843547821045,
      "entropy": 0.02332298830151558,
      "epoch": 16.0,
      "grad_norm": 0.43242138624191284,
      "learning_rate": 1.2853123169586503e-06,
      "loss": 0.043614618480205536,
      "mean_token_accuracy": 0.9981607794761658,
      "num_tokens": 107863408.0,
      "step": 672
    },
    {
      "aux_loss": 8.020530700683594,
      "entropy": 0.022393235005438328,
      "epoch": 16.02409638554217,
      "grad_norm": 0.2633357346057892,
      "learning_rate": 1.2848554796463979e-06,
      "loss": 0.0727771669626236,
      "mean_token_accuracy": 0.999247670173645,
      "num_tokens": 108050327.0,
      "step": 673
    },
    {
      "aux_loss": 8.02012825012207,
      "entropy": 0.022030208259820938,
      "epoch": 16.048192771084338,
      "grad_norm": 0.2506941258907318,
      "learning_rate": 1.2843981964148343e-06,
      "loss": 0.0728757455945015,
      "mean_token_accuracy": 0.9989919066429138,
      "num_tokens": 108217426.0,
      "step": 674
    },
    {
      "aux_loss": 8.021928787231445,
      "entropy": 0.023885431699454784,
      "epoch": 16.072289156626507,
      "grad_norm": 0.26206958293914795,
      "learning_rate": 1.2839404682971785e-06,
      "loss": 0.07347150146961212,
      "mean_token_accuracy": 0.9989467859268188,
      "num_tokens": 108392675.0,
      "step": 675
    },
    {
      "aux_loss": 8.020271301269531,
      "entropy": 0.022960077971220016,
      "epoch": 16.096385542168676,
      "grad_norm": 0.2646230161190033,
      "learning_rate": 1.2834822963276542e-06,
      "loss": 0.07312756776809692,
      "mean_token_accuracy": 0.9989570677280426,
      "num_tokens": 108553072.0,
      "step": 676
    },
    {
      "aux_loss": 8.019484043121338,
      "entropy": 0.020488368347287178,
      "epoch": 16.120481927710845,
      "grad_norm": 0.24604511260986328,
      "learning_rate": 1.283023681541489e-06,
      "loss": 0.0727168396115303,
      "mean_token_accuracy": 0.9987335801124573,
      "num_tokens": 108719154.0,
      "step": 677
    },
    {
      "aux_loss": 8.020560264587402,
      "entropy": 0.019702761434018612,
      "epoch": 16.14457831325301,
      "grad_norm": 0.2394089698791504,
      "learning_rate": 1.28256462497491e-06,
      "loss": 0.07267318665981293,
      "mean_token_accuracy": 0.9985304176807404,
      "num_tokens": 108904674.0,
      "step": 678
    },
    {
      "aux_loss": 8.020466327667236,
      "entropy": 0.020342765375971794,
      "epoch": 16.16867469879518,
      "grad_norm": 0.2705174684524536,
      "learning_rate": 1.2821051276651435e-06,
      "loss": 0.0721237063407898,
      "mean_token_accuracy": 0.998966246843338,
      "num_tokens": 109070935.0,
      "step": 679
    },
    {
      "aux_loss": 8.018497467041016,
      "entropy": 0.020449550822377205,
      "epoch": 16.19277108433735,
      "grad_norm": 0.2680540084838867,
      "learning_rate": 1.2816451906504105e-06,
      "loss": 0.07245158404111862,
      "mean_token_accuracy": 0.9988517463207245,
      "num_tokens": 109234529.0,
      "step": 680
    },
    {
      "aux_loss": 8.021220207214355,
      "entropy": 0.020716692321002483,
      "epoch": 16.216867469879517,
      "grad_norm": 0.28465089201927185,
      "learning_rate": 1.2811848149699267e-06,
      "loss": 0.07244610786437988,
      "mean_token_accuracy": 0.9989461600780487,
      "num_tokens": 109392460.0,
      "step": 681
    },
    {
      "aux_loss": 8.018983840942383,
      "entropy": 0.021451476030051708,
      "epoch": 16.240963855421686,
      "grad_norm": 0.2451764941215515,
      "learning_rate": 1.2807240016638979e-06,
      "loss": 0.0723300501704216,
      "mean_token_accuracy": 0.999213457107544,
      "num_tokens": 109554796.0,
      "step": 682
    },
    {
      "aux_loss": 8.019144535064697,
      "entropy": 0.02154405042529106,
      "epoch": 16.265060240963855,
      "grad_norm": 0.25080832839012146,
      "learning_rate": 1.2802627517735193e-06,
      "loss": 0.07240288704633713,
      "mean_token_accuracy": 0.9991842806339264,
      "num_tokens": 109713297.0,
      "step": 683
    },
    {
      "aux_loss": 8.019875526428223,
      "entropy": 0.022928071208298206,
      "epoch": 16.289156626506024,
      "grad_norm": 0.3149115741252899,
      "learning_rate": 1.2798010663409725e-06,
      "loss": 0.07373928278684616,
      "mean_token_accuracy": 0.9986364245414734,
      "num_tokens": 109873842.0,
      "step": 684
    },
    {
      "aux_loss": 8.020755290985107,
      "entropy": 0.019248541444540024,
      "epoch": 16.313253012048193,
      "grad_norm": 0.272822767496109,
      "learning_rate": 1.2793389464094231e-06,
      "loss": 0.07230706512928009,
      "mean_token_accuracy": 0.9988452792167664,
      "num_tokens": 110038995.0,
      "step": 685
    },
    {
      "aux_loss": 8.02060079574585,
      "entropy": 0.02429412119090557,
      "epoch": 16.337349397590362,
      "grad_norm": 0.2750745415687561,
      "learning_rate": 1.2788763930230185e-06,
      "loss": 0.07414013147354126,
      "mean_token_accuracy": 0.9989420175552368,
      "num_tokens": 110200379.0,
      "step": 686
    },
    {
      "aux_loss": 8.019941329956055,
      "entropy": 0.023073025979101658,
      "epoch": 16.36144578313253,
      "grad_norm": 0.28986501693725586,
      "learning_rate": 1.2784134072268853e-06,
      "loss": 0.07339712977409363,
      "mean_token_accuracy": 0.9983747899532318,
      "num_tokens": 110363989.0,
      "step": 687
    },
    {
      "aux_loss": 8.0198073387146,
      "entropy": 0.02075932454317808,
      "epoch": 16.3855421686747,
      "grad_norm": 0.26142221689224243,
      "learning_rate": 1.2779499900671275e-06,
      "loss": 0.07247196137905121,
      "mean_token_accuracy": 0.9989039599895477,
      "num_tokens": 110546477.0,
      "step": 688
    },
    {
      "aux_loss": 8.017828941345215,
      "entropy": 0.023983335122466087,
      "epoch": 16.40963855421687,
      "grad_norm": 0.26134753227233887,
      "learning_rate": 1.2774861425908232e-06,
      "loss": 0.07388327270746231,
      "mean_token_accuracy": 0.9985763430595398,
      "num_tokens": 110705279.0,
      "step": 689
    },
    {
      "aux_loss": 8.01949691772461,
      "entropy": 0.021404901519417763,
      "epoch": 16.433734939759034,
      "grad_norm": 0.29304608702659607,
      "learning_rate": 1.277021865846023e-06,
      "loss": 0.07277315855026245,
      "mean_token_accuracy": 0.9989573061466217,
      "num_tokens": 110872733.0,
      "step": 690
    },
    {
      "aux_loss": 8.019288539886475,
      "entropy": 0.02335297502577305,
      "epoch": 16.457831325301203,
      "grad_norm": 0.2603730857372284,
      "learning_rate": 1.2765571608817482e-06,
      "loss": 0.0734720528125763,
      "mean_token_accuracy": 0.9986677169799805,
      "num_tokens": 111048756.0,
      "step": 691
    },
    {
      "aux_loss": 8.017802238464355,
      "entropy": 0.024003656581044197,
      "epoch": 16.481927710843372,
      "grad_norm": 0.30309340357780457,
      "learning_rate": 1.2760920287479862e-06,
      "loss": 0.07363101094961166,
      "mean_token_accuracy": 0.9985250532627106,
      "num_tokens": 111204600.0,
      "step": 692
    },
    {
      "aux_loss": 8.02037239074707,
      "entropy": 0.019865051843225956,
      "epoch": 16.50602409638554,
      "grad_norm": 0.3331621289253235,
      "learning_rate": 1.2756264704956907e-06,
      "loss": 0.0723302960395813,
      "mean_token_accuracy": 0.9988258481025696,
      "num_tokens": 111383445.0,
      "step": 693
    },
    {
      "aux_loss": 8.020899295806885,
      "entropy": 0.02355680614709854,
      "epoch": 16.53012048192771,
      "grad_norm": 0.30058375000953674,
      "learning_rate": 1.2751604871767777e-06,
      "loss": 0.07387770712375641,
      "mean_token_accuracy": 0.9987203776836395,
      "num_tokens": 111526844.0,
      "step": 694
    },
    {
      "aux_loss": 8.019113063812256,
      "entropy": 0.02117983717471361,
      "epoch": 16.55421686746988,
      "grad_norm": 0.2666833996772766,
      "learning_rate": 1.2746940798441239e-06,
      "loss": 0.07312053442001343,
      "mean_token_accuracy": 0.9987923204898834,
      "num_tokens": 111698598.0,
      "step": 695
    },
    {
      "aux_loss": 8.019593715667725,
      "entropy": 0.023584443144500256,
      "epoch": 16.57831325301205,
      "grad_norm": 0.2818170189857483,
      "learning_rate": 1.2742272495515638e-06,
      "loss": 0.0736173540353775,
      "mean_token_accuracy": 0.9989428520202637,
      "num_tokens": 111854262.0,
      "step": 696
    },
    {
      "aux_loss": 8.018776893615723,
      "entropy": 0.02602121327072382,
      "epoch": 16.602409638554217,
      "grad_norm": 0.2953900396823883,
      "learning_rate": 1.2737599973538879e-06,
      "loss": 0.0743790864944458,
      "mean_token_accuracy": 0.9987208843231201,
      "num_tokens": 112000995.0,
      "step": 697
    },
    {
      "aux_loss": 8.020740032196045,
      "entropy": 0.019654186442494392,
      "epoch": 16.626506024096386,
      "grad_norm": 0.2816968262195587,
      "learning_rate": 1.2732923243068392e-06,
      "loss": 0.07228513062000275,
      "mean_token_accuracy": 0.9988078773021698,
      "num_tokens": 112156280.0,
      "step": 698
    },
    {
      "aux_loss": 8.019818305969238,
      "entropy": 0.02012883499264717,
      "epoch": 16.650602409638555,
      "grad_norm": 0.2908588647842407,
      "learning_rate": 1.272824231467113e-06,
      "loss": 0.07279551029205322,
      "mean_token_accuracy": 0.998571515083313,
      "num_tokens": 112317480.0,
      "step": 699
    },
    {
      "aux_loss": 8.017993450164795,
      "entropy": 0.022014505229890347,
      "epoch": 16.674698795180724,
      "grad_norm": 0.2733386754989624,
      "learning_rate": 1.2723557198923518e-06,
      "loss": 0.07319703698158264,
      "mean_token_accuracy": 0.998718798160553,
      "num_tokens": 112488475.0,
      "step": 700
    },
    {
      "aux_loss": 8.02134895324707,
      "entropy": 0.02212555892765522,
      "epoch": 16.698795180722893,
      "grad_norm": 0.27544257044792175,
      "learning_rate": 1.2718867906411449e-06,
      "loss": 0.0728733241558075,
      "mean_token_accuracy": 0.9989720582962036,
      "num_tokens": 112656711.0,
      "step": 701
    },
    {
      "aux_loss": 8.021170616149902,
      "entropy": 0.02474300842732191,
      "epoch": 16.72289156626506,
      "grad_norm": 0.2849322557449341,
      "learning_rate": 1.2714174447730245e-06,
      "loss": 0.0741940513253212,
      "mean_token_accuracy": 0.9986203908920288,
      "num_tokens": 112833393.0,
      "step": 702
    },
    {
      "aux_loss": 8.018834590911865,
      "entropy": 0.022908312268555164,
      "epoch": 16.746987951807228,
      "grad_norm": 0.28826624155044556,
      "learning_rate": 1.2709476833484658e-06,
      "loss": 0.0734931081533432,
      "mean_token_accuracy": 0.9986154139041901,
      "num_tokens": 112989879.0,
      "step": 703
    },
    {
      "aux_loss": 8.019938468933105,
      "entropy": 0.02371642179787159,
      "epoch": 16.771084337349397,
      "grad_norm": 0.28521883487701416,
      "learning_rate": 1.2704775074288811e-06,
      "loss": 0.07402516901493073,
      "mean_token_accuracy": 0.9986785352230072,
      "num_tokens": 113141308.0,
      "step": 704
    },
    {
      "aux_loss": 8.020092010498047,
      "entropy": 0.024456963874399662,
      "epoch": 16.795180722891565,
      "grad_norm": 0.2931254506111145,
      "learning_rate": 1.27000691807662e-06,
      "loss": 0.0738486722111702,
      "mean_token_accuracy": 0.9987576603889465,
      "num_tokens": 113301783.0,
      "step": 705
    },
    {
      "aux_loss": 8.018565654754639,
      "entropy": 0.0241565415635705,
      "epoch": 16.819277108433734,
      "grad_norm": 0.3001376986503601,
      "learning_rate": 1.2695359163549669e-06,
      "loss": 0.0737600326538086,
      "mean_token_accuracy": 0.9986201226711273,
      "num_tokens": 113477950.0,
      "step": 706
    },
    {
      "aux_loss": 8.01756477355957,
      "entropy": 0.025217771530151367,
      "epoch": 16.843373493975903,
      "grad_norm": 0.3046022355556488,
      "learning_rate": 1.2690645033281367e-06,
      "loss": 0.07499519735574722,
      "mean_token_accuracy": 0.9982928335666656,
      "num_tokens": 113621328.0,
      "step": 707
    },
    {
      "aux_loss": 8.018495559692383,
      "entropy": 0.026443729177117348,
      "epoch": 16.867469879518072,
      "grad_norm": 0.3249662220478058,
      "learning_rate": 1.268592680061275e-06,
      "loss": 0.07490746676921844,
      "mean_token_accuracy": 0.998531699180603,
      "num_tokens": 113768784.0,
      "step": 708
    },
    {
      "aux_loss": 8.018553256988525,
      "entropy": 0.02360593993216753,
      "epoch": 16.89156626506024,
      "grad_norm": 0.29597988724708557,
      "learning_rate": 1.2681204476204526e-06,
      "loss": 0.07375691831111908,
      "mean_token_accuracy": 0.9984996318817139,
      "num_tokens": 113913639.0,
      "step": 709
    },
    {
      "aux_loss": 8.018509864807129,
      "entropy": 0.022534790448844433,
      "epoch": 16.91566265060241,
      "grad_norm": 0.27990180253982544,
      "learning_rate": 1.2676478070726664e-06,
      "loss": 0.0731102004647255,
      "mean_token_accuracy": 0.9988978803157806,
      "num_tokens": 114063897.0,
      "step": 710
    },
    {
      "aux_loss": 8.018209457397461,
      "entropy": 0.02309778332710266,
      "epoch": 16.93975903614458,
      "grad_norm": 0.3171619176864624,
      "learning_rate": 1.2671747594858347e-06,
      "loss": 0.0735432580113411,
      "mean_token_accuracy": 0.9983865022659302,
      "num_tokens": 114211028.0,
      "step": 711
    },
    {
      "aux_loss": 8.020139694213867,
      "entropy": 0.020823744125664234,
      "epoch": 16.96385542168675,
      "grad_norm": 0.27658897638320923,
      "learning_rate": 1.2667013059287958e-06,
      "loss": 0.07257592678070068,
      "mean_token_accuracy": 0.9990012347698212,
      "num_tokens": 114382660.0,
      "step": 712
    },
    {
      "aux_loss": 8.01858901977539,
      "entropy": 0.022462548688054085,
      "epoch": 16.987951807228917,
      "grad_norm": 0.2987523078918457,
      "learning_rate": 1.2662274474713044e-06,
      "loss": 0.07326245307922363,
      "mean_token_accuracy": 0.9986851513385773,
      "num_tokens": 114563054.0,
      "step": 713
    },
    {
      "aux_loss": 8.018234252929688,
      "entropy": 0.022904684767127037,
      "epoch": 17.0,
      "grad_norm": 0.35216137766838074,
      "learning_rate": 1.2657531851840316e-06,
      "loss": 0.04157853499054909,
      "mean_token_accuracy": 0.9983097910881042,
      "num_tokens": 114604871.0,
      "step": 714
    },
    {
      "aux_loss": 8.020260334014893,
      "entropy": 0.01973626110702753,
      "epoch": 17.02409638554217,
      "grad_norm": 0.24988825619220734,
      "learning_rate": 1.2652785201385593e-06,
      "loss": 0.07225106656551361,
      "mean_token_accuracy": 0.9988470375537872,
      "num_tokens": 114781200.0,
      "step": 715
    },
    {
      "aux_loss": 8.018704891204834,
      "entropy": 0.021086208522319794,
      "epoch": 17.048192771084338,
      "grad_norm": 0.2632339298725128,
      "learning_rate": 1.2648034534073808e-06,
      "loss": 0.07266698777675629,
      "mean_token_accuracy": 0.9989010393619537,
      "num_tokens": 114951789.0,
      "step": 716
    },
    {
      "aux_loss": 8.01948356628418,
      "entropy": 0.020533667877316475,
      "epoch": 17.072289156626507,
      "grad_norm": 0.26556897163391113,
      "learning_rate": 1.264327986063896e-06,
      "loss": 0.07211188971996307,
      "mean_token_accuracy": 0.9989912211894989,
      "num_tokens": 115127254.0,
      "step": 717
    },
    {
      "aux_loss": 8.018863677978516,
      "entropy": 0.018902011215686798,
      "epoch": 17.096385542168676,
      "grad_norm": 0.24564693868160248,
      "learning_rate": 1.2638521191824104e-06,
      "loss": 0.07158875465393066,
      "mean_token_accuracy": 0.9990418553352356,
      "num_tokens": 115292268.0,
      "step": 718
    },
    {
      "aux_loss": 8.018441200256348,
      "entropy": 0.023904663510620594,
      "epoch": 17.120481927710845,
      "grad_norm": 0.29119113087654114,
      "learning_rate": 1.2633758538381323e-06,
      "loss": 0.07388065755367279,
      "mean_token_accuracy": 0.9987189471721649,
      "num_tokens": 115441718.0,
      "step": 719
    },
    {
      "aux_loss": 8.018760681152344,
      "entropy": 0.022885017096996307,
      "epoch": 17.14457831325301,
      "grad_norm": 0.2819487750530243,
      "learning_rate": 1.2628991911071704e-06,
      "loss": 0.07310710102319717,
      "mean_token_accuracy": 0.9988558888435364,
      "num_tokens": 115586091.0,
      "step": 720
    },
    {
      "aux_loss": 8.022488594055176,
      "entropy": 0.020207397639751434,
      "epoch": 17.16867469879518,
      "grad_norm": 0.27069467306137085,
      "learning_rate": 1.2624221320665308e-06,
      "loss": 0.0723872259259224,
      "mean_token_accuracy": 0.9989005327224731,
      "num_tokens": 115775635.0,
      "step": 721
    },
    {
      "aux_loss": 8.018838882446289,
      "entropy": 0.020961137488484383,
      "epoch": 17.19277108433735,
      "grad_norm": 0.24343232810497284,
      "learning_rate": 1.2619446777941157e-06,
      "loss": 0.07253874838352203,
      "mean_token_accuracy": 0.9989787936210632,
      "num_tokens": 115934942.0,
      "step": 722
    },
    {
      "aux_loss": 8.01854419708252,
      "entropy": 0.024558238685131073,
      "epoch": 17.216867469879517,
      "grad_norm": 0.2664203941822052,
      "learning_rate": 1.2614668293687195e-06,
      "loss": 0.07387381047010422,
      "mean_token_accuracy": 0.9987564980983734,
      "num_tokens": 116082257.0,
      "step": 723
    },
    {
      "aux_loss": 8.017542839050293,
      "entropy": 0.021965456195175648,
      "epoch": 17.240963855421686,
      "grad_norm": 0.2544980049133301,
      "learning_rate": 1.260988587870028e-06,
      "loss": 0.07254408299922943,
      "mean_token_accuracy": 0.9989876449108124,
      "num_tokens": 116241709.0,
      "step": 724
    },
    {
      "aux_loss": 8.017529964447021,
      "entropy": 0.02150173019617796,
      "epoch": 17.265060240963855,
      "grad_norm": 0.273670494556427,
      "learning_rate": 1.2605099543786148e-06,
      "loss": 0.07259450107812881,
      "mean_token_accuracy": 0.9989366829395294,
      "num_tokens": 116395455.0,
      "step": 725
    },
    {
      "aux_loss": 8.018130779266357,
      "entropy": 0.020396078005433083,
      "epoch": 17.289156626506024,
      "grad_norm": 0.25444889068603516,
      "learning_rate": 1.2600309299759394e-06,
      "loss": 0.07241903990507126,
      "mean_token_accuracy": 0.9989669620990753,
      "num_tokens": 116554317.0,
      "step": 726
    },
    {
      "aux_loss": 8.019949436187744,
      "entropy": 0.025264563970267773,
      "epoch": 17.313253012048193,
      "grad_norm": 0.2933806777000427,
      "learning_rate": 1.2595515157443437e-06,
      "loss": 0.07410993427038193,
      "mean_token_accuracy": 0.9988081157207489,
      "num_tokens": 116718223.0,
      "step": 727
    },
    {
      "aux_loss": 8.01746654510498,
      "entropy": 0.02322079986333847,
      "epoch": 17.337349397590362,
      "grad_norm": 0.28038111329078674,
      "learning_rate": 1.2590717127670515e-06,
      "loss": 0.07366527616977692,
      "mean_token_accuracy": 0.9986792206764221,
      "num_tokens": 116878403.0,
      "step": 728
    },
    {
      "aux_loss": 8.018659591674805,
      "entropy": 0.02247860375791788,
      "epoch": 17.36144578313253,
      "grad_norm": 0.30654484033584595,
      "learning_rate": 1.2585915221281644e-06,
      "loss": 0.0729723572731018,
      "mean_token_accuracy": 0.9987496137619019,
      "num_tokens": 117024804.0,
      "step": 729
    },
    {
      "aux_loss": 8.016965389251709,
      "entropy": 0.021469758823513985,
      "epoch": 17.3855421686747,
      "grad_norm": 0.2939203977584839,
      "learning_rate": 1.25811094491266e-06,
      "loss": 0.07287987321615219,
      "mean_token_accuracy": 0.9987414479255676,
      "num_tokens": 117176747.0,
      "step": 730
    },
    {
      "aux_loss": 8.018593311309814,
      "entropy": 0.018805145286023617,
      "epoch": 17.40963855421687,
      "grad_norm": 0.268107533454895,
      "learning_rate": 1.2576299822063895e-06,
      "loss": 0.07208384573459625,
      "mean_token_accuracy": 0.9986215829849243,
      "num_tokens": 117343678.0,
      "step": 731
    },
    {
      "aux_loss": 8.018801212310791,
      "entropy": 0.024666352197527885,
      "epoch": 17.433734939759034,
      "grad_norm": 0.31131765246391296,
      "learning_rate": 1.2571486350960748e-06,
      "loss": 0.07412542402744293,
      "mean_token_accuracy": 0.9986770153045654,
      "num_tokens": 117509421.0,
      "step": 732
    },
    {
      "aux_loss": 8.018770694732666,
      "entropy": 0.02042167354375124,
      "epoch": 17.457831325301203,
      "grad_norm": 0.2588518559932709,
      "learning_rate": 1.2566669046693068e-06,
      "loss": 0.07215803861618042,
      "mean_token_accuracy": 0.9989665746688843,
      "num_tokens": 117696015.0,
      "step": 733
    },
    {
      "aux_loss": 8.01615571975708,
      "entropy": 0.022597020491957664,
      "epoch": 17.481927710843372,
      "grad_norm": 0.30011555552482605,
      "learning_rate": 1.256184792014542e-06,
      "loss": 0.07340023666620255,
      "mean_token_accuracy": 0.9987127184867859,
      "num_tokens": 117841809.0,
      "step": 734
    },
    {
      "aux_loss": 8.019785404205322,
      "entropy": 0.02277126256376505,
      "epoch": 17.50602409638554,
      "grad_norm": 0.2965037226676941,
      "learning_rate": 1.255702298221101e-06,
      "loss": 0.0731470137834549,
      "mean_token_accuracy": 0.9989103674888611,
      "num_tokens": 118012997.0,
      "step": 735
    },
    {
      "aux_loss": 8.017552852630615,
      "entropy": 0.021831008605659008,
      "epoch": 17.53012048192771,
      "grad_norm": 0.308307945728302,
      "learning_rate": 1.255219424379165e-06,
      "loss": 0.07292138785123825,
      "mean_token_accuracy": 0.9988667964935303,
      "num_tokens": 118171296.0,
      "step": 736
    },
    {
      "aux_loss": 8.018458366394043,
      "entropy": 0.021706807427108288,
      "epoch": 17.55421686746988,
      "grad_norm": 0.26562947034835815,
      "learning_rate": 1.2547361715797744e-06,
      "loss": 0.07324767857789993,
      "mean_token_accuracy": 0.9985917508602142,
      "num_tokens": 118343766.0,
      "step": 737
    },
    {
      "aux_loss": 8.017994403839111,
      "entropy": 0.02541373111307621,
      "epoch": 17.57831325301205,
      "grad_norm": 0.4492263197898865,
      "learning_rate": 1.2542525409148257e-06,
      "loss": 0.07444741576910019,
      "mean_token_accuracy": 0.9987869262695312,
      "num_tokens": 118507347.0,
      "step": 738
    },
    {
      "aux_loss": 8.019275188446045,
      "entropy": 0.022211837582290173,
      "epoch": 17.602409638554217,
      "grad_norm": 0.2794819176197052,
      "learning_rate": 1.2537685334770693e-06,
      "loss": 0.07303763926029205,
      "mean_token_accuracy": 0.9987471699714661,
      "num_tokens": 118682794.0,
      "step": 739
    },
    {
      "aux_loss": 8.019274711608887,
      "entropy": 0.01959415804594755,
      "epoch": 17.626506024096386,
      "grad_norm": 0.2540050446987152,
      "learning_rate": 1.2532841503601068e-06,
      "loss": 0.07219386100769043,
      "mean_token_accuracy": 0.9988141357898712,
      "num_tokens": 118859309.0,
      "step": 740
    },
    {
      "aux_loss": 8.019009590148926,
      "entropy": 0.01926138438284397,
      "epoch": 17.650602409638555,
      "grad_norm": 0.28901535272598267,
      "learning_rate": 1.2527993926583883e-06,
      "loss": 0.07196243852376938,
      "mean_token_accuracy": 0.9989327490329742,
      "num_tokens": 119035893.0,
      "step": 741
    },
    {
      "aux_loss": 8.017775058746338,
      "entropy": 0.022144275717437267,
      "epoch": 17.674698795180724,
      "grad_norm": 0.2637416124343872,
      "learning_rate": 1.2523142614672107e-06,
      "loss": 0.07286165654659271,
      "mean_token_accuracy": 0.9988670945167542,
      "num_tokens": 119180158.0,
      "step": 742
    },
    {
      "aux_loss": 8.017905712127686,
      "entropy": 0.020680655725300312,
      "epoch": 17.698795180722893,
      "grad_norm": 0.27268412709236145,
      "learning_rate": 1.251828757882715e-06,
      "loss": 0.07280862331390381,
      "mean_token_accuracy": 0.9987437427043915,
      "num_tokens": 119321045.0,
      "step": 743
    },
    {
      "aux_loss": 8.019142150878906,
      "entropy": 0.022346139885485172,
      "epoch": 17.72289156626506,
      "grad_norm": 0.2602423131465912,
      "learning_rate": 1.2513428830018823e-06,
      "loss": 0.07306060194969177,
      "mean_token_accuracy": 0.9989902079105377,
      "num_tokens": 119496030.0,
      "step": 744
    },
    {
      "aux_loss": 8.018294334411621,
      "entropy": 0.01832838449627161,
      "epoch": 17.746987951807228,
      "grad_norm": 0.2646837532520294,
      "learning_rate": 1.2508566379225348e-06,
      "loss": 0.07142096012830734,
      "mean_token_accuracy": 0.9991815686225891,
      "num_tokens": 119678297.0,
      "step": 745
    },
    {
      "aux_loss": 8.018517971038818,
      "entropy": 0.021844751201570034,
      "epoch": 17.771084337349397,
      "grad_norm": 0.291924387216568,
      "learning_rate": 1.2503700237433296e-06,
      "loss": 0.0728534385561943,
      "mean_token_accuracy": 0.9987517893314362,
      "num_tokens": 119840233.0,
      "step": 746
    },
    {
      "aux_loss": 8.018216133117676,
      "entropy": 0.020785548724234104,
      "epoch": 17.795180722891565,
      "grad_norm": 0.28122949600219727,
      "learning_rate": 1.2498830415637577e-06,
      "loss": 0.07247965782880783,
      "mean_token_accuracy": 0.9988265633583069,
      "num_tokens": 120003358.0,
      "step": 747
    },
    {
      "aux_loss": 8.018074035644531,
      "entropy": 0.021714145317673683,
      "epoch": 17.819277108433734,
      "grad_norm": 0.2694195508956909,
      "learning_rate": 1.2493956924841425e-06,
      "loss": 0.07262789458036423,
      "mean_token_accuracy": 0.9990041851997375,
      "num_tokens": 120160933.0,
      "step": 748
    },
    {
      "aux_loss": 8.019499778747559,
      "entropy": 0.02373611554503441,
      "epoch": 17.843373493975903,
      "grad_norm": 0.2960670292377472,
      "learning_rate": 1.2489079776056356e-06,
      "loss": 0.07344628870487213,
      "mean_token_accuracy": 0.9987205564975739,
      "num_tokens": 120325227.0,
      "step": 749
    },
    {
      "aux_loss": 8.018202304840088,
      "entropy": 0.021645985543727875,
      "epoch": 17.867469879518072,
      "grad_norm": 0.28741884231567383,
      "learning_rate": 1.2484198980302157e-06,
      "loss": 0.07262778282165527,
      "mean_token_accuracy": 0.9989192187786102,
      "num_tokens": 120495791.0,
      "step": 750
    },
    {
      "aux_loss": 8.018044471740723,
      "entropy": 0.023350493051111698,
      "epoch": 17.89156626506024,
      "grad_norm": 0.28955328464508057,
      "learning_rate": 1.2479314548606851e-06,
      "loss": 0.0735691487789154,
      "mean_token_accuracy": 0.9987229704856873,
      "num_tokens": 120660773.0,
      "step": 751
    },
    {
      "aux_loss": 8.018717765808105,
      "entropy": 0.02127855271100998,
      "epoch": 17.91566265060241,
      "grad_norm": 0.2683342695236206,
      "learning_rate": 1.2474426492006677e-06,
      "loss": 0.07282005995512009,
      "mean_token_accuracy": 0.9986674189567566,
      "num_tokens": 120826550.0,
      "step": 752
    },
    {
      "aux_loss": 8.017789840698242,
      "entropy": 0.022071994841098785,
      "epoch": 17.93975903614458,
      "grad_norm": 0.29112520813941956,
      "learning_rate": 1.246953482154607e-06,
      "loss": 0.07279111444950104,
      "mean_token_accuracy": 0.9989712834358215,
      "num_tokens": 120979013.0,
      "step": 753
    },
    {
      "aux_loss": 8.016852855682373,
      "entropy": 0.019618223421275616,
      "epoch": 17.96385542168675,
      "grad_norm": 0.2747417986392975,
      "learning_rate": 1.2464639548277618e-06,
      "loss": 0.07161170989274979,
      "mean_token_accuracy": 0.9992184042930603,
      "num_tokens": 121130966.0,
      "step": 754
    },
    {
      "aux_loss": 8.018714904785156,
      "entropy": 0.01958319917321205,
      "epoch": 17.987951807228917,
      "grad_norm": 0.26990944147109985,
      "learning_rate": 1.2459740683262065e-06,
      "loss": 0.07182662934064865,
      "mean_token_accuracy": 0.9990403354167938,
      "num_tokens": 121306353.0,
      "step": 755
    },
    {
      "aux_loss": 8.018830299377441,
      "entropy": 0.01911022700369358,
      "epoch": 18.0,
      "grad_norm": 0.3378460705280304,
      "learning_rate": 1.2454838237568257e-06,
      "loss": 0.03958260267972946,
      "mean_token_accuracy": 0.9991428852081299,
      "num_tokens": 121346334.0,
      "step": 756
    },
    {
      "aux_loss": 8.018142223358154,
      "entropy": 0.01967654377222061,
      "epoch": 18.02409638554217,
      "grad_norm": 0.22108010947704315,
      "learning_rate": 1.2449932222273137e-06,
      "loss": 0.07172244787216187,
      "mean_token_accuracy": 0.9990315437316895,
      "num_tokens": 121523418.0,
      "step": 757
    },
    {
      "aux_loss": 8.017167091369629,
      "entropy": 0.021951429545879364,
      "epoch": 18.048192771084338,
      "grad_norm": 0.27572283148765564,
      "learning_rate": 1.2445022648461716e-06,
      "loss": 0.07266423106193542,
      "mean_token_accuracy": 0.9990328252315521,
      "num_tokens": 121678552.0,
      "step": 758
    },
    {
      "aux_loss": 8.016557693481445,
      "entropy": 0.0203837463632226,
      "epoch": 18.072289156626507,
      "grad_norm": 0.2632099688053131,
      "learning_rate": 1.2440109527227037e-06,
      "loss": 0.07195623219013214,
      "mean_token_accuracy": 0.9990858137607574,
      "num_tokens": 121825520.0,
      "step": 759
    },
    {
      "aux_loss": 8.018312931060791,
      "entropy": 0.020994133315980434,
      "epoch": 18.096385542168676,
      "grad_norm": 0.27051058411598206,
      "learning_rate": 1.2435192869670168e-06,
      "loss": 0.0721668004989624,
      "mean_token_accuracy": 0.9990926384925842,
      "num_tokens": 121986116.0,
      "step": 760
    },
    {
      "aux_loss": 8.017702102661133,
      "entropy": 0.01970060635358095,
      "epoch": 18.120481927710845,
      "grad_norm": 0.23090173304080963,
      "learning_rate": 1.243027268690016e-06,
      "loss": 0.07158677279949188,
      "mean_token_accuracy": 0.9991891384124756,
      "num_tokens": 122156242.0,
      "step": 761
    },
    {
      "aux_loss": 8.019289016723633,
      "entropy": 0.021130764856934547,
      "epoch": 18.14457831325301,
      "grad_norm": 0.2537887692451477,
      "learning_rate": 1.2425348990034026e-06,
      "loss": 0.07236440479755402,
      "mean_token_accuracy": 0.999125063419342,
      "num_tokens": 122340036.0,
      "step": 762
    },
    {
      "aux_loss": 8.018044471740723,
      "entropy": 0.01954676304012537,
      "epoch": 18.16867469879518,
      "grad_norm": 0.2520613372325897,
      "learning_rate": 1.2420421790196733e-06,
      "loss": 0.07214604318141937,
      "mean_token_accuracy": 0.9987333118915558,
      "num_tokens": 122521317.0,
      "step": 763
    },
    {
      "aux_loss": 8.01823902130127,
      "entropy": 0.01993868313729763,
      "epoch": 18.19277108433735,
      "grad_norm": 0.24693681299686432,
      "learning_rate": 1.2415491098521152e-06,
      "loss": 0.07197412103414536,
      "mean_token_accuracy": 0.9989577829837799,
      "num_tokens": 122689596.0,
      "step": 764
    },
    {
      "aux_loss": 8.017821788787842,
      "entropy": 0.019117838703095913,
      "epoch": 18.216867469879517,
      "grad_norm": 0.25819259881973267,
      "learning_rate": 1.2410556926148043e-06,
      "loss": 0.07151195406913757,
      "mean_token_accuracy": 0.9991266429424286,
      "num_tokens": 122848887.0,
      "step": 765
    },
    {
      "aux_loss": 8.017277240753174,
      "entropy": 0.02282050345093012,
      "epoch": 18.240963855421686,
      "grad_norm": 0.3336971700191498,
      "learning_rate": 1.2405619284226036e-06,
      "loss": 0.07297661900520325,
      "mean_token_accuracy": 0.9989360868930817,
      "num_tokens": 123007977.0,
      "step": 766
    },
    {
      "aux_loss": 8.020303726196289,
      "entropy": 0.018516553565859795,
      "epoch": 18.265060240963855,
      "grad_norm": 0.2913930416107178,
      "learning_rate": 1.24006781839116e-06,
      "loss": 0.07139115035533905,
      "mean_token_accuracy": 0.9990445077419281,
      "num_tokens": 123183325.0,
      "step": 767
    },
    {
      "aux_loss": 8.017969131469727,
      "entropy": 0.020907875150442123,
      "epoch": 18.289156626506024,
      "grad_norm": 0.26279398798942566,
      "learning_rate": 1.2395733636369012e-06,
      "loss": 0.0723220556974411,
      "mean_token_accuracy": 0.9988578855991364,
      "num_tokens": 123354501.0,
      "step": 768
    },
    {
      "aux_loss": 8.01951789855957,
      "entropy": 0.019883046858012676,
      "epoch": 18.313253012048193,
      "grad_norm": 0.2431030571460724,
      "learning_rate": 1.2390785652770344e-06,
      "loss": 0.07210540771484375,
      "mean_token_accuracy": 0.9989734888076782,
      "num_tokens": 123522136.0,
      "step": 769
    },
    {
      "aux_loss": 8.018885612487793,
      "entropy": 0.018920441158115864,
      "epoch": 18.337349397590362,
      "grad_norm": 0.25550684332847595,
      "learning_rate": 1.2385834244295428e-06,
      "loss": 0.0715964287519455,
      "mean_token_accuracy": 0.9991505444049835,
      "num_tokens": 123662037.0,
      "step": 770
    },
    {
      "aux_loss": 8.017975330352783,
      "entropy": 0.021694052033126354,
      "epoch": 18.36144578313253,
      "grad_norm": 0.30290576815605164,
      "learning_rate": 1.238087942213184e-06,
      "loss": 0.07307359576225281,
      "mean_token_accuracy": 0.9989704787731171,
      "num_tokens": 123814012.0,
      "step": 771
    },
    {
      "aux_loss": 8.017130851745605,
      "entropy": 0.02111757453531027,
      "epoch": 18.3855421686747,
      "grad_norm": 0.3017498552799225,
      "learning_rate": 1.2375921197474862e-06,
      "loss": 0.07237553596496582,
      "mean_token_accuracy": 0.9989622533321381,
      "num_tokens": 123984333.0,
      "step": 772
    },
    {
      "aux_loss": 8.015630722045898,
      "entropy": 0.02062998153269291,
      "epoch": 18.40963855421687,
      "grad_norm": 0.2949274778366089,
      "learning_rate": 1.2370959581527464e-06,
      "loss": 0.07229379564523697,
      "mean_token_accuracy": 0.9989545941352844,
      "num_tokens": 124155932.0,
      "step": 773
    },
    {
      "aux_loss": 8.016655921936035,
      "entropy": 0.01841769553720951,
      "epoch": 18.433734939759034,
      "grad_norm": 0.25348353385925293,
      "learning_rate": 1.236599458550029e-06,
      "loss": 0.0715688019990921,
      "mean_token_accuracy": 0.9989084601402283,
      "num_tokens": 124320385.0,
      "step": 774
    },
    {
      "aux_loss": 8.01733684539795,
      "entropy": 0.021025179885327816,
      "epoch": 18.457831325301203,
      "grad_norm": 0.27677661180496216,
      "learning_rate": 1.2361026220611607e-06,
      "loss": 0.07285228371620178,
      "mean_token_accuracy": 0.9987456500530243,
      "num_tokens": 124486615.0,
      "step": 775
    },
    {
      "aux_loss": 8.017730236053467,
      "entropy": 0.022088666446506977,
      "epoch": 18.481927710843372,
      "grad_norm": 0.2776980400085449,
      "learning_rate": 1.2356054498087302e-06,
      "loss": 0.07264924049377441,
      "mean_token_accuracy": 0.99912890791893,
      "num_tokens": 124645740.0,
      "step": 776
    },
    {
      "aux_loss": 8.0174560546875,
      "entropy": 0.020048685371875763,
      "epoch": 18.50602409638554,
      "grad_norm": 0.4569571912288666,
      "learning_rate": 1.2351079429160842e-06,
      "loss": 0.07212985306978226,
      "mean_token_accuracy": 0.9989519417285919,
      "num_tokens": 124805207.0,
      "step": 777
    },
    {
      "aux_loss": 8.015850067138672,
      "entropy": 0.022110003978013992,
      "epoch": 18.53012048192771,
      "grad_norm": 0.2736422121524811,
      "learning_rate": 1.2346101025073264e-06,
      "loss": 0.07289471477270126,
      "mean_token_accuracy": 0.9988590180873871,
      "num_tokens": 124971034.0,
      "step": 778
    },
    {
      "aux_loss": 8.016895294189453,
      "entropy": 0.02094773482531309,
      "epoch": 18.55421686746988,
      "grad_norm": 0.27527254819869995,
      "learning_rate": 1.2341119297073133e-06,
      "loss": 0.07241301238536835,
      "mean_token_accuracy": 0.9987749755382538,
      "num_tokens": 125134164.0,
      "step": 779
    },
    {
      "aux_loss": 8.018424987792969,
      "entropy": 0.019557715393602848,
      "epoch": 18.57831325301205,
      "grad_norm": 0.2686154246330261,
      "learning_rate": 1.233613425641653e-06,
      "loss": 0.07223023474216461,
      "mean_token_accuracy": 0.9988672733306885,
      "num_tokens": 125303080.0,
      "step": 780
    },
    {
      "aux_loss": 8.016579627990723,
      "entropy": 0.02069095242768526,
      "epoch": 18.602409638554217,
      "grad_norm": 0.2709297835826874,
      "learning_rate": 1.2331145914367016e-06,
      "loss": 0.07235056161880493,
      "mean_token_accuracy": 0.9991201758384705,
      "num_tokens": 125457326.0,
      "step": 781
    },
    {
      "aux_loss": 8.017444133758545,
      "entropy": 0.019050849601626396,
      "epoch": 18.626506024096386,
      "grad_norm": 0.2603260278701782,
      "learning_rate": 1.2326154282195612e-06,
      "loss": 0.07184787094593048,
      "mean_token_accuracy": 0.9990633726119995,
      "num_tokens": 125623507.0,
      "step": 782
    },
    {
      "aux_loss": 8.016646385192871,
      "entropy": 0.020856465213000774,
      "epoch": 18.650602409638555,
      "grad_norm": 0.26723363995552063,
      "learning_rate": 1.232115937118078e-06,
      "loss": 0.0722731500864029,
      "mean_token_accuracy": 0.9991081357002258,
      "num_tokens": 125794453.0,
      "step": 783
    },
    {
      "aux_loss": 8.016578197479248,
      "entropy": 0.020370217971503735,
      "epoch": 18.674698795180724,
      "grad_norm": 0.24809421598911285,
      "learning_rate": 1.2316161192608378e-06,
      "loss": 0.07225576043128967,
      "mean_token_accuracy": 0.9988513886928558,
      "num_tokens": 125968466.0,
      "step": 784
    },
    {
      "aux_loss": 8.017183780670166,
      "entropy": 0.019782189279794693,
      "epoch": 18.698795180722893,
      "grad_norm": 0.27581024169921875,
      "learning_rate": 1.231115975777166e-06,
      "loss": 0.0719655454158783,
      "mean_token_accuracy": 0.9990843832492828,
      "num_tokens": 126129579.0,
      "step": 785
    },
    {
      "aux_loss": 8.016780853271484,
      "entropy": 0.02061288245022297,
      "epoch": 18.72289156626506,
      "grad_norm": 0.2525731325149536,
      "learning_rate": 1.2306155077971227e-06,
      "loss": 0.07215370237827301,
      "mean_token_accuracy": 0.9992306530475616,
      "num_tokens": 126292719.0,
      "step": 786
    },
    {
      "aux_loss": 8.016438007354736,
      "entropy": 0.021378587931394577,
      "epoch": 18.746987951807228,
      "grad_norm": 0.3300939202308655,
      "learning_rate": 1.2301147164515017e-06,
      "loss": 0.07247316092252731,
      "mean_token_accuracy": 0.9989226460456848,
      "num_tokens": 126458152.0,
      "step": 787
    },
    {
      "aux_loss": 8.017562866210938,
      "entropy": 0.023026008158922195,
      "epoch": 18.771084337349397,
      "grad_norm": 0.29177817702293396,
      "learning_rate": 1.2296136028718276e-06,
      "loss": 0.07325838506221771,
      "mean_token_accuracy": 0.9986927509307861,
      "num_tokens": 126608387.0,
      "step": 788
    },
    {
      "aux_loss": 8.01653003692627,
      "entropy": 0.02121768146753311,
      "epoch": 18.795180722891565,
      "grad_norm": 0.24625295400619507,
      "learning_rate": 1.2291121681903523e-06,
      "loss": 0.07226230949163437,
      "mean_token_accuracy": 0.998975396156311,
      "num_tokens": 126774101.0,
      "step": 789
    },
    {
      "aux_loss": 8.017135620117188,
      "entropy": 0.02047717571258545,
      "epoch": 18.819277108433734,
      "grad_norm": 0.28403371572494507,
      "learning_rate": 1.2286104135400546e-06,
      "loss": 0.07196817547082901,
      "mean_token_accuracy": 0.99922975897789,
      "num_tokens": 126948810.0,
      "step": 790
    },
    {
      "aux_loss": 8.017651081085205,
      "entropy": 0.02271729800850153,
      "epoch": 18.843373493975903,
      "grad_norm": 0.2664547264575958,
      "learning_rate": 1.2281083400546346e-06,
      "loss": 0.07297229021787643,
      "mean_token_accuracy": 0.9989756643772125,
      "num_tokens": 127123705.0,
      "step": 791
    },
    {
      "aux_loss": 8.016648292541504,
      "entropy": 0.022724810987710953,
      "epoch": 18.867469879518072,
      "grad_norm": 0.27334722876548767,
      "learning_rate": 1.227605948868514e-06,
      "loss": 0.07290689647197723,
      "mean_token_accuracy": 0.9990213513374329,
      "num_tokens": 127286211.0,
      "step": 792
    },
    {
      "aux_loss": 8.018507480621338,
      "entropy": 0.02393777295947075,
      "epoch": 18.89156626506024,
      "grad_norm": 0.2772607207298279,
      "learning_rate": 1.227103241116832e-06,
      "loss": 0.07380357384681702,
      "mean_token_accuracy": 0.9988025724887848,
      "num_tokens": 127454841.0,
      "step": 793
    },
    {
      "aux_loss": 8.014709949493408,
      "entropy": 0.022906312718987465,
      "epoch": 18.91566265060241,
      "grad_norm": 0.2939465641975403,
      "learning_rate": 1.2266002179354427e-06,
      "loss": 0.07364203035831451,
      "mean_token_accuracy": 0.9987647533416748,
      "num_tokens": 127601787.0,
      "step": 794
    },
    {
      "aux_loss": 8.014947414398193,
      "entropy": 0.02199286874383688,
      "epoch": 18.93975903614458,
      "grad_norm": 0.3006232976913452,
      "learning_rate": 1.2260968804609137e-06,
      "loss": 0.07279400527477264,
      "mean_token_accuracy": 0.9985716938972473,
      "num_tokens": 127743717.0,
      "step": 795
    },
    {
      "aux_loss": 8.016969680786133,
      "entropy": 0.02244889084249735,
      "epoch": 18.96385542168675,
      "grad_norm": 0.3185865879058838,
      "learning_rate": 1.225593229830522e-06,
      "loss": 0.07299022376537323,
      "mean_token_accuracy": 0.9987007081508636,
      "num_tokens": 127897148.0,
      "step": 796
    },
    {
      "aux_loss": 8.016538143157959,
      "entropy": 0.018228761851787567,
      "epoch": 18.987951807228917,
      "grad_norm": 0.24646823108196259,
      "learning_rate": 1.2250892671822523e-06,
      "loss": 0.07121360301971436,
      "mean_token_accuracy": 0.9989063739776611,
      "num_tokens": 128063518.0,
      "step": 797
    },
    {
      "aux_loss": 8.016475677490234,
      "entropy": 0.029299454763531685,
      "epoch": 19.0,
      "grad_norm": 0.4041454792022705,
      "learning_rate": 1.2245849936547948e-06,
      "loss": 0.040493883192539215,
      "mean_token_accuracy": 0.9988755583763123,
      "num_tokens": 128087797.0,
      "step": 798
    },
    {
      "aux_loss": 8.016082286834717,
      "entropy": 0.01707358844578266,
      "epoch": 19.02409638554217,
      "grad_norm": 0.2209968864917755,
      "learning_rate": 1.224080410387541e-06,
      "loss": 0.07054603844881058,
      "mean_token_accuracy": 0.9994014203548431,
      "num_tokens": 128257323.0,
      "step": 799
    },
    {
      "aux_loss": 8.016879081726074,
      "entropy": 0.020540348254144192,
      "epoch": 19.048192771084338,
      "grad_norm": 0.28073349595069885,
      "learning_rate": 1.223575518520584e-06,
      "loss": 0.07254428416490555,
      "mean_token_accuracy": 0.9990403950214386,
      "num_tokens": 128408675.0,
      "step": 800
    },
    {
      "aux_loss": 8.016739845275879,
      "entropy": 0.0186675563454628,
      "epoch": 19.072289156626507,
      "grad_norm": 0.22932597994804382,
      "learning_rate": 1.2230703191947125e-06,
      "loss": 0.07110397517681122,
      "mean_token_accuracy": 0.9992808401584625,
      "num_tokens": 128576966.0,
      "step": 801
    },
    {
      "aux_loss": 8.014533519744873,
      "entropy": 0.0216646958142519,
      "epoch": 19.096385542168676,
      "grad_norm": 0.24630875885486603,
      "learning_rate": 1.2225648135514107e-06,
      "loss": 0.0723528265953064,
      "mean_token_accuracy": 0.9988449215888977,
      "num_tokens": 128722275.0,
      "step": 802
    },
    {
      "aux_loss": 8.016851902008057,
      "entropy": 0.01763029955327511,
      "epoch": 19.120481927710845,
      "grad_norm": 0.23657096922397614,
      "learning_rate": 1.222059002732855e-06,
      "loss": 0.07095248252153397,
      "mean_token_accuracy": 0.9990513622760773,
      "num_tokens": 128893797.0,
      "step": 803
    },
    {
      "aux_loss": 8.017178535461426,
      "entropy": 0.020476515404880047,
      "epoch": 19.14457831325301,
      "grad_norm": 0.2538764476776123,
      "learning_rate": 1.221552887881911e-06,
      "loss": 0.07223986089229584,
      "mean_token_accuracy": 0.9988578855991364,
      "num_tokens": 129068433.0,
      "step": 804
    },
    {
      "aux_loss": 8.016830921173096,
      "entropy": 0.018701947294175625,
      "epoch": 19.16867469879518,
      "grad_norm": 0.24616143107414246,
      "learning_rate": 1.2210464701421311e-06,
      "loss": 0.0714377760887146,
      "mean_token_accuracy": 0.9991999268531799,
      "num_tokens": 129255686.0,
      "step": 805
    },
    {
      "aux_loss": 8.01766586303711,
      "entropy": 0.02190525457262993,
      "epoch": 19.19277108433735,
      "grad_norm": 0.257795125246048,
      "learning_rate": 1.2205397506577528e-06,
      "loss": 0.07279151678085327,
      "mean_token_accuracy": 0.9989840984344482,
      "num_tokens": 129426600.0,
      "step": 806
    },
    {
      "aux_loss": 8.01670217514038,
      "entropy": 0.022698327898979187,
      "epoch": 19.216867469879517,
      "grad_norm": 0.27151861786842346,
      "learning_rate": 1.2200327305736947e-06,
      "loss": 0.072884202003479,
      "mean_token_accuracy": 0.9989652335643768,
      "num_tokens": 129602609.0,
      "step": 807
    },
    {
      "aux_loss": 8.016786098480225,
      "entropy": 0.020293287932872772,
      "epoch": 19.240963855421686,
      "grad_norm": 0.260490357875824,
      "learning_rate": 1.2195254110355547e-06,
      "loss": 0.07177005708217621,
      "mean_token_accuracy": 0.9990547299385071,
      "num_tokens": 129764750.0,
      "step": 808
    },
    {
      "aux_loss": 8.01784610748291,
      "entropy": 0.02007181290537119,
      "epoch": 19.265060240963855,
      "grad_norm": 0.24022895097732544,
      "learning_rate": 1.2190177931896074e-06,
      "loss": 0.07160848379135132,
      "mean_token_accuracy": 0.9992149472236633,
      "num_tokens": 129933157.0,
      "step": 809
    },
    {
      "aux_loss": 8.018133163452148,
      "entropy": 0.02162711601704359,
      "epoch": 19.289156626506024,
      "grad_norm": 0.30957019329071045,
      "learning_rate": 1.2185098781828017e-06,
      "loss": 0.07218888401985168,
      "mean_token_accuracy": 0.998899906873703,
      "num_tokens": 130082077.0,
      "step": 810
    },
    {
      "aux_loss": 8.016764640808105,
      "entropy": 0.017992699518799782,
      "epoch": 19.313253012048193,
      "grad_norm": 0.23729471862316132,
      "learning_rate": 1.2180016671627576e-06,
      "loss": 0.0710437223315239,
      "mean_token_accuracy": 0.9992759525775909,
      "num_tokens": 130256356.0,
      "step": 811
    },
    {
      "aux_loss": 8.018122673034668,
      "entropy": 0.02003295347094536,
      "epoch": 19.337349397590362,
      "grad_norm": 0.2603780925273895,
      "learning_rate": 1.217493161277764e-06,
      "loss": 0.07190095633268356,
      "mean_token_accuracy": 0.9992459416389465,
      "num_tokens": 130428786.0,
      "step": 812
    },
    {
      "aux_loss": 8.015380382537842,
      "entropy": 0.02138371951878071,
      "epoch": 19.36144578313253,
      "grad_norm": 0.23855821788311005,
      "learning_rate": 1.216984361676776e-06,
      "loss": 0.07250922918319702,
      "mean_token_accuracy": 0.9992147982120514,
      "num_tokens": 130569702.0,
      "step": 813
    },
    {
      "aux_loss": 8.014621257781982,
      "entropy": 0.024525553919374943,
      "epoch": 19.3855421686747,
      "grad_norm": 0.2766070067882538,
      "learning_rate": 1.2164752695094123e-06,
      "loss": 0.0731203556060791,
      "mean_token_accuracy": 0.9991021752357483,
      "num_tokens": 130728155.0,
      "step": 814
    },
    {
      "aux_loss": 8.017411231994629,
      "entropy": 0.01734110154211521,
      "epoch": 19.40963855421687,
      "grad_norm": 0.30324476957321167,
      "learning_rate": 1.215965885925953e-06,
      "loss": 0.07086675614118576,
      "mean_token_accuracy": 0.9991721510887146,
      "num_tokens": 130906575.0,
      "step": 815
    },
    {
      "aux_loss": 8.016311168670654,
      "entropy": 0.020214502699673176,
      "epoch": 19.433734939759034,
      "grad_norm": 0.37634822726249695,
      "learning_rate": 1.2154562120773358e-06,
      "loss": 0.07189343869686127,
      "mean_token_accuracy": 0.9990342259407043,
      "num_tokens": 131067413.0,
      "step": 816
    },
    {
      "aux_loss": 8.015110492706299,
      "entropy": 0.023175804875791073,
      "epoch": 19.457831325301203,
      "grad_norm": 0.3171541690826416,
      "learning_rate": 1.2149462491151556e-06,
      "loss": 0.07334239780902863,
      "mean_token_accuracy": 0.9987165927886963,
      "num_tokens": 131214226.0,
      "step": 817
    },
    {
      "aux_loss": 8.015056610107422,
      "entropy": 0.019055875949561596,
      "epoch": 19.481927710843372,
      "grad_norm": 0.33125078678131104,
      "learning_rate": 1.2144359981916595e-06,
      "loss": 0.07168950885534286,
      "mean_token_accuracy": 0.9990333318710327,
      "num_tokens": 131359757.0,
      "step": 818
    },
    {
      "aux_loss": 8.013885974884033,
      "entropy": 0.02209868934005499,
      "epoch": 19.50602409638554,
      "grad_norm": 0.2984984219074249,
      "learning_rate": 1.2139254604597453e-06,
      "loss": 0.07285237312316895,
      "mean_token_accuracy": 0.9989113807678223,
      "num_tokens": 131496939.0,
      "step": 819
    },
    {
      "aux_loss": 8.015743255615234,
      "entropy": 0.020269363187253475,
      "epoch": 19.53012048192771,
      "grad_norm": 0.24305978417396545,
      "learning_rate": 1.2134146370729593e-06,
      "loss": 0.07186189293861389,
      "mean_token_accuracy": 0.9991940855979919,
      "num_tokens": 131648339.0,
      "step": 820
    },
    {
      "aux_loss": 8.016966819763184,
      "entropy": 0.018671550787985325,
      "epoch": 19.55421686746988,
      "grad_norm": 0.25464141368865967,
      "learning_rate": 1.2129035291854927e-06,
      "loss": 0.07103103399276733,
      "mean_token_accuracy": 0.9992456138134003,
      "num_tokens": 131819044.0,
      "step": 821
    },
    {
      "aux_loss": 8.016483783721924,
      "entropy": 0.0185864744707942,
      "epoch": 19.57831325301205,
      "grad_norm": 0.24460941553115845,
      "learning_rate": 1.2123921379521802e-06,
      "loss": 0.0713551789522171,
      "mean_token_accuracy": 0.9990330636501312,
      "num_tokens": 131988935.0,
      "step": 822
    },
    {
      "aux_loss": 8.014330863952637,
      "entropy": 0.019629787653684616,
      "epoch": 19.602409638554217,
      "grad_norm": 0.27499374747276306,
      "learning_rate": 1.2118804645284957e-06,
      "loss": 0.07154914736747742,
      "mean_token_accuracy": 0.9992263913154602,
      "num_tokens": 132157225.0,
      "step": 823
    },
    {
      "aux_loss": 8.016732692718506,
      "entropy": 0.019671409390866756,
      "epoch": 19.626506024096386,
      "grad_norm": 0.2679067850112915,
      "learning_rate": 1.211368510070552e-06,
      "loss": 0.07195160537958145,
      "mean_token_accuracy": 0.9990942776203156,
      "num_tokens": 132319583.0,
      "step": 824
    },
    {
      "aux_loss": 8.016017436981201,
      "entropy": 0.01897235121577978,
      "epoch": 19.650602409638555,
      "grad_norm": 0.31477922201156616,
      "learning_rate": 1.2108562757350959e-06,
      "loss": 0.07175606489181519,
      "mean_token_accuracy": 0.998927503824234,
      "num_tokens": 132461657.0,
      "step": 825
    },
    {
      "aux_loss": 8.017291069030762,
      "entropy": 0.017452422529459,
      "epoch": 19.674698795180724,
      "grad_norm": 0.23533299565315247,
      "learning_rate": 1.2103437626795066e-06,
      "loss": 0.07113629579544067,
      "mean_token_accuracy": 0.9991660118103027,
      "num_tokens": 132654439.0,
      "step": 826
    },
    {
      "aux_loss": 8.014272689819336,
      "entropy": 0.022786526009440422,
      "epoch": 19.698795180722893,
      "grad_norm": 0.2727932333946228,
      "learning_rate": 1.2098309720617938e-06,
      "loss": 0.07298114150762558,
      "mean_token_accuracy": 0.9988896250724792,
      "num_tokens": 132807186.0,
      "step": 827
    },
    {
      "aux_loss": 8.015275001525879,
      "entropy": 0.02233956567943096,
      "epoch": 19.72289156626506,
      "grad_norm": 0.2817685008049011,
      "learning_rate": 1.2093179050405935e-06,
      "loss": 0.07318785786628723,
      "mean_token_accuracy": 0.998853325843811,
      "num_tokens": 132947861.0,
      "step": 828
    },
    {
      "aux_loss": 8.016317367553711,
      "entropy": 0.01885117031633854,
      "epoch": 19.746987951807228,
      "grad_norm": 0.2575323283672333,
      "learning_rate": 1.208804562775167e-06,
      "loss": 0.07110799103975296,
      "mean_token_accuracy": 0.9993088245391846,
      "num_tokens": 133118610.0,
      "step": 829
    },
    {
      "aux_loss": 8.013425827026367,
      "entropy": 0.020050152204930782,
      "epoch": 19.771084337349397,
      "grad_norm": 0.25590944290161133,
      "learning_rate": 1.2082909464253968e-06,
      "loss": 0.07188734412193298,
      "mean_token_accuracy": 0.9991053342819214,
      "num_tokens": 133265190.0,
      "step": 830
    },
    {
      "aux_loss": 8.018101692199707,
      "entropy": 0.020321492105722427,
      "epoch": 19.795180722891565,
      "grad_norm": 0.2820008397102356,
      "learning_rate": 1.2077770571517851e-06,
      "loss": 0.07242708653211594,
      "mean_token_accuracy": 0.9988300800323486,
      "num_tokens": 133443361.0,
      "step": 831
    },
    {
      "aux_loss": 8.016617774963379,
      "entropy": 0.021538270637392998,
      "epoch": 19.819277108433734,
      "grad_norm": 0.29813796281814575,
      "learning_rate": 1.207262896115451e-06,
      "loss": 0.07250789552927017,
      "mean_token_accuracy": 0.9990063607692719,
      "num_tokens": 133619744.0,
      "step": 832
    },
    {
      "aux_loss": 8.016489028930664,
      "entropy": 0.017469782382249832,
      "epoch": 19.843373493975903,
      "grad_norm": 0.2585788071155548,
      "learning_rate": 1.2067484644781265e-06,
      "loss": 0.0711778849363327,
      "mean_token_accuracy": 0.9991479218006134,
      "num_tokens": 133798588.0,
      "step": 833
    },
    {
      "aux_loss": 8.01574182510376,
      "entropy": 0.020052564330399036,
      "epoch": 19.867469879518072,
      "grad_norm": 0.2874489724636078,
      "learning_rate": 1.2062337634021566e-06,
      "loss": 0.0720338225364685,
      "mean_token_accuracy": 0.99901282787323,
      "num_tokens": 133949653.0,
      "step": 834
    },
    {
      "aux_loss": 8.014959812164307,
      "entropy": 0.020040088333189487,
      "epoch": 19.89156626506024,
      "grad_norm": 0.26376113295555115,
      "learning_rate": 1.2057187940504941e-06,
      "loss": 0.07194425165653229,
      "mean_token_accuracy": 0.9990537166595459,
      "num_tokens": 134113370.0,
      "step": 835
    },
    {
      "aux_loss": 8.015164375305176,
      "entropy": 0.023158026859164238,
      "epoch": 19.91566265060241,
      "grad_norm": 0.2815038859844208,
      "learning_rate": 1.205203557586698e-06,
      "loss": 0.07307106256484985,
      "mean_token_accuracy": 0.9989737272262573,
      "num_tokens": 134286525.0,
      "step": 836
    },
    {
      "aux_loss": 8.017555236816406,
      "entropy": 0.017343491315841675,
      "epoch": 19.93975903614458,
      "grad_norm": 0.2515917420387268,
      "learning_rate": 1.204688055174931e-06,
      "loss": 0.07106645405292511,
      "mean_token_accuracy": 0.9990668594837189,
      "num_tokens": 134455231.0,
      "step": 837
    },
    {
      "aux_loss": 8.015775203704834,
      "entropy": 0.023353048600256443,
      "epoch": 19.96385542168675,
      "grad_norm": 0.29685935378074646,
      "learning_rate": 1.2041722879799568e-06,
      "loss": 0.07338288426399231,
      "mean_token_accuracy": 0.9986701309680939,
      "num_tokens": 134603410.0,
      "step": 838
    },
    {
      "aux_loss": 8.01530408859253,
      "entropy": 0.018707728944718838,
      "epoch": 19.987951807228917,
      "grad_norm": 0.26528769731521606,
      "learning_rate": 1.2036562571671374e-06,
      "loss": 0.0717005729675293,
      "mean_token_accuracy": 0.9990585446357727,
      "num_tokens": 134779729.0,
      "step": 839
    },
    {
      "aux_loss": 8.01617431640625,
      "entropy": 0.0191100612282753,
      "epoch": 20.0,
      "grad_norm": 0.32825803756713867,
      "learning_rate": 1.20313996390243e-06,
      "loss": 0.03942277655005455,
      "mean_token_accuracy": 0.9991744160652161,
      "num_tokens": 134829260.0,
      "step": 840
    },
    {
      "aux_loss": 8.013209819793701,
      "entropy": 0.02059428207576275,
      "epoch": 20.02409638554217,
      "grad_norm": 0.24572697281837463,
      "learning_rate": 1.2026234093523856e-06,
      "loss": 0.07189934700727463,
      "mean_token_accuracy": 0.9991085827350616,
      "num_tokens": 134992831.0,
      "step": 841
    },
    {
      "aux_loss": 8.012920379638672,
      "entropy": 0.021370324306190014,
      "epoch": 20.048192771084338,
      "grad_norm": 0.42761075496673584,
      "learning_rate": 1.2021065946841448e-06,
      "loss": 0.07216686010360718,
      "mean_token_accuracy": 0.998961329460144,
      "num_tokens": 135132806.0,
      "step": 842
    },
    {
      "aux_loss": 8.015305042266846,
      "entropy": 0.018415560014545918,
      "epoch": 20.072289156626507,
      "grad_norm": 0.23072311282157898,
      "learning_rate": 1.2015895210654361e-06,
      "loss": 0.07108470797538757,
      "mean_token_accuracy": 0.9991589784622192,
      "num_tokens": 135310245.0,
      "step": 843
    },
    {
      "aux_loss": 8.016035079956055,
      "entropy": 0.018389324657619,
      "epoch": 20.096385542168676,
      "grad_norm": 0.24892204999923706,
      "learning_rate": 1.2010721896645732e-06,
      "loss": 0.0708564817905426,
      "mean_token_accuracy": 0.9994164109230042,
      "num_tokens": 135478692.0,
      "step": 844
    },
    {
      "aux_loss": 8.015366554260254,
      "entropy": 0.020714948885142803,
      "epoch": 20.120481927710845,
      "grad_norm": 0.24307778477668762,
      "learning_rate": 1.2005546016504526e-06,
      "loss": 0.07157693803310394,
      "mean_token_accuracy": 0.9993534982204437,
      "num_tokens": 135628890.0,
      "step": 845
    },
    {
      "aux_loss": 8.01463508605957,
      "entropy": 0.0212752353399992,
      "epoch": 20.14457831325301,
      "grad_norm": 0.24047866463661194,
      "learning_rate": 1.20003675819255e-06,
      "loss": 0.0722416490316391,
      "mean_token_accuracy": 0.999059796333313,
      "num_tokens": 135782823.0,
      "step": 846
    },
    {
      "aux_loss": 8.01609992980957,
      "entropy": 0.019625852815806866,
      "epoch": 20.16867469879518,
      "grad_norm": 0.24964730441570282,
      "learning_rate": 1.1995186604609185e-06,
      "loss": 0.07153446972370148,
      "mean_token_accuracy": 0.9991658329963684,
      "num_tokens": 135958992.0,
      "step": 847
    },
    {
      "aux_loss": 8.015625,
      "entropy": 0.02252290118485689,
      "epoch": 20.19277108433735,
      "grad_norm": 0.2563897669315338,
      "learning_rate": 1.199000309626186e-06,
      "loss": 0.07267315685749054,
      "mean_token_accuracy": 0.9989122748374939,
      "num_tokens": 136126642.0,
      "step": 848
    },
    {
      "aux_loss": 8.016855239868164,
      "entropy": 0.01888811308890581,
      "epoch": 20.216867469879517,
      "grad_norm": 0.23033533990383148,
      "learning_rate": 1.1984817068595518e-06,
      "loss": 0.07117213308811188,
      "mean_token_accuracy": 0.9991577565670013,
      "num_tokens": 136297196.0,
      "step": 849
    },
    {
      "aux_loss": 8.01485824584961,
      "entropy": 0.018825308419764042,
      "epoch": 20.240963855421686,
      "grad_norm": 0.2620784342288971,
      "learning_rate": 1.1979628533327847e-06,
      "loss": 0.07108195126056671,
      "mean_token_accuracy": 0.9991429448127747,
      "num_tokens": 136451872.0,
      "step": 850
    },
    {
      "aux_loss": 8.015204906463623,
      "entropy": 0.01622555498033762,
      "epoch": 20.265060240963855,
      "grad_norm": 0.22484785318374634,
      "learning_rate": 1.1974437502182203e-06,
      "loss": 0.07052718847990036,
      "mean_token_accuracy": 0.9991798996925354,
      "num_tokens": 136619483.0,
      "step": 851
    },
    {
      "aux_loss": 8.017267227172852,
      "entropy": 0.016841549426317215,
      "epoch": 20.289156626506024,
      "grad_norm": 0.21868085861206055,
      "learning_rate": 1.1969243986887578e-06,
      "loss": 0.07062295079231262,
      "mean_token_accuracy": 0.9991513788700104,
      "num_tokens": 136807598.0,
      "step": 852
    },
    {
      "aux_loss": 8.014912605285645,
      "entropy": 0.019336337223649025,
      "epoch": 20.313253012048193,
      "grad_norm": 0.23610550165176392,
      "learning_rate": 1.1964047999178572e-06,
      "loss": 0.07179201394319534,
      "mean_token_accuracy": 0.9992705881595612,
      "num_tokens": 136972464.0,
      "step": 853
    },
    {
      "aux_loss": 8.013875007629395,
      "entropy": 0.02350557316094637,
      "epoch": 20.337349397590362,
      "grad_norm": 0.2529645264148712,
      "learning_rate": 1.1958849550795387e-06,
      "loss": 0.07279883325099945,
      "mean_token_accuracy": 0.9991186857223511,
      "num_tokens": 137127060.0,
      "step": 854
    },
    {
      "aux_loss": 8.015921115875244,
      "entropy": 0.020983693189918995,
      "epoch": 20.36144578313253,
      "grad_norm": 0.23120775818824768,
      "learning_rate": 1.195364865348377e-06,
      "loss": 0.0721922218799591,
      "mean_token_accuracy": 0.999091386795044,
      "num_tokens": 137282054.0,
      "step": 855
    },
    {
      "aux_loss": 8.017494201660156,
      "entropy": 0.018297821283340454,
      "epoch": 20.3855421686747,
      "grad_norm": 0.2224910408258438,
      "learning_rate": 1.1948445318995005e-06,
      "loss": 0.07094530761241913,
      "mean_token_accuracy": 0.9991941154003143,
      "num_tokens": 137461718.0,
      "step": 856
    },
    {
      "aux_loss": 8.01477336883545,
      "entropy": 0.015757147688418627,
      "epoch": 20.40963855421687,
      "grad_norm": 0.2883114218711853,
      "learning_rate": 1.1943239559085886e-06,
      "loss": 0.07010868191719055,
      "mean_token_accuracy": 0.9994107782840729,
      "num_tokens": 137630346.0,
      "step": 857
    },
    {
      "aux_loss": 8.013188362121582,
      "entropy": 0.021769145503640175,
      "epoch": 20.433734939759034,
      "grad_norm": 0.2615295350551605,
      "learning_rate": 1.1938031385518685e-06,
      "loss": 0.07243753969669342,
      "mean_token_accuracy": 0.9989071488380432,
      "num_tokens": 137765802.0,
      "step": 858
    },
    {
      "aux_loss": 8.015521049499512,
      "entropy": 0.01864070724695921,
      "epoch": 20.457831325301203,
      "grad_norm": 0.231665700674057,
      "learning_rate": 1.1932820810061128e-06,
      "loss": 0.07132695615291595,
      "mean_token_accuracy": 0.9991524517536163,
      "num_tokens": 137931822.0,
      "step": 859
    },
    {
      "aux_loss": 8.014593601226807,
      "entropy": 0.01945006474852562,
      "epoch": 20.481927710843372,
      "grad_norm": 0.2958713173866272,
      "learning_rate": 1.1927607844486368e-06,
      "loss": 0.07161392271518707,
      "mean_token_accuracy": 0.9991827607154846,
      "num_tokens": 138095894.0,
      "step": 860
    },
    {
      "aux_loss": 8.01363468170166,
      "entropy": 0.019185454584658146,
      "epoch": 20.50602409638554,
      "grad_norm": 0.262090802192688,
      "learning_rate": 1.1922392500572957e-06,
      "loss": 0.07143920660018921,
      "mean_token_accuracy": 0.9991458058357239,
      "num_tokens": 138263788.0,
      "step": 861
    },
    {
      "aux_loss": 8.015174388885498,
      "entropy": 0.017392096109688282,
      "epoch": 20.53012048192771,
      "grad_norm": 0.250765323638916,
      "learning_rate": 1.1917174790104822e-06,
      "loss": 0.07097406685352325,
      "mean_token_accuracy": 0.9991758763790131,
      "num_tokens": 138433292.0,
      "step": 862
    },
    {
      "aux_loss": 8.016233921051025,
      "entropy": 0.02018314227461815,
      "epoch": 20.55421686746988,
      "grad_norm": 0.24786005914211273,
      "learning_rate": 1.1911954724871238e-06,
      "loss": 0.07198825478553772,
      "mean_token_accuracy": 0.9992509484291077,
      "num_tokens": 138594833.0,
      "step": 863
    },
    {
      "aux_loss": 8.014333724975586,
      "entropy": 0.01877683959901333,
      "epoch": 20.57831325301205,
      "grad_norm": 0.24034591019153595,
      "learning_rate": 1.1906732316666802e-06,
      "loss": 0.07137183845043182,
      "mean_token_accuracy": 0.9990993142127991,
      "num_tokens": 138750172.0,
      "step": 864
    },
    {
      "aux_loss": 8.015239238739014,
      "entropy": 0.019693774171173573,
      "epoch": 20.602409638554217,
      "grad_norm": 0.28064224123954773,
      "learning_rate": 1.1901507577291398e-06,
      "loss": 0.07180516421794891,
      "mean_token_accuracy": 0.998885989189148,
      "num_tokens": 138917870.0,
      "step": 865
    },
    {
      "aux_loss": 8.016077041625977,
      "entropy": 0.021774998866021633,
      "epoch": 20.626506024096386,
      "grad_norm": 0.25658896565437317,
      "learning_rate": 1.1896280518550184e-06,
      "loss": 0.07241100072860718,
      "mean_token_accuracy": 0.9991808533668518,
      "num_tokens": 139091508.0,
      "step": 866
    },
    {
      "aux_loss": 8.01535940170288,
      "entropy": 0.01949040312319994,
      "epoch": 20.650602409638555,
      "grad_norm": 0.2877894639968872,
      "learning_rate": 1.1891051152253557e-06,
      "loss": 0.07161968946456909,
      "mean_token_accuracy": 0.9990425705909729,
      "num_tokens": 139264973.0,
      "step": 867
    },
    {
      "aux_loss": 8.014946937561035,
      "entropy": 0.021726500242948532,
      "epoch": 20.674698795180724,
      "grad_norm": 0.28723615407943726,
      "learning_rate": 1.1885819490217124e-06,
      "loss": 0.07256714999675751,
      "mean_token_accuracy": 0.9991051852703094,
      "num_tokens": 139440246.0,
      "step": 868
    },
    {
      "aux_loss": 8.015090942382812,
      "entropy": 0.017755858600139618,
      "epoch": 20.698795180722893,
      "grad_norm": 0.2656784951686859,
      "learning_rate": 1.1880585544261689e-06,
      "loss": 0.07081268727779388,
      "mean_token_accuracy": 0.9991943538188934,
      "num_tokens": 139611484.0,
      "step": 869
    },
    {
      "aux_loss": 8.017644882202148,
      "entropy": 0.017926585860550404,
      "epoch": 20.72289156626506,
      "grad_norm": 0.25008681416511536,
      "learning_rate": 1.1875349326213202e-06,
      "loss": 0.07092377543449402,
      "mean_token_accuracy": 0.9991013407707214,
      "num_tokens": 139786451.0,
      "step": 870
    },
    {
      "aux_loss": 8.014115333557129,
      "entropy": 0.019910937640815973,
      "epoch": 20.746987951807228,
      "grad_norm": 0.2600031793117523,
      "learning_rate": 1.187011084790276e-06,
      "loss": 0.07185967266559601,
      "mean_token_accuracy": 0.9992063343524933,
      "num_tokens": 139940870.0,
      "step": 871
    },
    {
      "aux_loss": 8.014479637145996,
      "entropy": 0.01817234791815281,
      "epoch": 20.771084337349397,
      "grad_norm": 0.24683645367622375,
      "learning_rate": 1.1864870121166558e-06,
      "loss": 0.07097449898719788,
      "mean_token_accuracy": 0.9990246891975403,
      "num_tokens": 140110786.0,
      "step": 872
    },
    {
      "aux_loss": 8.0132417678833,
      "entropy": 0.01734836306422949,
      "epoch": 20.795180722891565,
      "grad_norm": 0.25990185141563416,
      "learning_rate": 1.1859627157845877e-06,
      "loss": 0.07099585235118866,
      "mean_token_accuracy": 0.9992154836654663,
      "num_tokens": 140281015.0,
      "step": 873
    },
    {
      "aux_loss": 8.015357971191406,
      "entropy": 0.018785839900374413,
      "epoch": 20.819277108433734,
      "grad_norm": 0.28464582562446594,
      "learning_rate": 1.185438196978705e-06,
      "loss": 0.07173855602741241,
      "mean_token_accuracy": 0.9988526701927185,
      "num_tokens": 140431751.0,
      "step": 874
    },
    {
      "aux_loss": 8.016085624694824,
      "entropy": 0.02032049559056759,
      "epoch": 20.843373493975903,
      "grad_norm": 0.29489409923553467,
      "learning_rate": 1.1849134568841436e-06,
      "loss": 0.07205380499362946,
      "mean_token_accuracy": 0.9989710748195648,
      "num_tokens": 140604033.0,
      "step": 875
    },
    {
      "aux_loss": 8.01557207107544,
      "entropy": 0.020371905528008938,
      "epoch": 20.867469879518072,
      "grad_norm": 0.26816612482070923,
      "learning_rate": 1.184388496686539e-06,
      "loss": 0.07203839719295502,
      "mean_token_accuracy": 0.9989346861839294,
      "num_tokens": 140762882.0,
      "step": 876
    },
    {
      "aux_loss": 8.011805057525635,
      "entropy": 0.01969855558127165,
      "epoch": 20.89156626506024,
      "grad_norm": 0.2511252164840698,
      "learning_rate": 1.1838633175720249e-06,
      "loss": 0.07142087072134018,
      "mean_token_accuracy": 0.9991007447242737,
      "num_tokens": 140913810.0,
      "step": 877
    },
    {
      "aux_loss": 8.015848636627197,
      "entropy": 0.018998843617737293,
      "epoch": 20.91566265060241,
      "grad_norm": 0.2623438835144043,
      "learning_rate": 1.183337920727229e-06,
      "loss": 0.07181046903133392,
      "mean_token_accuracy": 0.9990564286708832,
      "num_tokens": 141068865.0,
      "step": 878
    },
    {
      "aux_loss": 8.013355731964111,
      "entropy": 0.019784594886004925,
      "epoch": 20.93975903614458,
      "grad_norm": 0.28874844312667847,
      "learning_rate": 1.1828123073392708e-06,
      "loss": 0.0717635452747345,
      "mean_token_accuracy": 0.9990528523921967,
      "num_tokens": 141222703.0,
      "step": 879
    },
    {
      "aux_loss": 8.01409912109375,
      "entropy": 0.021358169615268707,
      "epoch": 20.96385542168675,
      "grad_norm": 0.2645207941532135,
      "learning_rate": 1.1822864785957597e-06,
      "loss": 0.07192955911159515,
      "mean_token_accuracy": 0.9991900324821472,
      "num_tokens": 141379277.0,
      "step": 880
    },
    {
      "aux_loss": 8.015501976013184,
      "entropy": 0.018178369849920273,
      "epoch": 20.987951807228917,
      "grad_norm": 0.2733187973499298,
      "learning_rate": 1.1817604356847914e-06,
      "loss": 0.07119777798652649,
      "mean_token_accuracy": 0.9990058541297913,
      "num_tokens": 141529476.0,
      "step": 881
    },
    {
      "aux_loss": 8.014284133911133,
      "entropy": 0.015183513052761555,
      "epoch": 21.0,
      "grad_norm": 0.34507277607917786,
      "learning_rate": 1.1812341797949455e-06,
      "loss": 0.03846766799688339,
      "mean_token_accuracy": 0.9994618892669678,
      "num_tokens": 141570723.0,
      "step": 882
    },
    {
      "aux_loss": 8.01323127746582,
      "entropy": 0.018154023215174675,
      "epoch": 21.02409638554217,
      "grad_norm": 0.24614425003528595,
      "learning_rate": 1.1807077121152827e-06,
      "loss": 0.07076417654752731,
      "mean_token_accuracy": 0.9993391036987305,
      "num_tokens": 141757507.0,
      "step": 883
    },
    {
      "aux_loss": 8.013922691345215,
      "entropy": 0.01983295939862728,
      "epoch": 21.048192771084338,
      "grad_norm": 0.24692152440547943,
      "learning_rate": 1.180181033835342e-06,
      "loss": 0.07120583951473236,
      "mean_token_accuracy": 0.9993802905082703,
      "num_tokens": 141894930.0,
      "step": 884
    },
    {
      "aux_loss": 8.014530181884766,
      "entropy": 0.019670534878969193,
      "epoch": 21.072289156626507,
      "grad_norm": 0.24641910195350647,
      "learning_rate": 1.179654146145139e-06,
      "loss": 0.0713525116443634,
      "mean_token_accuracy": 0.9992433786392212,
      "num_tokens": 142063547.0,
      "step": 885
    },
    {
      "aux_loss": 8.015149116516113,
      "entropy": 0.019047324545681477,
      "epoch": 21.096385542168676,
      "grad_norm": 0.24476729333400726,
      "learning_rate": 1.1791270502351621e-06,
      "loss": 0.07117617130279541,
      "mean_token_accuracy": 0.9990925788879395,
      "num_tokens": 142237959.0,
      "step": 886
    },
    {
      "aux_loss": 8.013380527496338,
      "entropy": 0.02015339396893978,
      "epoch": 21.120481927710845,
      "grad_norm": 0.23826949298381805,
      "learning_rate": 1.1785997472963697e-06,
      "loss": 0.07169397175312042,
      "mean_token_accuracy": 0.9991181790828705,
      "num_tokens": 142391101.0,
      "step": 887
    },
    {
      "aux_loss": 8.015638828277588,
      "entropy": 0.024460233747959137,
      "epoch": 21.14457831325301,
      "grad_norm": 0.27939584851264954,
      "learning_rate": 1.1780722385201889e-06,
      "loss": 0.07349646836519241,
      "mean_token_accuracy": 0.9990708231925964,
      "num_tokens": 142512794.0,
      "step": 888
    },
    {
      "aux_loss": 8.012683868408203,
      "entropy": 0.02155546471476555,
      "epoch": 21.16867469879518,
      "grad_norm": 0.24982911348342896,
      "learning_rate": 1.177544525098511e-06,
      "loss": 0.07213179767131805,
      "mean_token_accuracy": 0.9991197288036346,
      "num_tokens": 142670945.0,
      "step": 889
    },
    {
      "aux_loss": 8.01441478729248,
      "entropy": 0.02026781253516674,
      "epoch": 21.19277108433735,
      "grad_norm": 0.2505277097225189,
      "learning_rate": 1.17701660822369e-06,
      "loss": 0.07161730527877808,
      "mean_token_accuracy": 0.999284565448761,
      "num_tokens": 142817197.0,
      "step": 890
    },
    {
      "aux_loss": 8.015609741210938,
      "entropy": 0.01686427928507328,
      "epoch": 21.216867469879517,
      "grad_norm": 0.25400081276893616,
      "learning_rate": 1.1764884890885399e-06,
      "loss": 0.07048171758651733,
      "mean_token_accuracy": 0.9993577897548676,
      "num_tokens": 142979993.0,
      "step": 891
    },
    {
      "aux_loss": 8.015190601348877,
      "entropy": 0.016790298745036125,
      "epoch": 21.240963855421686,
      "grad_norm": 0.21269014477729797,
      "learning_rate": 1.175960168886331e-06,
      "loss": 0.07070891559123993,
      "mean_token_accuracy": 0.9992137253284454,
      "num_tokens": 143171670.0,
      "step": 892
    },
    {
      "aux_loss": 8.013529777526855,
      "entropy": 0.016728555783629417,
      "epoch": 21.265060240963855,
      "grad_norm": 0.2346455454826355,
      "learning_rate": 1.175431648810789e-06,
      "loss": 0.07016371190547943,
      "mean_token_accuracy": 0.9994739890098572,
      "num_tokens": 143335367.0,
      "step": 893
    },
    {
      "aux_loss": 8.01320505142212,
      "entropy": 0.017979947850108147,
      "epoch": 21.289156626506024,
      "grad_norm": 0.24540981650352478,
      "learning_rate": 1.1749029300560899e-06,
      "loss": 0.07062411308288574,
      "mean_token_accuracy": 0.9991055130958557,
      "num_tokens": 143494903.0,
      "step": 894
    },
    {
      "aux_loss": 8.013950824737549,
      "entropy": 0.020114672370254993,
      "epoch": 21.313253012048193,
      "grad_norm": 0.2551021873950958,
      "learning_rate": 1.17437401381686e-06,
      "loss": 0.0715552270412445,
      "mean_token_accuracy": 0.9991988241672516,
      "num_tokens": 143652508.0,
      "step": 895
    },
    {
      "aux_loss": 8.014235496520996,
      "entropy": 0.01615771744400263,
      "epoch": 21.337349397590362,
      "grad_norm": 0.24847005307674408,
      "learning_rate": 1.1738449012881706e-06,
      "loss": 0.07049012184143066,
      "mean_token_accuracy": 0.999138742685318,
      "num_tokens": 143815733.0,
      "step": 896
    },
    {
      "aux_loss": 8.014393329620361,
      "entropy": 0.022052861750125885,
      "epoch": 21.36144578313253,
      "grad_norm": 0.2577436864376068,
      "learning_rate": 1.1733155936655368e-06,
      "loss": 0.07279989123344421,
      "mean_token_accuracy": 0.9989136755466461,
      "num_tokens": 143986677.0,
      "step": 897
    },
    {
      "aux_loss": 8.01253890991211,
      "entropy": 0.018677118234336376,
      "epoch": 21.3855421686747,
      "grad_norm": 0.22350162267684937,
      "learning_rate": 1.1727860921449152e-06,
      "loss": 0.07126197218894958,
      "mean_token_accuracy": 0.9991850852966309,
      "num_tokens": 144150595.0,
      "step": 898
    },
    {
      "aux_loss": 8.015076160430908,
      "entropy": 0.017602560110390186,
      "epoch": 21.40963855421687,
      "grad_norm": 0.22461570799350739,
      "learning_rate": 1.1722563979226998e-06,
      "loss": 0.07061722129583359,
      "mean_token_accuracy": 0.9992894530296326,
      "num_tokens": 144321176.0,
      "step": 899
    },
    {
      "aux_loss": 8.014989852905273,
      "entropy": 0.01876959018409252,
      "epoch": 21.433734939759034,
      "grad_norm": 0.2587912082672119,
      "learning_rate": 1.17172651219572e-06,
      "loss": 0.07115986943244934,
      "mean_token_accuracy": 0.9990778565406799,
      "num_tokens": 144491907.0,
      "step": 900
    },
    {
      "aux_loss": 8.015000820159912,
      "entropy": 0.018395491875708103,
      "epoch": 21.457831325301203,
      "grad_norm": 0.23809660971164703,
      "learning_rate": 1.1711964361612385e-06,
      "loss": 0.07099337875843048,
      "mean_token_accuracy": 0.9992078840732574,
      "num_tokens": 144649417.0,
      "step": 901
    },
    {
      "aux_loss": 8.014244556427002,
      "entropy": 0.017675181850790977,
      "epoch": 21.481927710843372,
      "grad_norm": 0.24655362963676453,
      "learning_rate": 1.1706661710169475e-06,
      "loss": 0.0707675963640213,
      "mean_token_accuracy": 0.9993148446083069,
      "num_tokens": 144811954.0,
      "step": 902
    },
    {
      "aux_loss": 8.014013290405273,
      "entropy": 0.019099495373666286,
      "epoch": 21.50602409638554,
      "grad_norm": 0.24485772848129272,
      "learning_rate": 1.1701357179609666e-06,
      "loss": 0.07114608585834503,
      "mean_token_accuracy": 0.9993022382259369,
      "num_tokens": 144970306.0,
      "step": 903
    },
    {
      "aux_loss": 8.012389659881592,
      "entropy": 0.020937589928507805,
      "epoch": 21.53012048192771,
      "grad_norm": 0.2481389045715332,
      "learning_rate": 1.1696050781918398e-06,
      "loss": 0.07180596143007278,
      "mean_token_accuracy": 0.999099999666214,
      "num_tokens": 145115040.0,
      "step": 904
    },
    {
      "aux_loss": 8.014892578125,
      "entropy": 0.018934819847345352,
      "epoch": 21.55421686746988,
      "grad_norm": 0.2642694115638733,
      "learning_rate": 1.169074252908533e-06,
      "loss": 0.07122250646352768,
      "mean_token_accuracy": 0.9992123246192932,
      "num_tokens": 145279167.0,
      "step": 905
    },
    {
      "aux_loss": 8.014004707336426,
      "entropy": 0.021409288980066776,
      "epoch": 21.57831325301205,
      "grad_norm": 0.2766292989253998,
      "learning_rate": 1.168543243310432e-06,
      "loss": 0.07221560180187225,
      "mean_token_accuracy": 0.9993177652359009,
      "num_tokens": 145447208.0,
      "step": 906
    },
    {
      "aux_loss": 8.014603614807129,
      "entropy": 0.01751386933028698,
      "epoch": 21.602409638554217,
      "grad_norm": 0.25062957406044006,
      "learning_rate": 1.1680120505973382e-06,
      "loss": 0.0708167627453804,
      "mean_token_accuracy": 0.9991951882839203,
      "num_tokens": 145606489.0,
      "step": 907
    },
    {
      "aux_loss": 8.018440246582031,
      "entropy": 0.015230256132781506,
      "epoch": 21.626506024096386,
      "grad_norm": 0.25712889432907104,
      "learning_rate": 1.1674806759694668e-06,
      "loss": 0.07026419043540955,
      "mean_token_accuracy": 0.9993832409381866,
      "num_tokens": 145772295.0,
      "step": 908
    },
    {
      "aux_loss": 8.013906478881836,
      "entropy": 0.017842063680291176,
      "epoch": 21.650602409638555,
      "grad_norm": 0.2853424847126007,
      "learning_rate": 1.1669491206274444e-06,
      "loss": 0.0708966851234436,
      "mean_token_accuracy": 0.9990724325180054,
      "num_tokens": 145909545.0,
      "step": 909
    },
    {
      "aux_loss": 8.012400150299072,
      "entropy": 0.019418477080762386,
      "epoch": 21.674698795180724,
      "grad_norm": 0.2649625241756439,
      "learning_rate": 1.1664173857723058e-06,
      "loss": 0.0713847279548645,
      "mean_token_accuracy": 0.9992568492889404,
      "num_tokens": 146067220.0,
      "step": 910
    },
    {
      "aux_loss": 8.015399932861328,
      "entropy": 0.017952339723706245,
      "epoch": 21.698795180722893,
      "grad_norm": 0.2824140191078186,
      "learning_rate": 1.1658854726054913e-06,
      "loss": 0.07136330008506775,
      "mean_token_accuracy": 0.9991283416748047,
      "num_tokens": 146241091.0,
      "step": 911
    },
    {
      "aux_loss": 8.014923095703125,
      "entropy": 0.020302233286201954,
      "epoch": 21.72289156626506,
      "grad_norm": 0.27626627683639526,
      "learning_rate": 1.1653533823288444e-06,
      "loss": 0.07169334590435028,
      "mean_token_accuracy": 0.9992359578609467,
      "num_tokens": 146407090.0,
      "step": 912
    },
    {
      "aux_loss": 8.01394510269165,
      "entropy": 0.018771435134112835,
      "epoch": 21.746987951807228,
      "grad_norm": 0.2780453562736511,
      "learning_rate": 1.1648211161446084e-06,
      "loss": 0.07132960855960846,
      "mean_token_accuracy": 0.9989534914493561,
      "num_tokens": 146572741.0,
      "step": 913
    },
    {
      "aux_loss": 8.014733791351318,
      "entropy": 0.017568754963576794,
      "epoch": 21.771084337349397,
      "grad_norm": 0.24800656735897064,
      "learning_rate": 1.164288675255424e-06,
      "loss": 0.07059050351381302,
      "mean_token_accuracy": 0.9992434084415436,
      "num_tokens": 146771249.0,
      "step": 914
    },
    {
      "aux_loss": 8.013745307922363,
      "entropy": 0.018569269217550755,
      "epoch": 21.795180722891565,
      "grad_norm": 0.2566494047641754,
      "learning_rate": 1.1637560608643269e-06,
      "loss": 0.07117751240730286,
      "mean_token_accuracy": 0.9991358816623688,
      "num_tokens": 146927606.0,
      "step": 915
    },
    {
      "aux_loss": 8.014211654663086,
      "entropy": 0.019441496580839157,
      "epoch": 21.819277108433734,
      "grad_norm": 0.27913621068000793,
      "learning_rate": 1.1632232741747449e-06,
      "loss": 0.07145224511623383,
      "mean_token_accuracy": 0.9990966022014618,
      "num_tokens": 147078302.0,
      "step": 916
    },
    {
      "aux_loss": 8.012444496154785,
      "entropy": 0.019019109196960926,
      "epoch": 21.843373493975903,
      "grad_norm": 0.28188833594322205,
      "learning_rate": 1.1626903163904948e-06,
      "loss": 0.07167785614728928,
      "mean_token_accuracy": 0.9990155696868896,
      "num_tokens": 147233952.0,
      "step": 917
    },
    {
      "aux_loss": 8.01331901550293,
      "entropy": 0.01779519859701395,
      "epoch": 21.867469879518072,
      "grad_norm": 0.27154967188835144,
      "learning_rate": 1.1621571887157803e-06,
      "loss": 0.07083356380462646,
      "mean_token_accuracy": 0.9991532564163208,
      "num_tokens": 147400698.0,
      "step": 918
    },
    {
      "aux_loss": 8.013606548309326,
      "entropy": 0.017602672800421715,
      "epoch": 21.89156626506024,
      "grad_norm": 0.2656519114971161,
      "learning_rate": 1.1616238923551887e-06,
      "loss": 0.07089027762413025,
      "mean_token_accuracy": 0.9992519021034241,
      "num_tokens": 147582738.0,
      "step": 919
    },
    {
      "aux_loss": 8.015371799468994,
      "entropy": 0.018127653747797012,
      "epoch": 21.91566265060241,
      "grad_norm": 0.2636999189853668,
      "learning_rate": 1.1610904285136886e-06,
      "loss": 0.07094477117061615,
      "mean_token_accuracy": 0.9991376996040344,
      "num_tokens": 147775785.0,
      "step": 920
    },
    {
      "aux_loss": 8.013552188873291,
      "entropy": 0.01729942485690117,
      "epoch": 21.93975903614458,
      "grad_norm": 0.26023203134536743,
      "learning_rate": 1.1605567983966271e-06,
      "loss": 0.07063956558704376,
      "mean_token_accuracy": 0.9993418455123901,
      "num_tokens": 147936430.0,
      "step": 921
    },
    {
      "aux_loss": 8.014430522918701,
      "entropy": 0.015739817172288895,
      "epoch": 21.96385542168675,
      "grad_norm": 0.2076667696237564,
      "learning_rate": 1.1600230032097266e-06,
      "loss": 0.07040604948997498,
      "mean_token_accuracy": 0.999272495508194,
      "num_tokens": 148137575.0,
      "step": 922
    },
    {
      "aux_loss": 8.011709213256836,
      "entropy": 0.019408750347793102,
      "epoch": 21.987951807228917,
      "grad_norm": 0.2545049786567688,
      "learning_rate": 1.1594890441590829e-06,
      "loss": 0.07166232168674469,
      "mean_token_accuracy": 0.9990161955356598,
      "num_tokens": 148288575.0,
      "step": 923
    },
    {
      "aux_loss": 8.014568328857422,
      "entropy": 0.019053278490900993,
      "epoch": 22.0,
      "grad_norm": 0.4009021818637848,
      "learning_rate": 1.1589549224511617e-06,
      "loss": 0.03886584937572479,
      "mean_token_accuracy": 0.9992607831954956,
      "num_tokens": 148312186.0,
      "step": 924
    }
  ],
  "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": 3.075369309803905e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}