{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6,
  "eval_steps": 500,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004,
      "grad_norm": 2.131298542022705,
      "learning_rate": 0.00019962000000000002,
      "loss": 2.0121,
      "mean_token_accuracy": 0.6703190118074417,
      "num_tokens": 2348.0,
      "step": 20
    },
    {
      "epoch": 0.008,
      "grad_norm": 1.2890418767929077,
      "learning_rate": 0.00019922,
      "loss": 0.2751,
      "mean_token_accuracy": 0.9120438575744629,
      "num_tokens": 4697.0,
      "step": 40
    },
    {
      "epoch": 0.012,
      "grad_norm": 0.8126867413520813,
      "learning_rate": 0.00019882,
      "loss": 0.1966,
      "mean_token_accuracy": 0.9203487157821655,
      "num_tokens": 7014.0,
      "step": 60
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.6051881313323975,
      "learning_rate": 0.00019842000000000001,
      "loss": 0.1851,
      "mean_token_accuracy": 0.9291799515485764,
      "num_tokens": 9327.0,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6378348469734192,
      "learning_rate": 0.00019802,
      "loss": 0.1766,
      "mean_token_accuracy": 0.9286984890699387,
      "num_tokens": 11670.0,
      "step": 100
    },
    {
      "epoch": 0.024,
      "grad_norm": 0.624138593673706,
      "learning_rate": 0.00019762,
      "loss": 0.1784,
      "mean_token_accuracy": 0.9264282643795013,
      "num_tokens": 14000.0,
      "step": 120
    },
    {
      "epoch": 0.028,
      "grad_norm": 0.2111046463251114,
      "learning_rate": 0.00019722,
      "loss": 0.1702,
      "mean_token_accuracy": 0.9321970880031586,
      "num_tokens": 16329.0,
      "step": 140
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.5350440740585327,
      "learning_rate": 0.00019682,
      "loss": 0.171,
      "mean_token_accuracy": 0.9311463803052902,
      "num_tokens": 18667.0,
      "step": 160
    },
    {
      "epoch": 0.036,
      "grad_norm": 0.19237647950649261,
      "learning_rate": 0.00019642,
      "loss": 0.167,
      "mean_token_accuracy": 0.9345656305551528,
      "num_tokens": 20985.0,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.28153756260871887,
      "learning_rate": 0.00019602,
      "loss": 0.1674,
      "mean_token_accuracy": 0.9329667061567306,
      "num_tokens": 23325.0,
      "step": 200
    },
    {
      "epoch": 0.044,
      "grad_norm": 0.8545331954956055,
      "learning_rate": 0.00019562,
      "loss": 0.166,
      "mean_token_accuracy": 0.9344212204217911,
      "num_tokens": 25670.0,
      "step": 220
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.24941129982471466,
      "learning_rate": 0.00019522,
      "loss": 0.1661,
      "mean_token_accuracy": 0.934859549999237,
      "num_tokens": 28016.0,
      "step": 240
    },
    {
      "epoch": 0.052,
      "grad_norm": 0.29549548029899597,
      "learning_rate": 0.00019482,
      "loss": 0.1707,
      "mean_token_accuracy": 0.9345517784357071,
      "num_tokens": 30345.0,
      "step": 260
    },
    {
      "epoch": 0.056,
      "grad_norm": 0.20388178527355194,
      "learning_rate": 0.00019442,
      "loss": 0.1673,
      "mean_token_accuracy": 0.9353183209896088,
      "num_tokens": 32691.0,
      "step": 280
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.10762794315814972,
      "learning_rate": 0.00019402,
      "loss": 0.1642,
      "mean_token_accuracy": 0.9325390756130219,
      "num_tokens": 35030.0,
      "step": 300
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.07676753401756287,
      "learning_rate": 0.00019362,
      "loss": 0.1633,
      "mean_token_accuracy": 0.9348760217428207,
      "num_tokens": 37359.0,
      "step": 320
    },
    {
      "epoch": 0.068,
      "grad_norm": 0.06781225651502609,
      "learning_rate": 0.00019322,
      "loss": 0.1589,
      "mean_token_accuracy": 0.936154904961586,
      "num_tokens": 39707.0,
      "step": 340
    },
    {
      "epoch": 0.072,
      "grad_norm": 0.10010460019111633,
      "learning_rate": 0.00019282000000000001,
      "loss": 0.1583,
      "mean_token_accuracy": 0.9410124599933625,
      "num_tokens": 42071.0,
      "step": 360
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.07932794839143753,
      "learning_rate": 0.00019242,
      "loss": 0.1608,
      "mean_token_accuracy": 0.9380002528429031,
      "num_tokens": 44404.0,
      "step": 380
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.06678586453199387,
      "learning_rate": 0.00019202,
      "loss": 0.1633,
      "mean_token_accuracy": 0.9347729980945587,
      "num_tokens": 46715.0,
      "step": 400
    },
    {
      "epoch": 0.084,
      "grad_norm": 0.05118393525481224,
      "learning_rate": 0.00019162,
      "loss": 0.1621,
      "mean_token_accuracy": 0.9356200367212295,
      "num_tokens": 49030.0,
      "step": 420
    },
    {
      "epoch": 0.088,
      "grad_norm": 0.07563836127519608,
      "learning_rate": 0.00019122,
      "loss": 0.1603,
      "mean_token_accuracy": 0.9362942427396774,
      "num_tokens": 51366.0,
      "step": 440
    },
    {
      "epoch": 0.092,
      "grad_norm": 0.053388580679893494,
      "learning_rate": 0.00019082,
      "loss": 0.1585,
      "mean_token_accuracy": 0.9377258807420731,
      "num_tokens": 53706.0,
      "step": 460
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.05659119412302971,
      "learning_rate": 0.00019042,
      "loss": 0.1575,
      "mean_token_accuracy": 0.937972965836525,
      "num_tokens": 56052.0,
      "step": 480
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.04934714362025261,
      "learning_rate": 0.00019002,
      "loss": 0.1606,
      "mean_token_accuracy": 0.9356311202049256,
      "num_tokens": 58374.0,
      "step": 500
    },
    {
      "epoch": 0.104,
      "grad_norm": 0.05647804215550423,
      "learning_rate": 0.00018962000000000002,
      "loss": 0.1587,
      "mean_token_accuracy": 0.9353525519371033,
      "num_tokens": 60706.0,
      "step": 520
    },
    {
      "epoch": 0.108,
      "grad_norm": 0.058523017913103104,
      "learning_rate": 0.00018922,
      "loss": 0.1595,
      "mean_token_accuracy": 0.9371987581253052,
      "num_tokens": 63021.0,
      "step": 540
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.03793497756123543,
      "learning_rate": 0.00018882000000000003,
      "loss": 0.1584,
      "mean_token_accuracy": 0.9341553807258606,
      "num_tokens": 65369.0,
      "step": 560
    },
    {
      "epoch": 0.116,
      "grad_norm": 0.04743633046746254,
      "learning_rate": 0.00018842000000000002,
      "loss": 0.1586,
      "mean_token_accuracy": 0.936157900094986,
      "num_tokens": 67700.0,
      "step": 580
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.05463261529803276,
      "learning_rate": 0.00018802,
      "loss": 0.1578,
      "mean_token_accuracy": 0.937694975733757,
      "num_tokens": 70038.0,
      "step": 600
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.11279874294996262,
      "learning_rate": 0.00018762000000000002,
      "loss": 0.1584,
      "mean_token_accuracy": 0.9382745862007141,
      "num_tokens": 72367.0,
      "step": 620
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.050823234021663666,
      "learning_rate": 0.00018722,
      "loss": 0.1602,
      "mean_token_accuracy": 0.9388011395931244,
      "num_tokens": 74677.0,
      "step": 640
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.04983159899711609,
      "learning_rate": 0.00018682000000000003,
      "loss": 0.1584,
      "mean_token_accuracy": 0.9374112606048584,
      "num_tokens": 77014.0,
      "step": 660
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.05084273964166641,
      "learning_rate": 0.00018642000000000002,
      "loss": 0.1591,
      "mean_token_accuracy": 0.9387890577316285,
      "num_tokens": 79339.0,
      "step": 680
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.04936506226658821,
      "learning_rate": 0.00018602,
      "loss": 0.157,
      "mean_token_accuracy": 0.9359721839427948,
      "num_tokens": 81688.0,
      "step": 700
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.04119481146335602,
      "learning_rate": 0.00018562000000000003,
      "loss": 0.1584,
      "mean_token_accuracy": 0.9359663873910904,
      "num_tokens": 84025.0,
      "step": 720
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.03988514095544815,
      "learning_rate": 0.00018522000000000002,
      "loss": 0.1595,
      "mean_token_accuracy": 0.9372004926204681,
      "num_tokens": 86341.0,
      "step": 740
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.04381653666496277,
      "learning_rate": 0.00018482,
      "loss": 0.1591,
      "mean_token_accuracy": 0.9373360633850097,
      "num_tokens": 88665.0,
      "step": 760
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.03504428267478943,
      "learning_rate": 0.00018442000000000003,
      "loss": 0.1579,
      "mean_token_accuracy": 0.9403579801321029,
      "num_tokens": 90997.0,
      "step": 780
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.03745226562023163,
      "learning_rate": 0.00018402000000000002,
      "loss": 0.1584,
      "mean_token_accuracy": 0.9388439536094666,
      "num_tokens": 93322.0,
      "step": 800
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.033838506788015366,
      "learning_rate": 0.00018362,
      "loss": 0.1562,
      "mean_token_accuracy": 0.9360319077968597,
      "num_tokens": 95688.0,
      "step": 820
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.0314440056681633,
      "learning_rate": 0.00018322000000000002,
      "loss": 0.1596,
      "mean_token_accuracy": 0.9349893003702163,
      "num_tokens": 98005.0,
      "step": 840
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.06577116250991821,
      "learning_rate": 0.00018282000000000001,
      "loss": 0.1589,
      "mean_token_accuracy": 0.9364802747964859,
      "num_tokens": 100327.0,
      "step": 860
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.029735982418060303,
      "learning_rate": 0.00018242,
      "loss": 0.1576,
      "mean_token_accuracy": 0.9365677893161773,
      "num_tokens": 102671.0,
      "step": 880
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.03155644237995148,
      "learning_rate": 0.00018202000000000002,
      "loss": 0.1587,
      "mean_token_accuracy": 0.9387517213821411,
      "num_tokens": 104996.0,
      "step": 900
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.027411427348852158,
      "learning_rate": 0.00018162,
      "loss": 0.1581,
      "mean_token_accuracy": 0.9342827945947647,
      "num_tokens": 107337.0,
      "step": 920
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.024014495313167572,
      "learning_rate": 0.00018122,
      "loss": 0.1583,
      "mean_token_accuracy": 0.9355534523725509,
      "num_tokens": 109674.0,
      "step": 940
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.02990046516060829,
      "learning_rate": 0.00018082000000000002,
      "loss": 0.1577,
      "mean_token_accuracy": 0.9373938798904419,
      "num_tokens": 112015.0,
      "step": 960
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.020676879212260246,
      "learning_rate": 0.00018042,
      "loss": 0.1582,
      "mean_token_accuracy": 0.9363098949193954,
      "num_tokens": 114349.0,
      "step": 980
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.024556942284107208,
      "learning_rate": 0.00018002,
      "loss": 0.159,
      "mean_token_accuracy": 0.9355833351612091,
      "num_tokens": 116672.0,
      "step": 1000
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.22343556582927704,
      "learning_rate": 0.00017962000000000002,
      "loss": 0.157,
      "mean_token_accuracy": 0.9363467574119568,
      "num_tokens": 119023.0,
      "step": 1020
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.515048623085022,
      "learning_rate": 0.00017922,
      "loss": 0.1585,
      "mean_token_accuracy": 0.9374267637729645,
      "num_tokens": 121366.0,
      "step": 1040
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.2672664225101471,
      "learning_rate": 0.00017882,
      "loss": 0.1704,
      "mean_token_accuracy": 0.938492265343666,
      "num_tokens": 123719.0,
      "step": 1060
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.11909265071153641,
      "learning_rate": 0.00017842000000000002,
      "loss": 0.1752,
      "mean_token_accuracy": 0.9297878712415695,
      "num_tokens": 126053.0,
      "step": 1080
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.11977271735668182,
      "learning_rate": 0.00017802,
      "loss": 0.1652,
      "mean_token_accuracy": 0.9360411554574967,
      "num_tokens": 128379.0,
      "step": 1100
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.17722292244434357,
      "learning_rate": 0.00017762,
      "loss": 0.1697,
      "mean_token_accuracy": 0.9314894318580628,
      "num_tokens": 130700.0,
      "step": 1120
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.8375388979911804,
      "learning_rate": 0.00017722000000000001,
      "loss": 0.1895,
      "mean_token_accuracy": 0.9316652357578278,
      "num_tokens": 133026.0,
      "step": 1140
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.10569056123495102,
      "learning_rate": 0.00017682,
      "loss": 0.1679,
      "mean_token_accuracy": 0.9333775132894516,
      "num_tokens": 135371.0,
      "step": 1160
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.07626856118440628,
      "learning_rate": 0.00017642,
      "loss": 0.1613,
      "mean_token_accuracy": 0.9380175620317459,
      "num_tokens": 137695.0,
      "step": 1180
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.06852507591247559,
      "learning_rate": 0.00017602,
      "loss": 0.1693,
      "mean_token_accuracy": 0.9329774439334869,
      "num_tokens": 140007.0,
      "step": 1200
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.11691898107528687,
      "learning_rate": 0.00017562,
      "loss": 0.159,
      "mean_token_accuracy": 0.9360336065292358,
      "num_tokens": 142345.0,
      "step": 1220
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.10780195891857147,
      "learning_rate": 0.00017522000000000002,
      "loss": 0.1586,
      "mean_token_accuracy": 0.9376411676406861,
      "num_tokens": 144693.0,
      "step": 1240
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.07631397247314453,
      "learning_rate": 0.00017482,
      "loss": 0.1633,
      "mean_token_accuracy": 0.9390978574752807,
      "num_tokens": 147031.0,
      "step": 1260
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.05569858103990555,
      "learning_rate": 0.00017442,
      "loss": 0.1615,
      "mean_token_accuracy": 0.9400094121694564,
      "num_tokens": 149371.0,
      "step": 1280
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.04815123230218887,
      "learning_rate": 0.00017402000000000002,
      "loss": 0.1584,
      "mean_token_accuracy": 0.9370361328125,
      "num_tokens": 151716.0,
      "step": 1300
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.11904877424240112,
      "learning_rate": 0.00017362,
      "loss": 0.159,
      "mean_token_accuracy": 0.941026845574379,
      "num_tokens": 154046.0,
      "step": 1320
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.12382964044809341,
      "learning_rate": 0.00017322,
      "loss": 0.1594,
      "mean_token_accuracy": 0.9340965986251831,
      "num_tokens": 156382.0,
      "step": 1340
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.05523005872964859,
      "learning_rate": 0.00017282000000000002,
      "loss": 0.1593,
      "mean_token_accuracy": 0.9396583586931229,
      "num_tokens": 158705.0,
      "step": 1360
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.05591598525643349,
      "learning_rate": 0.00017242,
      "loss": 0.1592,
      "mean_token_accuracy": 0.9364501267671586,
      "num_tokens": 161025.0,
      "step": 1380
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.06780663877725601,
      "learning_rate": 0.00017202,
      "loss": 0.1618,
      "mean_token_accuracy": 0.936437115073204,
      "num_tokens": 163345.0,
      "step": 1400
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.03291817009449005,
      "learning_rate": 0.00017162000000000001,
      "loss": 0.1569,
      "mean_token_accuracy": 0.9380638599395752,
      "num_tokens": 165711.0,
      "step": 1420
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.04884820431470871,
      "learning_rate": 0.00017122,
      "loss": 0.1608,
      "mean_token_accuracy": 0.9370934247970581,
      "num_tokens": 168007.0,
      "step": 1440
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.04577581211924553,
      "learning_rate": 0.00017082,
      "loss": 0.1595,
      "mean_token_accuracy": 0.9374603897333145,
      "num_tokens": 170332.0,
      "step": 1460
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.03866467997431755,
      "learning_rate": 0.00017042,
      "loss": 0.1582,
      "mean_token_accuracy": 0.9355076909065246,
      "num_tokens": 172667.0,
      "step": 1480
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.06204424798488617,
      "learning_rate": 0.00017002,
      "loss": 0.1599,
      "mean_token_accuracy": 0.9372841835021972,
      "num_tokens": 174970.0,
      "step": 1500
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.03288702666759491,
      "learning_rate": 0.00016962,
      "loss": 0.1578,
      "mean_token_accuracy": 0.93585424721241,
      "num_tokens": 177317.0,
      "step": 1520
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.03605024516582489,
      "learning_rate": 0.00016922,
      "loss": 0.1597,
      "mean_token_accuracy": 0.9366346269845962,
      "num_tokens": 179629.0,
      "step": 1540
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.03328383341431618,
      "learning_rate": 0.00016882,
      "loss": 0.1582,
      "mean_token_accuracy": 0.9372009009122848,
      "num_tokens": 181960.0,
      "step": 1560
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.03522924706339836,
      "learning_rate": 0.00016842,
      "loss": 0.1572,
      "mean_token_accuracy": 0.9355730235576629,
      "num_tokens": 184314.0,
      "step": 1580
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.0317777544260025,
      "learning_rate": 0.00016802,
      "loss": 0.1572,
      "mean_token_accuracy": 0.9378984242677688,
      "num_tokens": 186658.0,
      "step": 1600
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.07111163437366486,
      "learning_rate": 0.00016762,
      "loss": 0.1613,
      "mean_token_accuracy": 0.9344337552785873,
      "num_tokens": 188974.0,
      "step": 1620
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.04765714704990387,
      "learning_rate": 0.00016722,
      "loss": 0.1608,
      "mean_token_accuracy": 0.9345141768455505,
      "num_tokens": 191276.0,
      "step": 1640
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.041960619390010834,
      "learning_rate": 0.00016682,
      "loss": 0.1594,
      "mean_token_accuracy": 0.9375987917184829,
      "num_tokens": 193597.0,
      "step": 1660
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.041757769882678986,
      "learning_rate": 0.00016642,
      "loss": 0.157,
      "mean_token_accuracy": 0.9367031455039978,
      "num_tokens": 195949.0,
      "step": 1680
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.05323236435651779,
      "learning_rate": 0.00016601999999999999,
      "loss": 0.1654,
      "mean_token_accuracy": 0.9359392642974853,
      "num_tokens": 198267.0,
      "step": 1700
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.08934314548969269,
      "learning_rate": 0.00016562,
      "loss": 0.1619,
      "mean_token_accuracy": 0.9331722050905228,
      "num_tokens": 200613.0,
      "step": 1720
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.033347178250551224,
      "learning_rate": 0.00016522,
      "loss": 0.1611,
      "mean_token_accuracy": 0.9317500472068787,
      "num_tokens": 202927.0,
      "step": 1740
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.03238425776362419,
      "learning_rate": 0.00016482,
      "loss": 0.1582,
      "mean_token_accuracy": 0.9358267247676849,
      "num_tokens": 205273.0,
      "step": 1760
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.03249628096818924,
      "learning_rate": 0.00016442000000000003,
      "loss": 0.1591,
      "mean_token_accuracy": 0.9367831707000732,
      "num_tokens": 207595.0,
      "step": 1780
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.034572117030620575,
      "learning_rate": 0.00016402000000000002,
      "loss": 0.1612,
      "mean_token_accuracy": 0.9356453567743301,
      "num_tokens": 209892.0,
      "step": 1800
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.04566624388098717,
      "learning_rate": 0.00016362,
      "loss": 0.1605,
      "mean_token_accuracy": 0.9359289228916168,
      "num_tokens": 212194.0,
      "step": 1820
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.029180865734815598,
      "learning_rate": 0.00016322000000000003,
      "loss": 0.1579,
      "mean_token_accuracy": 0.9378338158130646,
      "num_tokens": 214535.0,
      "step": 1840
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.04812979698181152,
      "learning_rate": 0.00016282000000000002,
      "loss": 0.1576,
      "mean_token_accuracy": 0.9391636937856674,
      "num_tokens": 216872.0,
      "step": 1860
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.06872449070215225,
      "learning_rate": 0.00016242,
      "loss": 0.1606,
      "mean_token_accuracy": 0.9389324098825454,
      "num_tokens": 219184.0,
      "step": 1880
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.05308040603995323,
      "learning_rate": 0.00016202000000000002,
      "loss": 0.1592,
      "mean_token_accuracy": 0.9385550439357757,
      "num_tokens": 221499.0,
      "step": 1900
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.13082465529441833,
      "learning_rate": 0.00016162000000000001,
      "loss": 0.1579,
      "mean_token_accuracy": 0.9380007416009903,
      "num_tokens": 223847.0,
      "step": 1920
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.03414028137922287,
      "learning_rate": 0.00016122,
      "loss": 0.1602,
      "mean_token_accuracy": 0.9369382321834564,
      "num_tokens": 226156.0,
      "step": 1940
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.04112333804368973,
      "learning_rate": 0.00016082000000000002,
      "loss": 0.1597,
      "mean_token_accuracy": 0.935492268204689,
      "num_tokens": 228475.0,
      "step": 1960
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.02955610118806362,
      "learning_rate": 0.00016042,
      "loss": 0.1591,
      "mean_token_accuracy": 0.93908212184906,
      "num_tokens": 230792.0,
      "step": 1980
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.024307863786816597,
      "learning_rate": 0.00016002,
      "loss": 0.1595,
      "mean_token_accuracy": 0.9393438696861267,
      "num_tokens": 233105.0,
      "step": 2000
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.031049860641360283,
      "learning_rate": 0.00015962000000000002,
      "loss": 0.157,
      "mean_token_accuracy": 0.9350471049547195,
      "num_tokens": 235457.0,
      "step": 2020
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.031201306730508804,
      "learning_rate": 0.00015922,
      "loss": 0.1597,
      "mean_token_accuracy": 0.9370091885328293,
      "num_tokens": 237770.0,
      "step": 2040
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.03218454122543335,
      "learning_rate": 0.00015882,
      "loss": 0.159,
      "mean_token_accuracy": 0.9386389076709747,
      "num_tokens": 240090.0,
      "step": 2060
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.03348623961210251,
      "learning_rate": 0.00015842000000000002,
      "loss": 0.1597,
      "mean_token_accuracy": 0.9362810254096985,
      "num_tokens": 242393.0,
      "step": 2080
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.04007818177342415,
      "learning_rate": 0.00015802,
      "loss": 0.1581,
      "mean_token_accuracy": 0.9370237767696381,
      "num_tokens": 244733.0,
      "step": 2100
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.03378809243440628,
      "learning_rate": 0.00015762,
      "loss": 0.1568,
      "mean_token_accuracy": 0.9377921044826507,
      "num_tokens": 247088.0,
      "step": 2120
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.02798735350370407,
      "learning_rate": 0.00015722000000000002,
      "loss": 0.1586,
      "mean_token_accuracy": 0.9383140057325363,
      "num_tokens": 249415.0,
      "step": 2140
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.02400992065668106,
      "learning_rate": 0.00015682,
      "loss": 0.1581,
      "mean_token_accuracy": 0.9389418184757232,
      "num_tokens": 251751.0,
      "step": 2160
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.028334975242614746,
      "learning_rate": 0.00015642000000000002,
      "loss": 0.1578,
      "mean_token_accuracy": 0.936533722281456,
      "num_tokens": 254093.0,
      "step": 2180
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.024794427677989006,
      "learning_rate": 0.00015602000000000001,
      "loss": 0.1587,
      "mean_token_accuracy": 0.9344067484140396,
      "num_tokens": 256422.0,
      "step": 2200
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.024761928245425224,
      "learning_rate": 0.00015562,
      "loss": 0.1589,
      "mean_token_accuracy": 0.936401879787445,
      "num_tokens": 258747.0,
      "step": 2220
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.023300737142562866,
      "learning_rate": 0.00015522000000000002,
      "loss": 0.1594,
      "mean_token_accuracy": 0.9409920126199722,
      "num_tokens": 261059.0,
      "step": 2240
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.023498738184571266,
      "learning_rate": 0.00015482,
      "loss": 0.1595,
      "mean_token_accuracy": 0.9378518283367157,
      "num_tokens": 263369.0,
      "step": 2260
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.020731788128614426,
      "learning_rate": 0.00015442,
      "loss": 0.1586,
      "mean_token_accuracy": 0.9395518034696579,
      "num_tokens": 265691.0,
      "step": 2280
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.03587990626692772,
      "learning_rate": 0.00015402000000000002,
      "loss": 0.1587,
      "mean_token_accuracy": 0.9378482937812805,
      "num_tokens": 268016.0,
      "step": 2300
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.03809090331196785,
      "learning_rate": 0.00015362,
      "loss": 0.1588,
      "mean_token_accuracy": 0.9367815405130386,
      "num_tokens": 270334.0,
      "step": 2320
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.04313996061682701,
      "learning_rate": 0.00015322,
      "loss": 0.1593,
      "mean_token_accuracy": 0.9363488733768464,
      "num_tokens": 272651.0,
      "step": 2340
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.033811088651418686,
      "learning_rate": 0.00015282000000000002,
      "loss": 0.1594,
      "mean_token_accuracy": 0.9361798793077469,
      "num_tokens": 274964.0,
      "step": 2360
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.03164658322930336,
      "learning_rate": 0.00015242,
      "loss": 0.1598,
      "mean_token_accuracy": 0.9361578047275543,
      "num_tokens": 277276.0,
      "step": 2380
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.030314739793539047,
      "learning_rate": 0.00015202,
      "loss": 0.1601,
      "mean_token_accuracy": 0.9364828914403915,
      "num_tokens": 279583.0,
      "step": 2400
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.03359575197100639,
      "learning_rate": 0.00015162000000000002,
      "loss": 0.1577,
      "mean_token_accuracy": 0.9370669215917588,
      "num_tokens": 281927.0,
      "step": 2420
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.025949697941541672,
      "learning_rate": 0.00015122,
      "loss": 0.1582,
      "mean_token_accuracy": 0.9364959686994553,
      "num_tokens": 284267.0,
      "step": 2440
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.031149016693234444,
      "learning_rate": 0.00015082,
      "loss": 0.1589,
      "mean_token_accuracy": 0.9384445637464524,
      "num_tokens": 286585.0,
      "step": 2460
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.031569018959999084,
      "learning_rate": 0.00015042,
      "loss": 0.1589,
      "mean_token_accuracy": 0.9359267175197601,
      "num_tokens": 288907.0,
      "step": 2480
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.02912713773548603,
      "learning_rate": 0.00015002,
      "loss": 0.1582,
      "mean_token_accuracy": 0.9354292452335358,
      "num_tokens": 291240.0,
      "step": 2500
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.029648004099726677,
      "learning_rate": 0.00014962,
      "loss": 0.1577,
      "mean_token_accuracy": 0.9371567130088806,
      "num_tokens": 293587.0,
      "step": 2520
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.01994331367313862,
      "learning_rate": 0.00014922,
      "loss": 0.1565,
      "mean_token_accuracy": 0.9374659866094589,
      "num_tokens": 295947.0,
      "step": 2540
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.022220291197299957,
      "learning_rate": 0.00014882,
      "loss": 0.1576,
      "mean_token_accuracy": 0.9352899432182312,
      "num_tokens": 298291.0,
      "step": 2560
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.019389133900403976,
      "learning_rate": 0.00014842,
      "loss": 0.1579,
      "mean_token_accuracy": 0.9369660496711731,
      "num_tokens": 300630.0,
      "step": 2580
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.025073856115341187,
      "learning_rate": 0.00014802,
      "loss": 0.1594,
      "mean_token_accuracy": 0.9384920775890351,
      "num_tokens": 302943.0,
      "step": 2600
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.02601858787238598,
      "learning_rate": 0.00014762,
      "loss": 0.1589,
      "mean_token_accuracy": 0.9373117983341217,
      "num_tokens": 305265.0,
      "step": 2620
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.0248605664819479,
      "learning_rate": 0.00014722,
      "loss": 0.1594,
      "mean_token_accuracy": 0.9345557481050492,
      "num_tokens": 307583.0,
      "step": 2640
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.022037120535969734,
      "learning_rate": 0.00014682,
      "loss": 0.1596,
      "mean_token_accuracy": 0.9348031014204026,
      "num_tokens": 309896.0,
      "step": 2660
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.03458873927593231,
      "learning_rate": 0.00014642,
      "loss": 0.1572,
      "mean_token_accuracy": 0.9376107037067414,
      "num_tokens": 312244.0,
      "step": 2680
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.016396528109908104,
      "learning_rate": 0.00014602,
      "loss": 0.1582,
      "mean_token_accuracy": 0.9370677560567856,
      "num_tokens": 314578.0,
      "step": 2700
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.01821085438132286,
      "learning_rate": 0.00014562,
      "loss": 0.1596,
      "mean_token_accuracy": 0.9343250393867493,
      "num_tokens": 316892.0,
      "step": 2720
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.025619324296712875,
      "learning_rate": 0.00014522,
      "loss": 0.1591,
      "mean_token_accuracy": 0.9362540364265441,
      "num_tokens": 319213.0,
      "step": 2740
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.01870078593492508,
      "learning_rate": 0.00014482,
      "loss": 0.1598,
      "mean_token_accuracy": 0.9338186293840408,
      "num_tokens": 321525.0,
      "step": 2760
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.018730677664279938,
      "learning_rate": 0.00014442,
      "loss": 0.1566,
      "mean_token_accuracy": 0.9394799619913101,
      "num_tokens": 323880.0,
      "step": 2780
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.019898803904652596,
      "learning_rate": 0.00014402,
      "loss": 0.1579,
      "mean_token_accuracy": 0.9384922862052918,
      "num_tokens": 326214.0,
      "step": 2800
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.021964257583022118,
      "learning_rate": 0.00014362,
      "loss": 0.1588,
      "mean_token_accuracy": 0.9378354996442795,
      "num_tokens": 328536.0,
      "step": 2820
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.02397042326629162,
      "learning_rate": 0.00014322,
      "loss": 0.157,
      "mean_token_accuracy": 0.9381829768419265,
      "num_tokens": 330888.0,
      "step": 2840
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.017819812521338463,
      "learning_rate": 0.00014282,
      "loss": 0.1569,
      "mean_token_accuracy": 0.940555801987648,
      "num_tokens": 333234.0,
      "step": 2860
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.019056344404816628,
      "learning_rate": 0.00014242,
      "loss": 0.1621,
      "mean_token_accuracy": 0.9321106940507888,
      "num_tokens": 335516.0,
      "step": 2880
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.020357482135295868,
      "learning_rate": 0.00014202,
      "loss": 0.1586,
      "mean_token_accuracy": 0.9384677648544312,
      "num_tokens": 337843.0,
      "step": 2900
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.02105647511780262,
      "learning_rate": 0.00014162,
      "loss": 0.1592,
      "mean_token_accuracy": 0.9367755681276322,
      "num_tokens": 340159.0,
      "step": 2920
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.02058851346373558,
      "learning_rate": 0.00014122,
      "loss": 0.1583,
      "mean_token_accuracy": 0.9383471548557282,
      "num_tokens": 342487.0,
      "step": 2940
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.022488698363304138,
      "learning_rate": 0.00014082,
      "loss": 0.1575,
      "mean_token_accuracy": 0.9383687317371369,
      "num_tokens": 344831.0,
      "step": 2960
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.01753912679851055,
      "learning_rate": 0.00014042,
      "loss": 0.1583,
      "mean_token_accuracy": 0.9367416232824326,
      "num_tokens": 347163.0,
      "step": 2980
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.01856599561870098,
      "learning_rate": 0.00014002,
      "loss": 0.156,
      "mean_token_accuracy": 0.9410459071397781,
      "num_tokens": 349528.0,
      "step": 3000
    }
  ],
  "logging_steps": 20,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4234514422376448.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}