{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 5925,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02531645569620253,
      "grad_norm": 0.2510090470314026,
      "learning_rate": 9.8e-05,
      "loss": 1.3649,
      "step": 50
    },
    {
      "epoch": 0.05063291139240506,
      "grad_norm": 0.17227400839328766,
      "learning_rate": 0.00019800000000000002,
      "loss": 0.1646,
      "step": 100
    },
    {
      "epoch": 0.0759493670886076,
      "grad_norm": 0.14617976546287537,
      "learning_rate": 0.00019831759656652363,
      "loss": 0.1474,
      "step": 150
    },
    {
      "epoch": 0.10126582278481013,
      "grad_norm": 0.14921978116035461,
      "learning_rate": 0.0001966008583690987,
      "loss": 0.1427,
      "step": 200
    },
    {
      "epoch": 0.12658227848101267,
      "grad_norm": 0.1575675755739212,
      "learning_rate": 0.00019488412017167384,
      "loss": 0.1425,
      "step": 250
    },
    {
      "epoch": 0.1518987341772152,
      "grad_norm": 0.15637104213237762,
      "learning_rate": 0.00019316738197424892,
      "loss": 0.1393,
      "step": 300
    },
    {
      "epoch": 0.17721518987341772,
      "grad_norm": 0.14315350353717804,
      "learning_rate": 0.00019145064377682405,
      "loss": 0.1401,
      "step": 350
    },
    {
      "epoch": 0.20253164556962025,
      "grad_norm": 0.15449273586273193,
      "learning_rate": 0.00018973390557939913,
      "loss": 0.1349,
      "step": 400
    },
    {
      "epoch": 0.22784810126582278,
      "grad_norm": 0.1331939995288849,
      "learning_rate": 0.00018801716738197426,
      "loss": 0.135,
      "step": 450
    },
    {
      "epoch": 0.25316455696202533,
      "grad_norm": 0.14222651720046997,
      "learning_rate": 0.00018630042918454937,
      "loss": 0.1371,
      "step": 500
    },
    {
      "epoch": 0.27848101265822783,
      "grad_norm": 0.1301516443490982,
      "learning_rate": 0.00018458369098712447,
      "loss": 0.1358,
      "step": 550
    },
    {
      "epoch": 0.3037974683544304,
      "grad_norm": 0.15532569587230682,
      "learning_rate": 0.00018286695278969958,
      "loss": 0.1379,
      "step": 600
    },
    {
      "epoch": 0.3291139240506329,
      "grad_norm": 0.13243968784809113,
      "learning_rate": 0.00018115021459227468,
      "loss": 0.1388,
      "step": 650
    },
    {
      "epoch": 0.35443037974683544,
      "grad_norm": 0.1436113715171814,
      "learning_rate": 0.00017943347639484979,
      "loss": 0.1381,
      "step": 700
    },
    {
      "epoch": 0.379746835443038,
      "grad_norm": 0.1405433863401413,
      "learning_rate": 0.0001777167381974249,
      "loss": 0.1327,
      "step": 750
    },
    {
      "epoch": 0.4050632911392405,
      "grad_norm": 0.13999509811401367,
      "learning_rate": 0.00017600000000000002,
      "loss": 0.1359,
      "step": 800
    },
    {
      "epoch": 0.43037974683544306,
      "grad_norm": 0.12786440551280975,
      "learning_rate": 0.00017428326180257513,
      "loss": 0.1348,
      "step": 850
    },
    {
      "epoch": 0.45569620253164556,
      "grad_norm": 0.15339377522468567,
      "learning_rate": 0.00017256652360515023,
      "loss": 0.1331,
      "step": 900
    },
    {
      "epoch": 0.4810126582278481,
      "grad_norm": 0.15028400719165802,
      "learning_rate": 0.00017084978540772534,
      "loss": 0.1351,
      "step": 950
    },
    {
      "epoch": 0.5063291139240507,
      "grad_norm": 0.13974353671073914,
      "learning_rate": 0.00016913304721030044,
      "loss": 0.1334,
      "step": 1000
    },
    {
      "epoch": 0.5316455696202531,
      "grad_norm": 0.16207924485206604,
      "learning_rate": 0.00016741630901287555,
      "loss": 0.1368,
      "step": 1050
    },
    {
      "epoch": 0.5569620253164557,
      "grad_norm": 0.16486062109470367,
      "learning_rate": 0.00016569957081545065,
      "loss": 0.1321,
      "step": 1100
    },
    {
      "epoch": 0.5822784810126582,
      "grad_norm": 0.17557835578918457,
      "learning_rate": 0.00016398283261802576,
      "loss": 0.1363,
      "step": 1150
    },
    {
      "epoch": 0.6075949367088608,
      "grad_norm": 0.1410171389579773,
      "learning_rate": 0.00016226609442060086,
      "loss": 0.1319,
      "step": 1200
    },
    {
      "epoch": 0.6329113924050633,
      "grad_norm": 0.1572587639093399,
      "learning_rate": 0.00016054935622317597,
      "loss": 0.1346,
      "step": 1250
    },
    {
      "epoch": 0.6582278481012658,
      "grad_norm": 0.1573045253753662,
      "learning_rate": 0.0001588326180257511,
      "loss": 0.136,
      "step": 1300
    },
    {
      "epoch": 0.6835443037974683,
      "grad_norm": 0.1620539426803589,
      "learning_rate": 0.00015711587982832618,
      "loss": 0.1286,
      "step": 1350
    },
    {
      "epoch": 0.7088607594936709,
      "grad_norm": 0.15751822292804718,
      "learning_rate": 0.0001553991416309013,
      "loss": 0.1317,
      "step": 1400
    },
    {
      "epoch": 0.7341772151898734,
      "grad_norm": 0.1773657500743866,
      "learning_rate": 0.0001536824034334764,
      "loss": 0.1319,
      "step": 1450
    },
    {
      "epoch": 0.759493670886076,
      "grad_norm": 0.14455097913742065,
      "learning_rate": 0.00015196566523605152,
      "loss": 0.1273,
      "step": 1500
    },
    {
      "epoch": 0.7848101265822784,
      "grad_norm": 0.15893496572971344,
      "learning_rate": 0.0001502489270386266,
      "loss": 0.1344,
      "step": 1550
    },
    {
      "epoch": 0.810126582278481,
      "grad_norm": 0.14044268429279327,
      "learning_rate": 0.00014853218884120173,
      "loss": 0.1278,
      "step": 1600
    },
    {
      "epoch": 0.8354430379746836,
      "grad_norm": 0.15982665121555328,
      "learning_rate": 0.00014681545064377683,
      "loss": 0.1329,
      "step": 1650
    },
    {
      "epoch": 0.8607594936708861,
      "grad_norm": 0.16502945125102997,
      "learning_rate": 0.00014509871244635194,
      "loss": 0.1291,
      "step": 1700
    },
    {
      "epoch": 0.8860759493670886,
      "grad_norm": 0.15228205919265747,
      "learning_rate": 0.00014338197424892704,
      "loss": 0.1321,
      "step": 1750
    },
    {
      "epoch": 0.9113924050632911,
      "grad_norm": 0.1459493339061737,
      "learning_rate": 0.00014166523605150215,
      "loss": 0.1298,
      "step": 1800
    },
    {
      "epoch": 0.9367088607594937,
      "grad_norm": 0.1607540249824524,
      "learning_rate": 0.00013994849785407725,
      "loss": 0.1301,
      "step": 1850
    },
    {
      "epoch": 0.9620253164556962,
      "grad_norm": 0.1760605126619339,
      "learning_rate": 0.00013823175965665236,
      "loss": 0.1291,
      "step": 1900
    },
    {
      "epoch": 0.9873417721518988,
      "grad_norm": 0.14624084532260895,
      "learning_rate": 0.00013651502145922746,
      "loss": 0.1303,
      "step": 1950
    },
    {
      "epoch": 1.0126582278481013,
      "grad_norm": 0.1362379789352417,
      "learning_rate": 0.00013479828326180257,
      "loss": 0.1173,
      "step": 2000
    },
    {
      "epoch": 1.0379746835443038,
      "grad_norm": 0.14143085479736328,
      "learning_rate": 0.0001330815450643777,
      "loss": 0.1036,
      "step": 2050
    },
    {
      "epoch": 1.0632911392405062,
      "grad_norm": 0.15834301710128784,
      "learning_rate": 0.0001313648068669528,
      "loss": 0.1026,
      "step": 2100
    },
    {
      "epoch": 1.0886075949367089,
      "grad_norm": 0.15556886792182922,
      "learning_rate": 0.0001296480686695279,
      "loss": 0.1044,
      "step": 2150
    },
    {
      "epoch": 1.1139240506329113,
      "grad_norm": 0.17505957186222076,
      "learning_rate": 0.00012793133047210302,
      "loss": 0.1029,
      "step": 2200
    },
    {
      "epoch": 1.139240506329114,
      "grad_norm": 0.16464762389659882,
      "learning_rate": 0.00012621459227467812,
      "loss": 0.1013,
      "step": 2250
    },
    {
      "epoch": 1.1645569620253164,
      "grad_norm": 0.17470216751098633,
      "learning_rate": 0.00012449785407725323,
      "loss": 0.1052,
      "step": 2300
    },
    {
      "epoch": 1.189873417721519,
      "grad_norm": 0.16448746621608734,
      "learning_rate": 0.00012278111587982833,
      "loss": 0.1039,
      "step": 2350
    },
    {
      "epoch": 1.2151898734177216,
      "grad_norm": 0.15587829053401947,
      "learning_rate": 0.00012106437768240344,
      "loss": 0.1035,
      "step": 2400
    },
    {
      "epoch": 1.240506329113924,
      "grad_norm": 0.1768651008605957,
      "learning_rate": 0.00011934763948497856,
      "loss": 0.1021,
      "step": 2450
    },
    {
      "epoch": 1.2658227848101267,
      "grad_norm": 0.1536482572555542,
      "learning_rate": 0.00011763090128755365,
      "loss": 0.1006,
      "step": 2500
    },
    {
      "epoch": 1.2911392405063291,
      "grad_norm": 0.16662020981311798,
      "learning_rate": 0.00011591416309012877,
      "loss": 0.1049,
      "step": 2550
    },
    {
      "epoch": 1.3164556962025316,
      "grad_norm": 0.16692157089710236,
      "learning_rate": 0.00011419742489270387,
      "loss": 0.1024,
      "step": 2600
    },
    {
      "epoch": 1.3417721518987342,
      "grad_norm": 0.17322711646556854,
      "learning_rate": 0.00011248068669527898,
      "loss": 0.0992,
      "step": 2650
    },
    {
      "epoch": 1.3670886075949367,
      "grad_norm": 0.16424308717250824,
      "learning_rate": 0.00011076394849785408,
      "loss": 0.1041,
      "step": 2700
    },
    {
      "epoch": 1.3924050632911391,
      "grad_norm": 0.17705556750297546,
      "learning_rate": 0.0001090472103004292,
      "loss": 0.1029,
      "step": 2750
    },
    {
      "epoch": 1.4177215189873418,
      "grad_norm": 0.19488489627838135,
      "learning_rate": 0.00010733047210300429,
      "loss": 0.1031,
      "step": 2800
    },
    {
      "epoch": 1.4430379746835442,
      "grad_norm": 0.1515897661447525,
      "learning_rate": 0.00010561373390557941,
      "loss": 0.1025,
      "step": 2850
    },
    {
      "epoch": 1.4683544303797469,
      "grad_norm": 0.17473047971725464,
      "learning_rate": 0.00010389699570815453,
      "loss": 0.103,
      "step": 2900
    },
    {
      "epoch": 1.4936708860759493,
      "grad_norm": 0.18762308359146118,
      "learning_rate": 0.00010218025751072962,
      "loss": 0.1041,
      "step": 2950
    },
    {
      "epoch": 1.518987341772152,
      "grad_norm": 0.17698685824871063,
      "learning_rate": 0.00010046351931330474,
      "loss": 0.1051,
      "step": 3000
    },
    {
      "epoch": 1.5443037974683544,
      "grad_norm": 0.14998982846736908,
      "learning_rate": 9.874678111587984e-05,
      "loss": 0.1033,
      "step": 3050
    },
    {
      "epoch": 1.5696202531645569,
      "grad_norm": 0.15886546671390533,
      "learning_rate": 9.703004291845495e-05,
      "loss": 0.1047,
      "step": 3100
    },
    {
      "epoch": 1.5949367088607596,
      "grad_norm": 0.16895173490047455,
      "learning_rate": 9.531330472103005e-05,
      "loss": 0.1021,
      "step": 3150
    },
    {
      "epoch": 1.620253164556962,
      "grad_norm": 0.19449743628501892,
      "learning_rate": 9.359656652360516e-05,
      "loss": 0.1029,
      "step": 3200
    },
    {
      "epoch": 1.6455696202531644,
      "grad_norm": 0.18704171478748322,
      "learning_rate": 9.187982832618026e-05,
      "loss": 0.1025,
      "step": 3250
    },
    {
      "epoch": 1.6708860759493671,
      "grad_norm": 0.18163681030273438,
      "learning_rate": 9.016309012875537e-05,
      "loss": 0.101,
      "step": 3300
    },
    {
      "epoch": 1.6962025316455698,
      "grad_norm": 0.16715924441814423,
      "learning_rate": 8.844635193133047e-05,
      "loss": 0.1024,
      "step": 3350
    },
    {
      "epoch": 1.721518987341772,
      "grad_norm": 0.1625606119632721,
      "learning_rate": 8.672961373390558e-05,
      "loss": 0.1014,
      "step": 3400
    },
    {
      "epoch": 1.7468354430379747,
      "grad_norm": 0.1782684028148651,
      "learning_rate": 8.50128755364807e-05,
      "loss": 0.1013,
      "step": 3450
    },
    {
      "epoch": 1.7721518987341773,
      "grad_norm": 0.1785702109336853,
      "learning_rate": 8.32961373390558e-05,
      "loss": 0.1002,
      "step": 3500
    },
    {
      "epoch": 1.7974683544303798,
      "grad_norm": 0.18035303056240082,
      "learning_rate": 8.15793991416309e-05,
      "loss": 0.1057,
      "step": 3550
    },
    {
      "epoch": 1.8227848101265822,
      "grad_norm": 0.18775810301303864,
      "learning_rate": 7.986266094420601e-05,
      "loss": 0.1026,
      "step": 3600
    },
    {
      "epoch": 1.8481012658227849,
      "grad_norm": 0.1495666652917862,
      "learning_rate": 7.814592274678112e-05,
      "loss": 0.1024,
      "step": 3650
    },
    {
      "epoch": 1.8734177215189873,
      "grad_norm": 0.1632242351770401,
      "learning_rate": 7.642918454935622e-05,
      "loss": 0.1028,
      "step": 3700
    },
    {
      "epoch": 1.8987341772151898,
      "grad_norm": 0.1862466186285019,
      "learning_rate": 7.471244635193133e-05,
      "loss": 0.1014,
      "step": 3750
    },
    {
      "epoch": 1.9240506329113924,
      "grad_norm": 0.18211400508880615,
      "learning_rate": 7.299570815450643e-05,
      "loss": 0.1014,
      "step": 3800
    },
    {
      "epoch": 1.9493670886075949,
      "grad_norm": 0.19410698115825653,
      "learning_rate": 7.127896995708155e-05,
      "loss": 0.1006,
      "step": 3850
    },
    {
      "epoch": 1.9746835443037973,
      "grad_norm": 0.15620633959770203,
      "learning_rate": 6.956223175965665e-05,
      "loss": 0.0983,
      "step": 3900
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.19822877645492554,
      "learning_rate": 6.784549356223176e-05,
      "loss": 0.1012,
      "step": 3950
    },
    {
      "epoch": 2.0253164556962027,
      "grad_norm": 0.13088855147361755,
      "learning_rate": 6.612875536480686e-05,
      "loss": 0.0726,
      "step": 4000
    },
    {
      "epoch": 2.050632911392405,
      "grad_norm": 0.14834293723106384,
      "learning_rate": 6.441201716738198e-05,
      "loss": 0.0748,
      "step": 4050
    },
    {
      "epoch": 2.0759493670886076,
      "grad_norm": 0.1417868286371231,
      "learning_rate": 6.269527896995709e-05,
      "loss": 0.0708,
      "step": 4100
    },
    {
      "epoch": 2.1012658227848102,
      "grad_norm": 0.14880657196044922,
      "learning_rate": 6.0978540772532187e-05,
      "loss": 0.0719,
      "step": 4150
    },
    {
      "epoch": 2.1265822784810124,
      "grad_norm": 0.17679744958877563,
      "learning_rate": 5.926180257510729e-05,
      "loss": 0.0749,
      "step": 4200
    },
    {
      "epoch": 2.151898734177215,
      "grad_norm": 0.1437467634677887,
      "learning_rate": 5.754506437768241e-05,
      "loss": 0.0754,
      "step": 4250
    },
    {
      "epoch": 2.1772151898734178,
      "grad_norm": 0.13949795067310333,
      "learning_rate": 5.5828326180257515e-05,
      "loss": 0.0715,
      "step": 4300
    },
    {
      "epoch": 2.2025316455696204,
      "grad_norm": 0.17855863273143768,
      "learning_rate": 5.411158798283262e-05,
      "loss": 0.073,
      "step": 4350
    },
    {
      "epoch": 2.2278481012658227,
      "grad_norm": 0.18084308505058289,
      "learning_rate": 5.239484978540773e-05,
      "loss": 0.0706,
      "step": 4400
    },
    {
      "epoch": 2.2531645569620253,
      "grad_norm": 0.14746320247650146,
      "learning_rate": 5.067811158798284e-05,
      "loss": 0.073,
      "step": 4450
    },
    {
      "epoch": 2.278481012658228,
      "grad_norm": 0.180175319314003,
      "learning_rate": 4.896137339055794e-05,
      "loss": 0.0755,
      "step": 4500
    },
    {
      "epoch": 2.3037974683544302,
      "grad_norm": 0.1923772394657135,
      "learning_rate": 4.7244635193133054e-05,
      "loss": 0.0746,
      "step": 4550
    },
    {
      "epoch": 2.329113924050633,
      "grad_norm": 0.1428052932024002,
      "learning_rate": 4.552789699570816e-05,
      "loss": 0.073,
      "step": 4600
    },
    {
      "epoch": 2.3544303797468356,
      "grad_norm": 0.16815955936908722,
      "learning_rate": 4.3811158798283264e-05,
      "loss": 0.0727,
      "step": 4650
    },
    {
      "epoch": 2.379746835443038,
      "grad_norm": 0.1623232066631317,
      "learning_rate": 4.209442060085837e-05,
      "loss": 0.0702,
      "step": 4700
    },
    {
      "epoch": 2.4050632911392404,
      "grad_norm": 0.14946827292442322,
      "learning_rate": 4.037768240343348e-05,
      "loss": 0.0721,
      "step": 4750
    },
    {
      "epoch": 2.430379746835443,
      "grad_norm": 0.14381594955921173,
      "learning_rate": 3.8660944206008585e-05,
      "loss": 0.0697,
      "step": 4800
    },
    {
      "epoch": 2.4556962025316453,
      "grad_norm": 0.19395038485527039,
      "learning_rate": 3.694420600858369e-05,
      "loss": 0.0748,
      "step": 4850
    },
    {
      "epoch": 2.481012658227848,
      "grad_norm": 0.18333947658538818,
      "learning_rate": 3.5227467811158795e-05,
      "loss": 0.0715,
      "step": 4900
    },
    {
      "epoch": 2.5063291139240507,
      "grad_norm": 0.17668786644935608,
      "learning_rate": 3.351072961373391e-05,
      "loss": 0.07,
      "step": 4950
    },
    {
      "epoch": 2.5316455696202533,
      "grad_norm": 0.2075100541114807,
      "learning_rate": 3.179399141630902e-05,
      "loss": 0.0737,
      "step": 5000
    },
    {
      "epoch": 2.5569620253164556,
      "grad_norm": 0.14891228079795837,
      "learning_rate": 3.007725321888412e-05,
      "loss": 0.0704,
      "step": 5050
    },
    {
      "epoch": 2.5822784810126582,
      "grad_norm": 0.17841076850891113,
      "learning_rate": 2.836051502145923e-05,
      "loss": 0.0724,
      "step": 5100
    },
    {
      "epoch": 2.607594936708861,
      "grad_norm": 0.184221088886261,
      "learning_rate": 2.6643776824034334e-05,
      "loss": 0.0712,
      "step": 5150
    },
    {
      "epoch": 2.632911392405063,
      "grad_norm": 0.16586528718471527,
      "learning_rate": 2.4927038626609442e-05,
      "loss": 0.0719,
      "step": 5200
    },
    {
      "epoch": 2.6582278481012658,
      "grad_norm": 0.19546498358249664,
      "learning_rate": 2.321030042918455e-05,
      "loss": 0.0728,
      "step": 5250
    },
    {
      "epoch": 2.6835443037974684,
      "grad_norm": 0.16904571652412415,
      "learning_rate": 2.1493562231759656e-05,
      "loss": 0.0712,
      "step": 5300
    },
    {
      "epoch": 2.708860759493671,
      "grad_norm": 0.14570939540863037,
      "learning_rate": 1.9776824034334764e-05,
      "loss": 0.0702,
      "step": 5350
    },
    {
      "epoch": 2.7341772151898733,
      "grad_norm": 0.1445421725511551,
      "learning_rate": 1.806008583690987e-05,
      "loss": 0.0694,
      "step": 5400
    },
    {
      "epoch": 2.759493670886076,
      "grad_norm": 0.19805610179901123,
      "learning_rate": 1.634334763948498e-05,
      "loss": 0.0728,
      "step": 5450
    },
    {
      "epoch": 2.7848101265822782,
      "grad_norm": 0.1762501746416092,
      "learning_rate": 1.4626609442060086e-05,
      "loss": 0.072,
      "step": 5500
    },
    {
      "epoch": 2.810126582278481,
      "grad_norm": 0.15009669959545135,
      "learning_rate": 1.2909871244635194e-05,
      "loss": 0.0703,
      "step": 5550
    },
    {
      "epoch": 2.8354430379746836,
      "grad_norm": 0.15821507573127747,
      "learning_rate": 1.1193133047210301e-05,
      "loss": 0.0727,
      "step": 5600
    },
    {
      "epoch": 2.8607594936708862,
      "grad_norm": 0.1787072718143463,
      "learning_rate": 9.476394849785408e-06,
      "loss": 0.0714,
      "step": 5650
    },
    {
      "epoch": 2.8860759493670884,
      "grad_norm": 0.18544761836528778,
      "learning_rate": 7.759656652360516e-06,
      "loss": 0.073,
      "step": 5700
    },
    {
      "epoch": 2.911392405063291,
      "grad_norm": 0.16766466200351715,
      "learning_rate": 6.042918454935623e-06,
      "loss": 0.0712,
      "step": 5750
    },
    {
      "epoch": 2.9367088607594938,
      "grad_norm": 0.15858995914459229,
      "learning_rate": 4.3261802575107296e-06,
      "loss": 0.0709,
      "step": 5800
    },
    {
      "epoch": 2.962025316455696,
      "grad_norm": 0.18093442916870117,
      "learning_rate": 2.609442060085837e-06,
      "loss": 0.0714,
      "step": 5850
    },
    {
      "epoch": 2.9873417721518987,
      "grad_norm": 0.13015761971473694,
      "learning_rate": 8.927038626609442e-07,
      "loss": 0.0695,
      "step": 5900
    }
  ],
  "logging_steps": 50,
  "max_steps": 5925,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.706065573669376e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}