{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.961636828644501,
  "eval_steps": 500,
  "global_step": 485,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010230179028132993,
      "grad_norm": 2.9682868161196363,
      "learning_rate": 3.2653061224489794e-06,
      "loss": 1.4918,
      "step": 1
    },
    {
      "epoch": 0.020460358056265986,
      "grad_norm": 2.9796653677109557,
      "learning_rate": 6.530612244897959e-06,
      "loss": 1.4922,
      "step": 2
    },
    {
      "epoch": 0.030690537084398978,
      "grad_norm": 2.8660911829024167,
      "learning_rate": 9.795918367346939e-06,
      "loss": 1.503,
      "step": 3
    },
    {
      "epoch": 0.04092071611253197,
      "grad_norm": 1.9805780511052813,
      "learning_rate": 1.3061224489795918e-05,
      "loss": 1.4675,
      "step": 4
    },
    {
      "epoch": 0.05115089514066496,
      "grad_norm": 1.998834758427541,
      "learning_rate": 1.63265306122449e-05,
      "loss": 1.4533,
      "step": 5
    },
    {
      "epoch": 0.061381074168797956,
      "grad_norm": 2.050049149995988,
      "learning_rate": 1.9591836734693877e-05,
      "loss": 1.4122,
      "step": 6
    },
    {
      "epoch": 0.07161125319693094,
      "grad_norm": 2.399845329221154,
      "learning_rate": 2.2857142857142858e-05,
      "loss": 1.4056,
      "step": 7
    },
    {
      "epoch": 0.08184143222506395,
      "grad_norm": 1.7232075850783724,
      "learning_rate": 2.6122448979591835e-05,
      "loss": 1.3833,
      "step": 8
    },
    {
      "epoch": 0.09207161125319693,
      "grad_norm": 1.410904151553637,
      "learning_rate": 2.938775510204082e-05,
      "loss": 1.3843,
      "step": 9
    },
    {
      "epoch": 0.10230179028132992,
      "grad_norm": 1.1900105733063466,
      "learning_rate": 3.26530612244898e-05,
      "loss": 1.3515,
      "step": 10
    },
    {
      "epoch": 0.11253196930946291,
      "grad_norm": 1.2160713923639859,
      "learning_rate": 3.591836734693878e-05,
      "loss": 1.3224,
      "step": 11
    },
    {
      "epoch": 0.12276214833759591,
      "grad_norm": 1.0389820905559795,
      "learning_rate": 3.9183673469387755e-05,
      "loss": 1.322,
      "step": 12
    },
    {
      "epoch": 0.1329923273657289,
      "grad_norm": 0.9606384846443378,
      "learning_rate": 4.244897959183674e-05,
      "loss": 1.2956,
      "step": 13
    },
    {
      "epoch": 0.1432225063938619,
      "grad_norm": 1.032658088356086,
      "learning_rate": 4.5714285714285716e-05,
      "loss": 1.3067,
      "step": 14
    },
    {
      "epoch": 0.1534526854219949,
      "grad_norm": 1.006086286942022,
      "learning_rate": 4.89795918367347e-05,
      "loss": 1.2961,
      "step": 15
    },
    {
      "epoch": 0.1636828644501279,
      "grad_norm": 1.1671471070443895,
      "learning_rate": 5.224489795918367e-05,
      "loss": 1.301,
      "step": 16
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 0.8672174854035504,
      "learning_rate": 5.551020408163266e-05,
      "loss": 1.2656,
      "step": 17
    },
    {
      "epoch": 0.18414322250639387,
      "grad_norm": 0.7154082384610629,
      "learning_rate": 5.877551020408164e-05,
      "loss": 1.246,
      "step": 18
    },
    {
      "epoch": 0.19437340153452684,
      "grad_norm": 0.694176256293125,
      "learning_rate": 6.204081632653062e-05,
      "loss": 1.2565,
      "step": 19
    },
    {
      "epoch": 0.20460358056265984,
      "grad_norm": 0.7538553851540746,
      "learning_rate": 6.53061224489796e-05,
      "loss": 1.241,
      "step": 20
    },
    {
      "epoch": 0.21483375959079284,
      "grad_norm": 0.7434156335821637,
      "learning_rate": 6.857142857142857e-05,
      "loss": 1.232,
      "step": 21
    },
    {
      "epoch": 0.22506393861892582,
      "grad_norm": 1.2022520877750795,
      "learning_rate": 7.183673469387756e-05,
      "loss": 1.2567,
      "step": 22
    },
    {
      "epoch": 0.23529411764705882,
      "grad_norm": 1.150528684217217,
      "learning_rate": 7.510204081632654e-05,
      "loss": 1.2448,
      "step": 23
    },
    {
      "epoch": 0.24552429667519182,
      "grad_norm": 0.8365697619839101,
      "learning_rate": 7.836734693877551e-05,
      "loss": 1.2357,
      "step": 24
    },
    {
      "epoch": 0.2557544757033248,
      "grad_norm": 1.1469314213901645,
      "learning_rate": 8.16326530612245e-05,
      "loss": 1.2164,
      "step": 25
    },
    {
      "epoch": 0.2659846547314578,
      "grad_norm": 0.9677681397432715,
      "learning_rate": 8.489795918367348e-05,
      "loss": 1.2115,
      "step": 26
    },
    {
      "epoch": 0.27621483375959077,
      "grad_norm": 1.106826210998024,
      "learning_rate": 8.816326530612245e-05,
      "loss": 1.2004,
      "step": 27
    },
    {
      "epoch": 0.2864450127877238,
      "grad_norm": 1.1744803496009477,
      "learning_rate": 9.142857142857143e-05,
      "loss": 1.2235,
      "step": 28
    },
    {
      "epoch": 0.2966751918158568,
      "grad_norm": 1.3610877277166293,
      "learning_rate": 9.469387755102041e-05,
      "loss": 1.208,
      "step": 29
    },
    {
      "epoch": 0.3069053708439898,
      "grad_norm": 0.7447757731907216,
      "learning_rate": 9.79591836734694e-05,
      "loss": 1.2118,
      "step": 30
    },
    {
      "epoch": 0.3171355498721228,
      "grad_norm": 1.3296162158652725,
      "learning_rate": 0.00010122448979591839,
      "loss": 1.2032,
      "step": 31
    },
    {
      "epoch": 0.3273657289002558,
      "grad_norm": 1.144264777414957,
      "learning_rate": 0.00010448979591836734,
      "loss": 1.1912,
      "step": 32
    },
    {
      "epoch": 0.3375959079283887,
      "grad_norm": 1.0342650793984924,
      "learning_rate": 0.00010775510204081634,
      "loss": 1.2025,
      "step": 33
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 1.9336445124108894,
      "learning_rate": 0.00011102040816326532,
      "loss": 1.2118,
      "step": 34
    },
    {
      "epoch": 0.35805626598465473,
      "grad_norm": 0.9195357493046576,
      "learning_rate": 0.0001142857142857143,
      "loss": 1.1986,
      "step": 35
    },
    {
      "epoch": 0.36828644501278773,
      "grad_norm": 1.588300333871633,
      "learning_rate": 0.00011755102040816328,
      "loss": 1.1939,
      "step": 36
    },
    {
      "epoch": 0.37851662404092073,
      "grad_norm": 1.4710948512542918,
      "learning_rate": 0.00012081632653061224,
      "loss": 1.1997,
      "step": 37
    },
    {
      "epoch": 0.3887468030690537,
      "grad_norm": 1.2164757524877552,
      "learning_rate": 0.00012408163265306124,
      "loss": 1.1791,
      "step": 38
    },
    {
      "epoch": 0.3989769820971867,
      "grad_norm": 1.4375602427525511,
      "learning_rate": 0.0001273469387755102,
      "loss": 1.1896,
      "step": 39
    },
    {
      "epoch": 0.4092071611253197,
      "grad_norm": 1.5736548612793597,
      "learning_rate": 0.0001306122448979592,
      "loss": 1.1876,
      "step": 40
    },
    {
      "epoch": 0.4194373401534527,
      "grad_norm": 1.0735537900578096,
      "learning_rate": 0.00013387755102040817,
      "loss": 1.1709,
      "step": 41
    },
    {
      "epoch": 0.4296675191815857,
      "grad_norm": 1.8510870420688348,
      "learning_rate": 0.00013714285714285713,
      "loss": 1.1795,
      "step": 42
    },
    {
      "epoch": 0.4398976982097187,
      "grad_norm": 1.1080619334489261,
      "learning_rate": 0.00014040816326530613,
      "loss": 1.1753,
      "step": 43
    },
    {
      "epoch": 0.45012787723785164,
      "grad_norm": 1.9521299298489614,
      "learning_rate": 0.00014367346938775512,
      "loss": 1.2028,
      "step": 44
    },
    {
      "epoch": 0.46035805626598464,
      "grad_norm": 1.41657060512908,
      "learning_rate": 0.0001469387755102041,
      "loss": 1.1822,
      "step": 45
    },
    {
      "epoch": 0.47058823529411764,
      "grad_norm": 1.1977389520948274,
      "learning_rate": 0.00015020408163265308,
      "loss": 1.1836,
      "step": 46
    },
    {
      "epoch": 0.48081841432225064,
      "grad_norm": 1.5288668435257222,
      "learning_rate": 0.00015346938775510205,
      "loss": 1.1901,
      "step": 47
    },
    {
      "epoch": 0.49104859335038364,
      "grad_norm": 0.9552067161901755,
      "learning_rate": 0.00015673469387755102,
      "loss": 1.1562,
      "step": 48
    },
    {
      "epoch": 0.5012787723785166,
      "grad_norm": 1.6547871414680237,
      "learning_rate": 0.00016,
      "loss": 1.1794,
      "step": 49
    },
    {
      "epoch": 0.5115089514066496,
      "grad_norm": 1.8368364441109388,
      "learning_rate": 0.00015999792324684382,
      "loss": 1.1877,
      "step": 50
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 1.1718896076543888,
      "learning_rate": 0.00015999169309519789,
      "loss": 1.1585,
      "step": 51
    },
    {
      "epoch": 0.5319693094629157,
      "grad_norm": 2.015243385621693,
      "learning_rate": 0.0001599813098685243,
      "loss": 1.1732,
      "step": 52
    },
    {
      "epoch": 0.5421994884910486,
      "grad_norm": 1.638588399609178,
      "learning_rate": 0.0001599667741059081,
      "loss": 1.1617,
      "step": 53
    },
    {
      "epoch": 0.5524296675191815,
      "grad_norm": 1.1352479883236335,
      "learning_rate": 0.00015994808656202904,
      "loss": 1.1584,
      "step": 54
    },
    {
      "epoch": 0.5626598465473146,
      "grad_norm": 1.2074962340041349,
      "learning_rate": 0.00015992524820712252,
      "loss": 1.1478,
      "step": 55
    },
    {
      "epoch": 0.5728900255754475,
      "grad_norm": 1.4410034215063858,
      "learning_rate": 0.00015989826022692918,
      "loss": 1.1493,
      "step": 56
    },
    {
      "epoch": 0.5831202046035806,
      "grad_norm": 0.8679551934822172,
      "learning_rate": 0.00015986712402263334,
      "loss": 1.148,
      "step": 57
    },
    {
      "epoch": 0.5933503836317136,
      "grad_norm": 1.3034239933412972,
      "learning_rate": 0.00015983184121079024,
      "loss": 1.1532,
      "step": 58
    },
    {
      "epoch": 0.6035805626598465,
      "grad_norm": 0.9780209051309727,
      "learning_rate": 0.00015979241362324223,
      "loss": 1.1313,
      "step": 59
    },
    {
      "epoch": 0.6138107416879796,
      "grad_norm": 1.057690857573897,
      "learning_rate": 0.0001597488433070234,
      "loss": 1.138,
      "step": 60
    },
    {
      "epoch": 0.6240409207161125,
      "grad_norm": 1.456389761257225,
      "learning_rate": 0.00015970113252425356,
      "loss": 1.1546,
      "step": 61
    },
    {
      "epoch": 0.6342710997442456,
      "grad_norm": 2.3692148862526485,
      "learning_rate": 0.00015964928375202068,
      "loss": 1.163,
      "step": 62
    },
    {
      "epoch": 0.6445012787723785,
      "grad_norm": 1.047447123295706,
      "learning_rate": 0.00015959329968225232,
      "loss": 1.1564,
      "step": 63
    },
    {
      "epoch": 0.6547314578005116,
      "grad_norm": 3.5383718747156534,
      "learning_rate": 0.0001595331832215758,
      "loss": 1.1684,
      "step": 64
    },
    {
      "epoch": 0.6649616368286445,
      "grad_norm": 2.968800505278846,
      "learning_rate": 0.00015946893749116734,
      "loss": 1.1788,
      "step": 65
    },
    {
      "epoch": 0.6751918158567775,
      "grad_norm": 1.9185023523212743,
      "learning_rate": 0.00015940056582659006,
      "loss": 1.1537,
      "step": 66
    },
    {
      "epoch": 0.6854219948849105,
      "grad_norm": 1.404723257524876,
      "learning_rate": 0.0001593280717776207,
      "loss": 1.1487,
      "step": 67
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 1.9443271708932357,
      "learning_rate": 0.0001592514591080654,
      "loss": 1.15,
      "step": 68
    },
    {
      "epoch": 0.7058823529411765,
      "grad_norm": 1.0325757666109745,
      "learning_rate": 0.0001591707317955642,
      "loss": 1.1487,
      "step": 69
    },
    {
      "epoch": 0.7161125319693095,
      "grad_norm": 2.1150324509757006,
      "learning_rate": 0.00015908589403138468,
      "loss": 1.1766,
      "step": 70
    },
    {
      "epoch": 0.7263427109974424,
      "grad_norm": 1.7663965705197828,
      "learning_rate": 0.00015899695022020415,
      "loss": 1.1464,
      "step": 71
    },
    {
      "epoch": 0.7365728900255755,
      "grad_norm": 1.4582930405137742,
      "learning_rate": 0.00015890390497988116,
      "loss": 1.1487,
      "step": 72
    },
    {
      "epoch": 0.7468030690537084,
      "grad_norm": 1.394985742880279,
      "learning_rate": 0.0001588067631412156,
      "loss": 1.1447,
      "step": 73
    },
    {
      "epoch": 0.7570332480818415,
      "grad_norm": 1.1897882316318342,
      "learning_rate": 0.000158705529747698,
      "loss": 1.1299,
      "step": 74
    },
    {
      "epoch": 0.7672634271099744,
      "grad_norm": 1.1626082561212874,
      "learning_rate": 0.0001586002100552476,
      "loss": 1.1473,
      "step": 75
    },
    {
      "epoch": 0.7774936061381074,
      "grad_norm": 1.0736945268719569,
      "learning_rate": 0.00015849080953193943,
      "loss": 1.1368,
      "step": 76
    },
    {
      "epoch": 0.7877237851662404,
      "grad_norm": 1.5654343283068946,
      "learning_rate": 0.00015837733385772062,
      "loss": 1.123,
      "step": 77
    },
    {
      "epoch": 0.7979539641943734,
      "grad_norm": 0.8523221287975669,
      "learning_rate": 0.00015825978892411522,
      "loss": 1.1403,
      "step": 78
    },
    {
      "epoch": 0.8081841432225064,
      "grad_norm": 1.035617914076512,
      "learning_rate": 0.00015813818083391858,
      "loss": 1.1229,
      "step": 79
    },
    {
      "epoch": 0.8184143222506394,
      "grad_norm": 1.4757207346915424,
      "learning_rate": 0.0001580125159008803,
      "loss": 1.129,
      "step": 80
    },
    {
      "epoch": 0.8286445012787724,
      "grad_norm": 1.0711086873355355,
      "learning_rate": 0.00015788280064937655,
      "loss": 1.1351,
      "step": 81
    },
    {
      "epoch": 0.8388746803069054,
      "grad_norm": 1.2110898160193528,
      "learning_rate": 0.00015774904181407127,
      "loss": 1.1334,
      "step": 82
    },
    {
      "epoch": 0.8491048593350383,
      "grad_norm": 1.2411189602993498,
      "learning_rate": 0.00015761124633956652,
      "loss": 1.1363,
      "step": 83
    },
    {
      "epoch": 0.8593350383631714,
      "grad_norm": 1.1341606871236802,
      "learning_rate": 0.00015746942138004203,
      "loss": 1.1142,
      "step": 84
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.9942900226210764,
      "learning_rate": 0.00015732357429888355,
      "loss": 1.13,
      "step": 85
    },
    {
      "epoch": 0.8797953964194374,
      "grad_norm": 1.1602205285909497,
      "learning_rate": 0.00015717371266830076,
      "loss": 1.1225,
      "step": 86
    },
    {
      "epoch": 0.8900255754475703,
      "grad_norm": 1.2773792554197163,
      "learning_rate": 0.000157019844268934,
      "loss": 1.1237,
      "step": 87
    },
    {
      "epoch": 0.9002557544757033,
      "grad_norm": 1.0698753174294156,
      "learning_rate": 0.0001568619770894504,
      "loss": 1.1223,
      "step": 88
    },
    {
      "epoch": 0.9104859335038363,
      "grad_norm": 1.1279028063380314,
      "learning_rate": 0.000156700119326129,
      "loss": 1.1117,
      "step": 89
    },
    {
      "epoch": 0.9207161125319693,
      "grad_norm": 1.047346608848377,
      "learning_rate": 0.00015653427938243532,
      "loss": 1.1195,
      "step": 90
    },
    {
      "epoch": 0.9309462915601023,
      "grad_norm": 1.1063879184523018,
      "learning_rate": 0.0001563644658685851,
      "loss": 1.1112,
      "step": 91
    },
    {
      "epoch": 0.9411764705882353,
      "grad_norm": 1.403473791127022,
      "learning_rate": 0.00015619068760109703,
      "loss": 1.1334,
      "step": 92
    },
    {
      "epoch": 0.9514066496163683,
      "grad_norm": 1.1738079096105647,
      "learning_rate": 0.00015601295360233528,
      "loss": 1.123,
      "step": 93
    },
    {
      "epoch": 0.9616368286445013,
      "grad_norm": 1.2622400384084083,
      "learning_rate": 0.0001558312731000409,
      "loss": 1.1245,
      "step": 94
    },
    {
      "epoch": 0.9718670076726342,
      "grad_norm": 1.5103980615396817,
      "learning_rate": 0.00015564565552685282,
      "loss": 1.1159,
      "step": 95
    },
    {
      "epoch": 0.9820971867007673,
      "grad_norm": 0.9924832092739461,
      "learning_rate": 0.00015545611051981807,
      "loss": 1.1086,
      "step": 96
    },
    {
      "epoch": 0.9923273657289002,
      "grad_norm": 1.2306800508291864,
      "learning_rate": 0.00015526264791989144,
      "loss": 1.1396,
      "step": 97
    },
    {
      "epoch": 1.0025575447570332,
      "grad_norm": 1.7155461125801181,
      "learning_rate": 0.00015506527777142453,
      "loss": 1.4022,
      "step": 98
    },
    {
      "epoch": 1.0127877237851663,
      "grad_norm": 1.0329131057981322,
      "learning_rate": 0.00015486401032164434,
      "loss": 1.0962,
      "step": 99
    },
    {
      "epoch": 1.0230179028132993,
      "grad_norm": 1.5658900134919402,
      "learning_rate": 0.00015465885602012117,
      "loss": 1.1252,
      "step": 100
    },
    {
      "epoch": 1.0332480818414322,
      "grad_norm": 0.8015741030349912,
      "learning_rate": 0.00015444982551822604,
      "loss": 1.1044,
      "step": 101
    },
    {
      "epoch": 1.0434782608695652,
      "grad_norm": 1.3138293387151978,
      "learning_rate": 0.00015423692966857788,
      "loss": 1.1138,
      "step": 102
    },
    {
      "epoch": 1.0537084398976981,
      "grad_norm": 0.8910062603103389,
      "learning_rate": 0.00015402017952447983,
      "loss": 1.0804,
      "step": 103
    },
    {
      "epoch": 1.0639386189258313,
      "grad_norm": 1.4375718907843533,
      "learning_rate": 0.00015379958633934555,
      "loss": 1.1212,
      "step": 104
    },
    {
      "epoch": 1.0741687979539642,
      "grad_norm": 1.2721636337229971,
      "learning_rate": 0.0001535751615661149,
      "loss": 1.1,
      "step": 105
    },
    {
      "epoch": 1.0843989769820972,
      "grad_norm": 1.0105810629121363,
      "learning_rate": 0.00015334691685665928,
      "loss": 1.096,
      "step": 106
    },
    {
      "epoch": 1.0946291560102301,
      "grad_norm": 1.0900734221547557,
      "learning_rate": 0.00015311486406117668,
      "loss": 1.0882,
      "step": 107
    },
    {
      "epoch": 1.104859335038363,
      "grad_norm": 1.1336612336243286,
      "learning_rate": 0.00015287901522757652,
      "loss": 1.1214,
      "step": 108
    },
    {
      "epoch": 1.1150895140664963,
      "grad_norm": 1.1569430105162481,
      "learning_rate": 0.000152639382600854,
      "loss": 1.0963,
      "step": 109
    },
    {
      "epoch": 1.1253196930946292,
      "grad_norm": 1.1569271586356937,
      "learning_rate": 0.00015239597862245452,
      "loss": 1.0855,
      "step": 110
    },
    {
      "epoch": 1.1355498721227621,
      "grad_norm": 1.137534239049028,
      "learning_rate": 0.00015214881592962753,
      "loss": 1.094,
      "step": 111
    },
    {
      "epoch": 1.145780051150895,
      "grad_norm": 1.2673351118799951,
      "learning_rate": 0.00015189790735477062,
      "loss": 1.0819,
      "step": 112
    },
    {
      "epoch": 1.156010230179028,
      "grad_norm": 0.8230868750468249,
      "learning_rate": 0.00015164326592476316,
      "loss": 1.099,
      "step": 113
    },
    {
      "epoch": 1.1662404092071612,
      "grad_norm": 1.0550234081618097,
      "learning_rate": 0.00015138490486028998,
      "loss": 1.0887,
      "step": 114
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 1.2808691339484497,
      "learning_rate": 0.000151122837575155,
      "loss": 1.0879,
      "step": 115
    },
    {
      "epoch": 1.186700767263427,
      "grad_norm": 1.0207057907200907,
      "learning_rate": 0.00015085707767558475,
      "loss": 1.0974,
      "step": 116
    },
    {
      "epoch": 1.19693094629156,
      "grad_norm": 1.7224394340286069,
      "learning_rate": 0.00015058763895952194,
      "loss": 1.1016,
      "step": 117
    },
    {
      "epoch": 1.207161125319693,
      "grad_norm": 0.5474668831444041,
      "learning_rate": 0.00015031453541590925,
      "loss": 1.0789,
      "step": 118
    },
    {
      "epoch": 1.2173913043478262,
      "grad_norm": 1.2675844883943275,
      "learning_rate": 0.00015003778122396277,
      "loss": 1.0851,
      "step": 119
    },
    {
      "epoch": 1.227621483375959,
      "grad_norm": 1.2798593237994211,
      "learning_rate": 0.0001497573907524361,
      "loss": 1.1047,
      "step": 120
    },
    {
      "epoch": 1.237851662404092,
      "grad_norm": 1.1502336899283057,
      "learning_rate": 0.00014947337855887406,
      "loss": 1.0943,
      "step": 121
    },
    {
      "epoch": 1.248081841432225,
      "grad_norm": 1.2423026547868323,
      "learning_rate": 0.00014918575938885725,
      "loss": 1.0896,
      "step": 122
    },
    {
      "epoch": 1.258312020460358,
      "grad_norm": 0.8127468643676744,
      "learning_rate": 0.00014889454817523608,
      "loss": 1.0984,
      "step": 123
    },
    {
      "epoch": 1.2685421994884911,
      "grad_norm": 1.1236998604735062,
      "learning_rate": 0.00014859976003735572,
      "loss": 1.091,
      "step": 124
    },
    {
      "epoch": 1.278772378516624,
      "grad_norm": 1.0625241691652207,
      "learning_rate": 0.0001483014102802711,
      "loss": 1.0749,
      "step": 125
    },
    {
      "epoch": 1.289002557544757,
      "grad_norm": 1.2427086503303664,
      "learning_rate": 0.00014799951439395221,
      "loss": 1.0901,
      "step": 126
    },
    {
      "epoch": 1.29923273657289,
      "grad_norm": 0.9886108496701173,
      "learning_rate": 0.00014769408805247986,
      "loss": 1.0848,
      "step": 127
    },
    {
      "epoch": 1.309462915601023,
      "grad_norm": 1.017219160331291,
      "learning_rate": 0.0001473851471132321,
      "loss": 1.0897,
      "step": 128
    },
    {
      "epoch": 1.319693094629156,
      "grad_norm": 1.4542703012009928,
      "learning_rate": 0.00014707270761606063,
      "loss": 1.0695,
      "step": 129
    },
    {
      "epoch": 1.329923273657289,
      "grad_norm": 1.0782706459189146,
      "learning_rate": 0.00014675678578245828,
      "loss": 1.0895,
      "step": 130
    },
    {
      "epoch": 1.340153452685422,
      "grad_norm": 1.0968316967604133,
      "learning_rate": 0.00014643739801471667,
      "loss": 1.1003,
      "step": 131
    },
    {
      "epoch": 1.350383631713555,
      "grad_norm": 0.9003561398004416,
      "learning_rate": 0.00014611456089507464,
      "loss": 1.098,
      "step": 132
    },
    {
      "epoch": 1.3606138107416879,
      "grad_norm": 1.351034619458769,
      "learning_rate": 0.00014578829118485742,
      "loss": 1.0698,
      "step": 133
    },
    {
      "epoch": 1.370843989769821,
      "grad_norm": 0.8658183503538551,
      "learning_rate": 0.00014545860582360624,
      "loss": 1.1071,
      "step": 134
    },
    {
      "epoch": 1.381074168797954,
      "grad_norm": 1.0177123952310823,
      "learning_rate": 0.00014512552192819897,
      "loss": 1.0869,
      "step": 135
    },
    {
      "epoch": 1.391304347826087,
      "grad_norm": 1.4195927565971158,
      "learning_rate": 0.0001447890567919614,
      "loss": 1.0954,
      "step": 136
    },
    {
      "epoch": 1.40153452685422,
      "grad_norm": 0.9931491307968467,
      "learning_rate": 0.00014444922788376934,
      "loss": 1.0784,
      "step": 137
    },
    {
      "epoch": 1.4117647058823528,
      "grad_norm": 1.270639141634098,
      "learning_rate": 0.00014410605284714175,
      "loss": 1.0888,
      "step": 138
    },
    {
      "epoch": 1.421994884910486,
      "grad_norm": 0.8408471380854269,
      "learning_rate": 0.0001437595494993246,
      "loss": 1.0842,
      "step": 139
    },
    {
      "epoch": 1.432225063938619,
      "grad_norm": 1.380164251857079,
      "learning_rate": 0.000143409735830366,
      "loss": 1.0795,
      "step": 140
    },
    {
      "epoch": 1.4424552429667519,
      "grad_norm": 0.9896679635139878,
      "learning_rate": 0.00014305663000218193,
      "loss": 1.0907,
      "step": 141
    },
    {
      "epoch": 1.452685421994885,
      "grad_norm": 1.412072301199199,
      "learning_rate": 0.00014270025034761352,
      "loss": 1.0817,
      "step": 142
    },
    {
      "epoch": 1.4629156010230178,
      "grad_norm": 0.770940428011719,
      "learning_rate": 0.000142340615369475,
      "loss": 1.0819,
      "step": 143
    },
    {
      "epoch": 1.473145780051151,
      "grad_norm": 0.7983918736084521,
      "learning_rate": 0.00014197774373959327,
      "loss": 1.0931,
      "step": 144
    },
    {
      "epoch": 1.4833759590792839,
      "grad_norm": 1.3538690920598053,
      "learning_rate": 0.00014161165429783844,
      "loss": 1.0884,
      "step": 145
    },
    {
      "epoch": 1.4936061381074168,
      "grad_norm": 0.8169757443568035,
      "learning_rate": 0.0001412423660511456,
      "loss": 1.0924,
      "step": 146
    },
    {
      "epoch": 1.50383631713555,
      "grad_norm": 0.9256987946838048,
      "learning_rate": 0.00014086989817252803,
      "loss": 1.0785,
      "step": 147
    },
    {
      "epoch": 1.5140664961636827,
      "grad_norm": 1.1631633494660514,
      "learning_rate": 0.00014049427000008185,
      "loss": 1.0699,
      "step": 148
    },
    {
      "epoch": 1.5242966751918159,
      "grad_norm": 1.212333243880141,
      "learning_rate": 0.00014011550103598176,
      "loss": 1.064,
      "step": 149
    },
    {
      "epoch": 1.5345268542199488,
      "grad_norm": 0.9967182911949714,
      "learning_rate": 0.0001397336109454689,
      "loss": 1.1002,
      "step": 150
    },
    {
      "epoch": 1.5447570332480818,
      "grad_norm": 1.1502313981018193,
      "learning_rate": 0.0001393486195558295,
      "loss": 1.0709,
      "step": 151
    },
    {
      "epoch": 1.554987212276215,
      "grad_norm": 1.311911957474106,
      "learning_rate": 0.00013896054685536566,
      "loss": 1.0717,
      "step": 152
    },
    {
      "epoch": 1.5652173913043477,
      "grad_norm": 0.8141917822120857,
      "learning_rate": 0.00013856941299235752,
      "loss": 1.0714,
      "step": 153
    },
    {
      "epoch": 1.5754475703324808,
      "grad_norm": 0.5826303398204319,
      "learning_rate": 0.00013817523827401715,
      "loss": 1.0825,
      "step": 154
    },
    {
      "epoch": 1.5856777493606138,
      "grad_norm": 1.1035174246946824,
      "learning_rate": 0.00013777804316543438,
      "loss": 1.0583,
      "step": 155
    },
    {
      "epoch": 1.5959079283887467,
      "grad_norm": 1.483439389847773,
      "learning_rate": 0.00013737784828851405,
      "loss": 1.0998,
      "step": 156
    },
    {
      "epoch": 1.60613810741688,
      "grad_norm": 0.5590063439764236,
      "learning_rate": 0.0001369746744209055,
      "loss": 1.0814,
      "step": 157
    },
    {
      "epoch": 1.6163682864450126,
      "grad_norm": 1.427302602685382,
      "learning_rate": 0.00013656854249492382,
      "loss": 1.0672,
      "step": 158
    },
    {
      "epoch": 1.6265984654731458,
      "grad_norm": 0.9907320262762309,
      "learning_rate": 0.00013615947359646295,
      "loss": 1.077,
      "step": 159
    },
    {
      "epoch": 1.6368286445012787,
      "grad_norm": 1.3253726577143705,
      "learning_rate": 0.00013574748896390105,
      "loss": 1.0831,
      "step": 160
    },
    {
      "epoch": 1.6470588235294117,
      "grad_norm": 0.6170129994977172,
      "learning_rate": 0.00013533260998699776,
      "loss": 1.0808,
      "step": 161
    },
    {
      "epoch": 1.6572890025575449,
      "grad_norm": 0.8974715249718852,
      "learning_rate": 0.00013491485820578373,
      "loss": 1.0609,
      "step": 162
    },
    {
      "epoch": 1.6675191815856778,
      "grad_norm": 0.8218782327367921,
      "learning_rate": 0.00013449425530944218,
      "loss": 1.0822,
      "step": 163
    },
    {
      "epoch": 1.6777493606138107,
      "grad_norm": 0.6783603661700464,
      "learning_rate": 0.00013407082313518292,
      "loss": 1.0771,
      "step": 164
    },
    {
      "epoch": 1.6879795396419437,
      "grad_norm": 0.6976792612146404,
      "learning_rate": 0.0001336445836671086,
      "loss": 1.0853,
      "step": 165
    },
    {
      "epoch": 1.6982097186700766,
      "grad_norm": 0.823428019130252,
      "learning_rate": 0.0001332155590350732,
      "loss": 1.0838,
      "step": 166
    },
    {
      "epoch": 1.7084398976982098,
      "grad_norm": 1.0953340521038164,
      "learning_rate": 0.0001327837715135332,
      "loss": 1.082,
      "step": 167
    },
    {
      "epoch": 1.7186700767263428,
      "grad_norm": 0.9474195557921864,
      "learning_rate": 0.00013234924352039103,
      "loss": 1.0802,
      "step": 168
    },
    {
      "epoch": 1.7289002557544757,
      "grad_norm": 0.9444470757389084,
      "learning_rate": 0.00013191199761583124,
      "loss": 1.0887,
      "step": 169
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 1.101212086185109,
      "learning_rate": 0.00013147205650114913,
      "loss": 1.0718,
      "step": 170
    },
    {
      "epoch": 1.7493606138107416,
      "grad_norm": 1.0553531873256305,
      "learning_rate": 0.0001310294430175722,
      "loss": 1.0788,
      "step": 171
    },
    {
      "epoch": 1.7595907928388748,
      "grad_norm": 1.05582162764911,
      "learning_rate": 0.00013058418014507412,
      "loss": 1.0879,
      "step": 172
    },
    {
      "epoch": 1.7698209718670077,
      "grad_norm": 0.9944536264830629,
      "learning_rate": 0.00013013629100118183,
      "loss": 1.0721,
      "step": 173
    },
    {
      "epoch": 1.7800511508951407,
      "grad_norm": 1.1307659477843484,
      "learning_rate": 0.00012968579883977508,
      "loss": 1.0737,
      "step": 174
    },
    {
      "epoch": 1.7902813299232738,
      "grad_norm": 1.2420812191214468,
      "learning_rate": 0.00012923272704987943,
      "loss": 1.0742,
      "step": 175
    },
    {
      "epoch": 1.8005115089514065,
      "grad_norm": 1.0472853919924061,
      "learning_rate": 0.00012877709915445155,
      "loss": 1.0721,
      "step": 176
    },
    {
      "epoch": 1.8107416879795397,
      "grad_norm": 0.6947884508219546,
      "learning_rate": 0.00012831893880915822,
      "loss": 1.0555,
      "step": 177
    },
    {
      "epoch": 1.8209718670076727,
      "grad_norm": 0.7757404896202875,
      "learning_rate": 0.00012785826980114798,
      "loss": 1.0804,
      "step": 178
    },
    {
      "epoch": 1.8312020460358056,
      "grad_norm": 1.253778618594067,
      "learning_rate": 0.0001273951160478163,
      "loss": 1.063,
      "step": 179
    },
    {
      "epoch": 1.8414322250639388,
      "grad_norm": 0.5979958514242996,
      "learning_rate": 0.00012692950159556358,
      "loss": 1.0666,
      "step": 180
    },
    {
      "epoch": 1.8516624040920715,
      "grad_norm": 0.8662747748641294,
      "learning_rate": 0.00012646145061854697,
      "loss": 1.0703,
      "step": 181
    },
    {
      "epoch": 1.8618925831202047,
      "grad_norm": 0.8287006788806506,
      "learning_rate": 0.00012599098741742504,
      "loss": 1.0571,
      "step": 182
    },
    {
      "epoch": 1.8721227621483376,
      "grad_norm": 1.0265713289334373,
      "learning_rate": 0.00012551813641809622,
      "loss": 1.0706,
      "step": 183
    },
    {
      "epoch": 1.8823529411764706,
      "grad_norm": 1.2306509484226458,
      "learning_rate": 0.0001250429221704306,
      "loss": 1.0779,
      "step": 184
    },
    {
      "epoch": 1.8925831202046037,
      "grad_norm": 0.8951251773522786,
      "learning_rate": 0.00012456536934699552,
      "loss": 1.064,
      "step": 185
    },
    {
      "epoch": 1.9028132992327365,
      "grad_norm": 1.1181307738147266,
      "learning_rate": 0.0001240855027417742,
      "loss": 1.0585,
      "step": 186
    },
    {
      "epoch": 1.9130434782608696,
      "grad_norm": 0.977542614018667,
      "learning_rate": 0.00012360334726887887,
      "loss": 1.0672,
      "step": 187
    },
    {
      "epoch": 1.9232736572890026,
      "grad_norm": 1.1637804133002638,
      "learning_rate": 0.00012311892796125704,
      "loss": 1.0713,
      "step": 188
    },
    {
      "epoch": 1.9335038363171355,
      "grad_norm": 0.445419436951453,
      "learning_rate": 0.0001226322699693918,
      "loss": 1.0536,
      "step": 189
    },
    {
      "epoch": 1.9437340153452687,
      "grad_norm": 0.9034412973205753,
      "learning_rate": 0.00012214339855999624,
      "loss": 1.0807,
      "step": 190
    },
    {
      "epoch": 1.9539641943734014,
      "grad_norm": 0.5236086539347802,
      "learning_rate": 0.00012165233911470136,
      "loss": 1.0777,
      "step": 191
    },
    {
      "epoch": 1.9641943734015346,
      "grad_norm": 0.6745807390622581,
      "learning_rate": 0.00012115911712873851,
      "loss": 1.0525,
      "step": 192
    },
    {
      "epoch": 1.9744245524296675,
      "grad_norm": 0.5551428173478703,
      "learning_rate": 0.00012066375820961558,
      "loss": 1.0617,
      "step": 193
    },
    {
      "epoch": 1.9846547314578005,
      "grad_norm": 0.5930456521809883,
      "learning_rate": 0.00012016628807578756,
      "loss": 1.0682,
      "step": 194
    },
    {
      "epoch": 1.9948849104859336,
      "grad_norm": 0.5333039982670191,
      "learning_rate": 0.00011966673255532119,
      "loss": 1.1518,
      "step": 195
    },
    {
      "epoch": 2.0051150895140664,
      "grad_norm": 0.7368227183464856,
      "learning_rate": 0.00011916511758455407,
      "loss": 1.226,
      "step": 196
    },
    {
      "epoch": 2.0153452685421995,
      "grad_norm": 1.0538129513540742,
      "learning_rate": 0.00011866146920674807,
      "loss": 1.068,
      "step": 197
    },
    {
      "epoch": 2.0255754475703327,
      "grad_norm": 1.2094094160759923,
      "learning_rate": 0.0001181558135707371,
      "loss": 1.0502,
      "step": 198
    },
    {
      "epoch": 2.0358056265984654,
      "grad_norm": 0.915224534709575,
      "learning_rate": 0.00011764817692956966,
      "loss": 1.0286,
      "step": 199
    },
    {
      "epoch": 2.0460358056265986,
      "grad_norm": 1.0711022599910567,
      "learning_rate": 0.00011713858563914562,
      "loss": 1.0747,
      "step": 200
    },
    {
      "epoch": 2.0562659846547313,
      "grad_norm": 1.0787442473268498,
      "learning_rate": 0.00011662706615684803,
      "loss": 1.045,
      "step": 201
    },
    {
      "epoch": 2.0664961636828645,
      "grad_norm": 0.8778392977870085,
      "learning_rate": 0.00011611364504016935,
      "loss": 1.0678,
      "step": 202
    },
    {
      "epoch": 2.0767263427109977,
      "grad_norm": 0.9910713045345557,
      "learning_rate": 0.00011559834894533275,
      "loss": 1.0458,
      "step": 203
    },
    {
      "epoch": 2.0869565217391304,
      "grad_norm": 1.1592891619781418,
      "learning_rate": 0.00011508120462590794,
      "loss": 1.0461,
      "step": 204
    },
    {
      "epoch": 2.0971867007672635,
      "grad_norm": 0.6931145505443533,
      "learning_rate": 0.00011456223893142238,
      "loss": 1.0407,
      "step": 205
    },
    {
      "epoch": 2.1074168797953963,
      "grad_norm": 0.8009294768801537,
      "learning_rate": 0.0001140414788059672,
      "loss": 1.0534,
      "step": 206
    },
    {
      "epoch": 2.1176470588235294,
      "grad_norm": 0.6527102431957467,
      "learning_rate": 0.00011351895128679823,
      "loss": 1.0577,
      "step": 207
    },
    {
      "epoch": 2.1278772378516626,
      "grad_norm": 0.5808810881566437,
      "learning_rate": 0.00011299468350293232,
      "loss": 1.0592,
      "step": 208
    },
    {
      "epoch": 2.1381074168797953,
      "grad_norm": 0.8896635325022703,
      "learning_rate": 0.00011246870267373885,
      "loss": 1.069,
      "step": 209
    },
    {
      "epoch": 2.1483375959079285,
      "grad_norm": 1.0873242378114067,
      "learning_rate": 0.00011194103610752655,
      "loss": 1.0454,
      "step": 210
    },
    {
      "epoch": 2.1585677749360612,
      "grad_norm": 0.9489754319671625,
      "learning_rate": 0.00011141171120012552,
      "loss": 1.0723,
      "step": 211
    },
    {
      "epoch": 2.1687979539641944,
      "grad_norm": 0.8030522232247485,
      "learning_rate": 0.0001108807554334651,
      "loss": 1.0428,
      "step": 212
    },
    {
      "epoch": 2.1790281329923276,
      "grad_norm": 0.5529249514757153,
      "learning_rate": 0.00011034819637414686,
      "loss": 1.061,
      "step": 213
    },
    {
      "epoch": 2.1892583120204603,
      "grad_norm": 0.4631583007632592,
      "learning_rate": 0.00010981406167201354,
      "loss": 1.0355,
      "step": 214
    },
    {
      "epoch": 2.1994884910485935,
      "grad_norm": 0.4099197978603843,
      "learning_rate": 0.0001092783790587133,
      "loss": 1.0777,
      "step": 215
    },
    {
      "epoch": 2.209718670076726,
      "grad_norm": 0.4121757237952397,
      "learning_rate": 0.00010874117634626011,
      "loss": 1.0541,
      "step": 216
    },
    {
      "epoch": 2.2199488491048593,
      "grad_norm": 0.4349846280052258,
      "learning_rate": 0.00010820248142558965,
      "loss": 1.0435,
      "step": 217
    },
    {
      "epoch": 2.2301790281329925,
      "grad_norm": 0.3960108429870153,
      "learning_rate": 0.00010766232226511142,
      "loss": 1.0513,
      "step": 218
    },
    {
      "epoch": 2.2404092071611252,
      "grad_norm": 0.4522590868845693,
      "learning_rate": 0.00010712072690925638,
      "loss": 1.0509,
      "step": 219
    },
    {
      "epoch": 2.2506393861892584,
      "grad_norm": 0.4862701100635931,
      "learning_rate": 0.00010657772347702118,
      "loss": 1.0325,
      "step": 220
    },
    {
      "epoch": 2.260869565217391,
      "grad_norm": 0.44763620371165586,
      "learning_rate": 0.00010603334016050808,
      "loss": 1.0369,
      "step": 221
    },
    {
      "epoch": 2.2710997442455243,
      "grad_norm": 0.5580650650662916,
      "learning_rate": 0.00010548760522346138,
      "loss": 1.0414,
      "step": 222
    },
    {
      "epoch": 2.2813299232736575,
      "grad_norm": 0.6352266377825914,
      "learning_rate": 0.00010494054699979992,
      "loss": 1.056,
      "step": 223
    },
    {
      "epoch": 2.29156010230179,
      "grad_norm": 0.617790095766774,
      "learning_rate": 0.00010439219389214595,
      "loss": 1.0573,
      "step": 224
    },
    {
      "epoch": 2.3017902813299234,
      "grad_norm": 0.6417970276744828,
      "learning_rate": 0.0001038425743703507,
      "loss": 1.0412,
      "step": 225
    },
    {
      "epoch": 2.312020460358056,
      "grad_norm": 0.7082597904375986,
      "learning_rate": 0.00010329171697001608,
      "loss": 1.0366,
      "step": 226
    },
    {
      "epoch": 2.3222506393861893,
      "grad_norm": 0.8496266841221832,
      "learning_rate": 0.0001027396502910132,
      "loss": 1.0451,
      "step": 227
    },
    {
      "epoch": 2.3324808184143224,
      "grad_norm": 1.0308330001421908,
      "learning_rate": 0.0001021864029959975,
      "loss": 1.0428,
      "step": 228
    },
    {
      "epoch": 2.342710997442455,
      "grad_norm": 1.026355847644825,
      "learning_rate": 0.00010163200380892063,
      "loss": 1.0612,
      "step": 229
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 0.7560985076254375,
      "learning_rate": 0.00010107648151353916,
      "loss": 1.0247,
      "step": 230
    },
    {
      "epoch": 2.363171355498721,
      "grad_norm": 0.41561156978885305,
      "learning_rate": 0.00010051986495192008,
      "loss": 1.0363,
      "step": 231
    },
    {
      "epoch": 2.373401534526854,
      "grad_norm": 0.38675030483907014,
      "learning_rate": 9.99621830229434e-05,
      "loss": 1.05,
      "step": 232
    },
    {
      "epoch": 2.3836317135549874,
      "grad_norm": 0.4804093390079753,
      "learning_rate": 9.94034646808018e-05,
      "loss": 1.0537,
      "step": 233
    },
    {
      "epoch": 2.39386189258312,
      "grad_norm": 0.5877095225822028,
      "learning_rate": 9.884373893349725e-05,
      "loss": 1.0273,
      "step": 234
    },
    {
      "epoch": 2.4040920716112533,
      "grad_norm": 0.6010921886045744,
      "learning_rate": 9.828303484133515e-05,
      "loss": 1.053,
      "step": 235
    },
    {
      "epoch": 2.414322250639386,
      "grad_norm": 0.4918851450838751,
      "learning_rate": 9.772138151541522e-05,
      "loss": 1.0364,
      "step": 236
    },
    {
      "epoch": 2.424552429667519,
      "grad_norm": 0.3407705594769831,
      "learning_rate": 9.715880811612044e-05,
      "loss": 1.0331,
      "step": 237
    },
    {
      "epoch": 2.4347826086956523,
      "grad_norm": 0.2468632428242675,
      "learning_rate": 9.659534385160289e-05,
      "loss": 1.0323,
      "step": 238
    },
    {
      "epoch": 2.445012787723785,
      "grad_norm": 0.26175126594915626,
      "learning_rate": 9.603101797626729e-05,
      "loss": 1.0491,
      "step": 239
    },
    {
      "epoch": 2.455242966751918,
      "grad_norm": 0.3250566758468044,
      "learning_rate": 9.546585978925221e-05,
      "loss": 1.0127,
      "step": 240
    },
    {
      "epoch": 2.4654731457800514,
      "grad_norm": 0.3733853550848411,
      "learning_rate": 9.489989863290885e-05,
      "loss": 1.0637,
      "step": 241
    },
    {
      "epoch": 2.475703324808184,
      "grad_norm": 0.45748627752418863,
      "learning_rate": 9.433316389127768e-05,
      "loss": 1.038,
      "step": 242
    },
    {
      "epoch": 2.4859335038363173,
      "grad_norm": 0.432153145698656,
      "learning_rate": 9.37656849885628e-05,
      "loss": 1.0441,
      "step": 243
    },
    {
      "epoch": 2.49616368286445,
      "grad_norm": 0.3361885905419595,
      "learning_rate": 9.319749138760424e-05,
      "loss": 1.0317,
      "step": 244
    },
    {
      "epoch": 2.506393861892583,
      "grad_norm": 0.26392084893370804,
      "learning_rate": 9.262861258834833e-05,
      "loss": 1.0353,
      "step": 245
    },
    {
      "epoch": 2.516624040920716,
      "grad_norm": 0.25278206548366783,
      "learning_rate": 9.205907812631616e-05,
      "loss": 1.0211,
      "step": 246
    },
    {
      "epoch": 2.526854219948849,
      "grad_norm": 0.2853155639467533,
      "learning_rate": 9.148891757106999e-05,
      "loss": 1.0381,
      "step": 247
    },
    {
      "epoch": 2.5370843989769822,
      "grad_norm": 0.28133618292890095,
      "learning_rate": 9.091816052467817e-05,
      "loss": 1.045,
      "step": 248
    },
    {
      "epoch": 2.547314578005115,
      "grad_norm": 0.26415138394214893,
      "learning_rate": 9.034683662017812e-05,
      "loss": 1.0339,
      "step": 249
    },
    {
      "epoch": 2.557544757033248,
      "grad_norm": 0.23100116731609785,
      "learning_rate": 8.977497552003785e-05,
      "loss": 1.0297,
      "step": 250
    },
    {
      "epoch": 2.5677749360613813,
      "grad_norm": 0.25137211273391913,
      "learning_rate": 8.920260691461602e-05,
      "loss": 1.0474,
      "step": 251
    },
    {
      "epoch": 2.578005115089514,
      "grad_norm": 0.25925530222353527,
      "learning_rate": 8.862976052062034e-05,
      "loss": 1.0478,
      "step": 252
    },
    {
      "epoch": 2.588235294117647,
      "grad_norm": 0.21181713461857493,
      "learning_rate": 8.805646607956467e-05,
      "loss": 1.0384,
      "step": 253
    },
    {
      "epoch": 2.59846547314578,
      "grad_norm": 0.20383781215036986,
      "learning_rate": 8.748275335622506e-05,
      "loss": 1.0352,
      "step": 254
    },
    {
      "epoch": 2.608695652173913,
      "grad_norm": 0.22216426938201625,
      "learning_rate": 8.69086521370942e-05,
      "loss": 1.0251,
      "step": 255
    },
    {
      "epoch": 2.618925831202046,
      "grad_norm": 0.2339550911616179,
      "learning_rate": 8.633419222883508e-05,
      "loss": 1.0388,
      "step": 256
    },
    {
      "epoch": 2.629156010230179,
      "grad_norm": 0.29608330426201757,
      "learning_rate": 8.575940345673337e-05,
      "loss": 1.0415,
      "step": 257
    },
    {
      "epoch": 2.639386189258312,
      "grad_norm": 0.282754017428277,
      "learning_rate": 8.518431566314901e-05,
      "loss": 1.0338,
      "step": 258
    },
    {
      "epoch": 2.649616368286445,
      "grad_norm": 0.21590808755680316,
      "learning_rate": 8.460895870596675e-05,
      "loss": 1.0455,
      "step": 259
    },
    {
      "epoch": 2.659846547314578,
      "grad_norm": 0.22038275314532527,
      "learning_rate": 8.4033362457046e-05,
      "loss": 1.0446,
      "step": 260
    },
    {
      "epoch": 2.670076726342711,
      "grad_norm": 0.19903936327716265,
      "learning_rate": 8.345755680066993e-05,
      "loss": 1.0282,
      "step": 261
    },
    {
      "epoch": 2.680306905370844,
      "grad_norm": 0.17558239320808622,
      "learning_rate": 8.288157163199389e-05,
      "loss": 1.0278,
      "step": 262
    },
    {
      "epoch": 2.690537084398977,
      "grad_norm": 0.20248735135033116,
      "learning_rate": 8.230543685549333e-05,
      "loss": 1.0317,
      "step": 263
    },
    {
      "epoch": 2.70076726342711,
      "grad_norm": 0.23742588094542533,
      "learning_rate": 8.17291823834111e-05,
      "loss": 1.0326,
      "step": 264
    },
    {
      "epoch": 2.710997442455243,
      "grad_norm": 0.3032160882786812,
      "learning_rate": 8.115283813420459e-05,
      "loss": 1.0375,
      "step": 265
    },
    {
      "epoch": 2.7212276214833757,
      "grad_norm": 0.2962448679587415,
      "learning_rate": 8.057643403099221e-05,
      "loss": 1.0584,
      "step": 266
    },
    {
      "epoch": 2.731457800511509,
      "grad_norm": 0.30722625928358643,
      "learning_rate": 8e-05,
      "loss": 1.0395,
      "step": 267
    },
    {
      "epoch": 2.741687979539642,
      "grad_norm": 0.28307960709841534,
      "learning_rate": 7.94235659690078e-05,
      "loss": 1.0369,
      "step": 268
    },
    {
      "epoch": 2.7519181585677748,
      "grad_norm": 0.21321155330317557,
      "learning_rate": 7.884716186579545e-05,
      "loss": 1.0532,
      "step": 269
    },
    {
      "epoch": 2.762148337595908,
      "grad_norm": 0.2443283053996415,
      "learning_rate": 7.827081761658892e-05,
      "loss": 1.0266,
      "step": 270
    },
    {
      "epoch": 2.772378516624041,
      "grad_norm": 0.2904460725842854,
      "learning_rate": 7.76945631445067e-05,
      "loss": 1.0344,
      "step": 271
    },
    {
      "epoch": 2.782608695652174,
      "grad_norm": 0.25431043958852784,
      "learning_rate": 7.711842836800614e-05,
      "loss": 1.0285,
      "step": 272
    },
    {
      "epoch": 2.792838874680307,
      "grad_norm": 0.25586265632907745,
      "learning_rate": 7.654244319933009e-05,
      "loss": 1.0272,
      "step": 273
    },
    {
      "epoch": 2.80306905370844,
      "grad_norm": 0.23485094953105404,
      "learning_rate": 7.596663754295404e-05,
      "loss": 1.0427,
      "step": 274
    },
    {
      "epoch": 2.813299232736573,
      "grad_norm": 0.17957023788842033,
      "learning_rate": 7.539104129403327e-05,
      "loss": 1.0474,
      "step": 275
    },
    {
      "epoch": 2.8235294117647056,
      "grad_norm": 0.160508230528404,
      "learning_rate": 7.4815684336851e-05,
      "loss": 1.0445,
      "step": 276
    },
    {
      "epoch": 2.833759590792839,
      "grad_norm": 0.21356478024874126,
      "learning_rate": 7.424059654326664e-05,
      "loss": 1.04,
      "step": 277
    },
    {
      "epoch": 2.843989769820972,
      "grad_norm": 0.21888445256566497,
      "learning_rate": 7.366580777116495e-05,
      "loss": 1.0406,
      "step": 278
    },
    {
      "epoch": 2.8542199488491047,
      "grad_norm": 0.2041365287775187,
      "learning_rate": 7.309134786290583e-05,
      "loss": 1.0321,
      "step": 279
    },
    {
      "epoch": 2.864450127877238,
      "grad_norm": 0.19151321727068246,
      "learning_rate": 7.251724664377497e-05,
      "loss": 1.0371,
      "step": 280
    },
    {
      "epoch": 2.874680306905371,
      "grad_norm": 0.18344693741146628,
      "learning_rate": 7.194353392043534e-05,
      "loss": 1.039,
      "step": 281
    },
    {
      "epoch": 2.8849104859335037,
      "grad_norm": 0.19330950475902522,
      "learning_rate": 7.13702394793797e-05,
      "loss": 1.0364,
      "step": 282
    },
    {
      "epoch": 2.895140664961637,
      "grad_norm": 0.1990648393315987,
      "learning_rate": 7.079739308538399e-05,
      "loss": 1.0277,
      "step": 283
    },
    {
      "epoch": 2.90537084398977,
      "grad_norm": 0.2262885121158685,
      "learning_rate": 7.022502447996215e-05,
      "loss": 1.0275,
      "step": 284
    },
    {
      "epoch": 2.915601023017903,
      "grad_norm": 0.20233291351840074,
      "learning_rate": 6.965316337982191e-05,
      "loss": 1.0381,
      "step": 285
    },
    {
      "epoch": 2.9258312020460355,
      "grad_norm": 0.19513363820566396,
      "learning_rate": 6.908183947532184e-05,
      "loss": 1.0342,
      "step": 286
    },
    {
      "epoch": 2.9360613810741687,
      "grad_norm": 0.20380750046128057,
      "learning_rate": 6.851108242893002e-05,
      "loss": 1.0377,
      "step": 287
    },
    {
      "epoch": 2.946291560102302,
      "grad_norm": 0.1866318852041668,
      "learning_rate": 6.794092187368387e-05,
      "loss": 1.0428,
      "step": 288
    },
    {
      "epoch": 2.9565217391304346,
      "grad_norm": 0.15744510473495013,
      "learning_rate": 6.737138741165168e-05,
      "loss": 1.0503,
      "step": 289
    },
    {
      "epoch": 2.9667519181585678,
      "grad_norm": 0.1652822650571465,
      "learning_rate": 6.680250861239581e-05,
      "loss": 1.035,
      "step": 290
    },
    {
      "epoch": 2.976982097186701,
      "grad_norm": 0.15546583646318948,
      "learning_rate": 6.623431501143723e-05,
      "loss": 1.0313,
      "step": 291
    },
    {
      "epoch": 2.9872122762148337,
      "grad_norm": 0.29006532173297445,
      "learning_rate": 6.566683610872231e-05,
      "loss": 1.0564,
      "step": 292
    },
    {
      "epoch": 2.997442455242967,
      "grad_norm": 0.2268887255063016,
      "learning_rate": 6.510010136709118e-05,
      "loss": 1.2037,
      "step": 293
    },
    {
      "epoch": 3.0076726342710995,
      "grad_norm": 0.2539730951218038,
      "learning_rate": 6.453414021074781e-05,
      "loss": 1.1394,
      "step": 294
    },
    {
      "epoch": 3.0179028132992327,
      "grad_norm": 0.2231294668730024,
      "learning_rate": 6.396898202373277e-05,
      "loss": 1.0223,
      "step": 295
    },
    {
      "epoch": 3.028132992327366,
      "grad_norm": 0.20401904180469313,
      "learning_rate": 6.340465614839714e-05,
      "loss": 1.0336,
      "step": 296
    },
    {
      "epoch": 3.0383631713554986,
      "grad_norm": 0.18979057876237626,
      "learning_rate": 6.284119188387957e-05,
      "loss": 1.0107,
      "step": 297
    },
    {
      "epoch": 3.0485933503836318,
      "grad_norm": 0.16567071328991922,
      "learning_rate": 6.227861848458481e-05,
      "loss": 1.0134,
      "step": 298
    },
    {
      "epoch": 3.0588235294117645,
      "grad_norm": 0.2788005001606031,
      "learning_rate": 6.171696515866488e-05,
      "loss": 1.0289,
      "step": 299
    },
    {
      "epoch": 3.0690537084398977,
      "grad_norm": 0.13520429300745568,
      "learning_rate": 6.115626106650273e-05,
      "loss": 1.0297,
      "step": 300
    },
    {
      "epoch": 3.079283887468031,
      "grad_norm": 0.21854013343576806,
      "learning_rate": 6.059653531919823e-05,
      "loss": 1.0282,
      "step": 301
    },
    {
      "epoch": 3.0895140664961636,
      "grad_norm": 0.1818832819994029,
      "learning_rate": 6.0037816977056625e-05,
      "loss": 1.0531,
      "step": 302
    },
    {
      "epoch": 3.0997442455242967,
      "grad_norm": 0.18589522528315505,
      "learning_rate": 5.9480135048079964e-05,
      "loss": 1.0113,
      "step": 303
    },
    {
      "epoch": 3.10997442455243,
      "grad_norm": 0.18063564176637392,
      "learning_rate": 5.892351848646087e-05,
      "loss": 1.0394,
      "step": 304
    },
    {
      "epoch": 3.1202046035805626,
      "grad_norm": 0.147586744880159,
      "learning_rate": 5.836799619107937e-05,
      "loss": 1.0365,
      "step": 305
    },
    {
      "epoch": 3.130434782608696,
      "grad_norm": 0.17675912154183307,
      "learning_rate": 5.781359700400254e-05,
      "loss": 1.0039,
      "step": 306
    },
    {
      "epoch": 3.1406649616368285,
      "grad_norm": 0.16586699917434417,
      "learning_rate": 5.726034970898682e-05,
      "loss": 1.0243,
      "step": 307
    },
    {
      "epoch": 3.1508951406649617,
      "grad_norm": 0.17820259338383218,
      "learning_rate": 5.670828302998393e-05,
      "loss": 1.0314,
      "step": 308
    },
    {
      "epoch": 3.1611253196930944,
      "grad_norm": 0.15458082339803622,
      "learning_rate": 5.6157425629649314e-05,
      "loss": 1.0485,
      "step": 309
    },
    {
      "epoch": 3.1713554987212276,
      "grad_norm": 0.1643571380269719,
      "learning_rate": 5.560780610785406e-05,
      "loss": 1.018,
      "step": 310
    },
    {
      "epoch": 3.1815856777493607,
      "grad_norm": 0.156856618068632,
      "learning_rate": 5.5059453000200125e-05,
      "loss": 1.0061,
      "step": 311
    },
    {
      "epoch": 3.1918158567774935,
      "grad_norm": 0.14086641088136265,
      "learning_rate": 5.451239477653864e-05,
      "loss": 1.0205,
      "step": 312
    },
    {
      "epoch": 3.2020460358056266,
      "grad_norm": 0.14820087407296167,
      "learning_rate": 5.3966659839491936e-05,
      "loss": 1.0226,
      "step": 313
    },
    {
      "epoch": 3.21227621483376,
      "grad_norm": 0.13884477979100748,
      "learning_rate": 5.342227652297887e-05,
      "loss": 1.001,
      "step": 314
    },
    {
      "epoch": 3.2225063938618925,
      "grad_norm": 0.14309328916359365,
      "learning_rate": 5.287927309074365e-05,
      "loss": 1.031,
      "step": 315
    },
    {
      "epoch": 3.2327365728900257,
      "grad_norm": 0.15252565103321644,
      "learning_rate": 5.233767773488859e-05,
      "loss": 1.015,
      "step": 316
    },
    {
      "epoch": 3.2429667519181584,
      "grad_norm": 0.13230625482589622,
      "learning_rate": 5.179751857441036e-05,
      "loss": 1.0053,
      "step": 317
    },
    {
      "epoch": 3.2531969309462916,
      "grad_norm": 0.1556826070734683,
      "learning_rate": 5.1258823653739914e-05,
      "loss": 1.0211,
      "step": 318
    },
    {
      "epoch": 3.2634271099744243,
      "grad_norm": 0.1371375107565863,
      "learning_rate": 5.0721620941286735e-05,
      "loss": 1.0143,
      "step": 319
    },
    {
      "epoch": 3.2736572890025575,
      "grad_norm": 0.13573969073646522,
      "learning_rate": 5.018593832798649e-05,
      "loss": 1.0375,
      "step": 320
    },
    {
      "epoch": 3.2838874680306906,
      "grad_norm": 0.11657280744271552,
      "learning_rate": 4.965180362585315e-05,
      "loss": 1.0253,
      "step": 321
    },
    {
      "epoch": 3.2941176470588234,
      "grad_norm": 0.11616693511840327,
      "learning_rate": 4.911924456653494e-05,
      "loss": 1.0209,
      "step": 322
    },
    {
      "epoch": 3.3043478260869565,
      "grad_norm": 0.12512977851281545,
      "learning_rate": 4.8588288799874514e-05,
      "loss": 1.0112,
      "step": 323
    },
    {
      "epoch": 3.3145780051150897,
      "grad_norm": 0.10387393823484337,
      "learning_rate": 4.805896389247348e-05,
      "loss": 1.0077,
      "step": 324
    },
    {
      "epoch": 3.3248081841432224,
      "grad_norm": 0.13924261830307932,
      "learning_rate": 4.753129732626116e-05,
      "loss": 1.0229,
      "step": 325
    },
    {
      "epoch": 3.3350383631713556,
      "grad_norm": 0.12135011031399273,
      "learning_rate": 4.70053164970677e-05,
      "loss": 1.0184,
      "step": 326
    },
    {
      "epoch": 3.3452685421994883,
      "grad_norm": 0.1183989328140461,
      "learning_rate": 4.6481048713201825e-05,
      "loss": 1.0058,
      "step": 327
    },
    {
      "epoch": 3.3554987212276215,
      "grad_norm": 0.1284311973823529,
      "learning_rate": 4.595852119403282e-05,
      "loss": 1.0278,
      "step": 328
    },
    {
      "epoch": 3.3657289002557547,
      "grad_norm": 0.11430293939389213,
      "learning_rate": 4.543776106857765e-05,
      "loss": 1.012,
      "step": 329
    },
    {
      "epoch": 3.3759590792838874,
      "grad_norm": 0.1263289266506516,
      "learning_rate": 4.491879537409211e-05,
      "loss": 1.0242,
      "step": 330
    },
    {
      "epoch": 3.3861892583120206,
      "grad_norm": 0.1316142612111793,
      "learning_rate": 4.4401651054667274e-05,
      "loss": 1.0078,
      "step": 331
    },
    {
      "epoch": 3.3964194373401533,
      "grad_norm": 0.10490352230994067,
      "learning_rate": 4.3886354959830625e-05,
      "loss": 1.0141,
      "step": 332
    },
    {
      "epoch": 3.4066496163682864,
      "grad_norm": 0.1587962681364766,
      "learning_rate": 4.3372933843152e-05,
      "loss": 1.031,
      "step": 333
    },
    {
      "epoch": 3.4168797953964196,
      "grad_norm": 0.14806873434228535,
      "learning_rate": 4.2861414360854387e-05,
      "loss": 1.0261,
      "step": 334
    },
    {
      "epoch": 3.4271099744245523,
      "grad_norm": 0.12428110117787702,
      "learning_rate": 4.2351823070430376e-05,
      "loss": 1.0191,
      "step": 335
    },
    {
      "epoch": 3.4373401534526855,
      "grad_norm": 0.1700524980071232,
      "learning_rate": 4.184418642926289e-05,
      "loss": 1.0267,
      "step": 336
    },
    {
      "epoch": 3.4475703324808182,
      "grad_norm": 0.15240023162277883,
      "learning_rate": 4.133853079325196e-05,
      "loss": 1.025,
      "step": 337
    },
    {
      "epoch": 3.4578005115089514,
      "grad_norm": 0.11788415451483955,
      "learning_rate": 4.083488241544595e-05,
      "loss": 1.0459,
      "step": 338
    },
    {
      "epoch": 3.4680306905370846,
      "grad_norm": 0.16921030624641467,
      "learning_rate": 4.033326744467882e-05,
      "loss": 1.0112,
      "step": 339
    },
    {
      "epoch": 3.4782608695652173,
      "grad_norm": 0.11606787523018822,
      "learning_rate": 3.983371192421246e-05,
      "loss": 1.0306,
      "step": 340
    },
    {
      "epoch": 3.4884910485933505,
      "grad_norm": 0.1682374876165679,
      "learning_rate": 3.933624179038446e-05,
      "loss": 1.0185,
      "step": 341
    },
    {
      "epoch": 3.498721227621483,
      "grad_norm": 0.1565962574632331,
      "learning_rate": 3.884088287126151e-05,
      "loss": 1.0293,
      "step": 342
    },
    {
      "epoch": 3.5089514066496164,
      "grad_norm": 0.14324776346066015,
      "learning_rate": 3.834766088529867e-05,
      "loss": 1.0458,
      "step": 343
    },
    {
      "epoch": 3.5191815856777495,
      "grad_norm": 0.19529159534261803,
      "learning_rate": 3.785660144000378e-05,
      "loss": 1.0056,
      "step": 344
    },
    {
      "epoch": 3.5294117647058822,
      "grad_norm": 0.12801441909483788,
      "learning_rate": 3.736773003060821e-05,
      "loss": 1.0297,
      "step": 345
    },
    {
      "epoch": 3.5396419437340154,
      "grad_norm": 0.1503941568635443,
      "learning_rate": 3.688107203874301e-05,
      "loss": 1.0416,
      "step": 346
    },
    {
      "epoch": 3.5498721227621486,
      "grad_norm": 0.13361192505091346,
      "learning_rate": 3.6396652731121136e-05,
      "loss": 1.0204,
      "step": 347
    },
    {
      "epoch": 3.5601023017902813,
      "grad_norm": 0.12402122906765294,
      "learning_rate": 3.5914497258225815e-05,
      "loss": 1.0281,
      "step": 348
    },
    {
      "epoch": 3.5703324808184145,
      "grad_norm": 0.1370545977754261,
      "learning_rate": 3.543463065300452e-05,
      "loss": 1.0271,
      "step": 349
    },
    {
      "epoch": 3.580562659846547,
      "grad_norm": 0.10522140449656789,
      "learning_rate": 3.49570778295694e-05,
      "loss": 1.0253,
      "step": 350
    },
    {
      "epoch": 3.5907928388746804,
      "grad_norm": 0.13288330038665777,
      "learning_rate": 3.448186358190383e-05,
      "loss": 1.0155,
      "step": 351
    },
    {
      "epoch": 3.601023017902813,
      "grad_norm": 0.11850947440760301,
      "learning_rate": 3.400901258257501e-05,
      "loss": 1.0316,
      "step": 352
    },
    {
      "epoch": 3.6112531969309463,
      "grad_norm": 0.1207639690615278,
      "learning_rate": 3.3538549381453046e-05,
      "loss": 1.0147,
      "step": 353
    },
    {
      "epoch": 3.6214833759590794,
      "grad_norm": 0.10436214175934275,
      "learning_rate": 3.307049840443644e-05,
      "loss": 1.0158,
      "step": 354
    },
    {
      "epoch": 3.631713554987212,
      "grad_norm": 0.10939591730050287,
      "learning_rate": 3.2604883952183716e-05,
      "loss": 1.0219,
      "step": 355
    },
    {
      "epoch": 3.6419437340153453,
      "grad_norm": 0.10305381344145671,
      "learning_rate": 3.214173019885202e-05,
      "loss": 1.0165,
      "step": 356
    },
    {
      "epoch": 3.6521739130434785,
      "grad_norm": 0.10617153493263774,
      "learning_rate": 3.1681061190841806e-05,
      "loss": 1.0193,
      "step": 357
    },
    {
      "epoch": 3.662404092071611,
      "grad_norm": 0.1052450380146473,
      "learning_rate": 3.122290084554845e-05,
      "loss": 1.0309,
      "step": 358
    },
    {
      "epoch": 3.6726342710997444,
      "grad_norm": 0.09779940640870793,
      "learning_rate": 3.076727295012059e-05,
      "loss": 1.0106,
      "step": 359
    },
    {
      "epoch": 3.682864450127877,
      "grad_norm": 0.09865111890866658,
      "learning_rate": 3.031420116022493e-05,
      "loss": 1.0237,
      "step": 360
    },
    {
      "epoch": 3.6930946291560103,
      "grad_norm": 0.09761460812236314,
      "learning_rate": 2.98637089988182e-05,
      "loss": 1.0071,
      "step": 361
    },
    {
      "epoch": 3.703324808184143,
      "grad_norm": 0.09657195902847425,
      "learning_rate": 2.94158198549259e-05,
      "loss": 1.0213,
      "step": 362
    },
    {
      "epoch": 3.713554987212276,
      "grad_norm": 0.10306766572443878,
      "learning_rate": 2.8970556982427836e-05,
      "loss": 1.0114,
      "step": 363
    },
    {
      "epoch": 3.7237851662404093,
      "grad_norm": 0.0892256925639781,
      "learning_rate": 2.852794349885087e-05,
      "loss": 1.0141,
      "step": 364
    },
    {
      "epoch": 3.734015345268542,
      "grad_norm": 0.0984967903902541,
      "learning_rate": 2.8088002384168783e-05,
      "loss": 1.0309,
      "step": 365
    },
    {
      "epoch": 3.7442455242966752,
      "grad_norm": 0.09702833968633048,
      "learning_rate": 2.765075647960898e-05,
      "loss": 1.0133,
      "step": 366
    },
    {
      "epoch": 3.7544757033248084,
      "grad_norm": 0.10259443786625837,
      "learning_rate": 2.7216228486466856e-05,
      "loss": 1.0158,
      "step": 367
    },
    {
      "epoch": 3.764705882352941,
      "grad_norm": 0.09444360606393558,
      "learning_rate": 2.678444096492683e-05,
      "loss": 1.02,
      "step": 368
    },
    {
      "epoch": 3.7749360613810743,
      "grad_norm": 0.11306816830082422,
      "learning_rate": 2.6355416332891404e-05,
      "loss": 1.0185,
      "step": 369
    },
    {
      "epoch": 3.785166240409207,
      "grad_norm": 0.09142683583600977,
      "learning_rate": 2.592917686481708e-05,
      "loss": 1.0038,
      "step": 370
    },
    {
      "epoch": 3.79539641943734,
      "grad_norm": 0.10949950748671738,
      "learning_rate": 2.5505744690557846e-05,
      "loss": 1.0376,
      "step": 371
    },
    {
      "epoch": 3.805626598465473,
      "grad_norm": 0.11343467361764166,
      "learning_rate": 2.508514179421629e-05,
      "loss": 1.0358,
      "step": 372
    },
    {
      "epoch": 3.815856777493606,
      "grad_norm": 0.09342791259699781,
      "learning_rate": 2.4667390013002254e-05,
      "loss": 1.0211,
      "step": 373
    },
    {
      "epoch": 3.8260869565217392,
      "grad_norm": 0.10858137240897216,
      "learning_rate": 2.425251103609898e-05,
      "loss": 1.0332,
      "step": 374
    },
    {
      "epoch": 3.836317135549872,
      "grad_norm": 0.0886186909107238,
      "learning_rate": 2.3840526403537095e-05,
      "loss": 1.0143,
      "step": 375
    },
    {
      "epoch": 3.846547314578005,
      "grad_norm": 0.09895029034827527,
      "learning_rate": 2.3431457505076205e-05,
      "loss": 1.0173,
      "step": 376
    },
    {
      "epoch": 3.8567774936061383,
      "grad_norm": 0.10546368891288044,
      "learning_rate": 2.3025325579094498e-05,
      "loss": 1.0362,
      "step": 377
    },
    {
      "epoch": 3.867007672634271,
      "grad_norm": 0.08194245174545557,
      "learning_rate": 2.2622151711485962e-05,
      "loss": 1.0124,
      "step": 378
    },
    {
      "epoch": 3.877237851662404,
      "grad_norm": 0.1114016593112589,
      "learning_rate": 2.2221956834565647e-05,
      "loss": 1.0139,
      "step": 379
    },
    {
      "epoch": 3.887468030690537,
      "grad_norm": 0.09816527002580688,
      "learning_rate": 2.1824761725982874e-05,
      "loss": 1.0523,
      "step": 380
    },
    {
      "epoch": 3.89769820971867,
      "grad_norm": 0.0799950179310954,
      "learning_rate": 2.1430587007642513e-05,
      "loss": 0.991,
      "step": 381
    },
    {
      "epoch": 3.907928388746803,
      "grad_norm": 0.1080492642683735,
      "learning_rate": 2.1039453144634364e-05,
      "loss": 1.0355,
      "step": 382
    },
    {
      "epoch": 3.918158567774936,
      "grad_norm": 0.09178099069537385,
      "learning_rate": 2.0651380444170527e-05,
      "loss": 1.015,
      "step": 383
    },
    {
      "epoch": 3.928388746803069,
      "grad_norm": 0.09031457518884235,
      "learning_rate": 2.026638905453111e-05,
      "loss": 1.0229,
      "step": 384
    },
    {
      "epoch": 3.938618925831202,
      "grad_norm": 0.09836817148107584,
      "learning_rate": 1.9884498964018233e-05,
      "loss": 1.0135,
      "step": 385
    },
    {
      "epoch": 3.948849104859335,
      "grad_norm": 0.08002277345563297,
      "learning_rate": 1.9505729999918194e-05,
      "loss": 1.0044,
      "step": 386
    },
    {
      "epoch": 3.959079283887468,
      "grad_norm": 0.08738866259057297,
      "learning_rate": 1.913010182747196e-05,
      "loss": 1.0103,
      "step": 387
    },
    {
      "epoch": 3.969309462915601,
      "grad_norm": 0.08459856390870199,
      "learning_rate": 1.875763394885441e-05,
      "loss": 1.0491,
      "step": 388
    },
    {
      "epoch": 3.979539641943734,
      "grad_norm": 0.07894640689767103,
      "learning_rate": 1.8388345702161556e-05,
      "loss": 1.0355,
      "step": 389
    },
    {
      "epoch": 3.9897698209718673,
      "grad_norm": 0.07643936790355793,
      "learning_rate": 1.8022256260406756e-05,
      "loss": 1.021,
      "step": 390
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.09661289153754662,
      "learning_rate": 1.765938463052506e-05,
      "loss": 1.288,
      "step": 391
    },
    {
      "epoch": 4.010230179028133,
      "grad_norm": 0.08762181265193057,
      "learning_rate": 1.729974965238651e-05,
      "loss": 1.001,
      "step": 392
    },
    {
      "epoch": 4.020460358056266,
      "grad_norm": 0.08647609139764888,
      "learning_rate": 1.6943369997818066e-05,
      "loss": 1.015,
      "step": 393
    },
    {
      "epoch": 4.030690537084399,
      "grad_norm": 0.08788066772345464,
      "learning_rate": 1.659026416963401e-05,
      "loss": 1.0076,
      "step": 394
    },
    {
      "epoch": 4.040920716112532,
      "grad_norm": 0.07750381562018518,
      "learning_rate": 1.6240450500675393e-05,
      "loss": 1.0148,
      "step": 395
    },
    {
      "epoch": 4.051150895140665,
      "grad_norm": 0.07635694926329481,
      "learning_rate": 1.5893947152858285e-05,
      "loss": 1.0016,
      "step": 396
    },
    {
      "epoch": 4.061381074168798,
      "grad_norm": 0.08148076505063037,
      "learning_rate": 1.55507721162307e-05,
      "loss": 1.0043,
      "step": 397
    },
    {
      "epoch": 4.071611253196931,
      "grad_norm": 0.07751240865988411,
      "learning_rate": 1.5210943208038634e-05,
      "loss": 1.0288,
      "step": 398
    },
    {
      "epoch": 4.081841432225064,
      "grad_norm": 0.07904341557445363,
      "learning_rate": 1.4874478071801055e-05,
      "loss": 1.0302,
      "step": 399
    },
    {
      "epoch": 4.092071611253197,
      "grad_norm": 0.08036549522985713,
      "learning_rate": 1.454139417639377e-05,
      "loss": 1.0021,
      "step": 400
    },
    {
      "epoch": 4.10230179028133,
      "grad_norm": 0.07597724481740194,
      "learning_rate": 1.4211708815142599e-05,
      "loss": 1.008,
      "step": 401
    },
    {
      "epoch": 4.112531969309463,
      "grad_norm": 0.07493385254201147,
      "learning_rate": 1.3885439104925387e-05,
      "loss": 1.0082,
      "step": 402
    },
    {
      "epoch": 4.122762148337596,
      "grad_norm": 0.08492760833944613,
      "learning_rate": 1.3562601985283358e-05,
      "loss": 1.0103,
      "step": 403
    },
    {
      "epoch": 4.132992327365729,
      "grad_norm": 0.0819518128021678,
      "learning_rate": 1.3243214217541751e-05,
      "loss": 1.0186,
      "step": 404
    },
    {
      "epoch": 4.143222506393862,
      "grad_norm": 0.07813515029694298,
      "learning_rate": 1.2927292383939407e-05,
      "loss": 1.0103,
      "step": 405
    },
    {
      "epoch": 4.153452685421995,
      "grad_norm": 0.07571122126240162,
      "learning_rate": 1.2614852886767932e-05,
      "loss": 1.0172,
      "step": 406
    },
    {
      "epoch": 4.163682864450128,
      "grad_norm": 0.07867401677851067,
      "learning_rate": 1.2305911947520159e-05,
      "loss": 1.0172,
      "step": 407
    },
    {
      "epoch": 4.173913043478261,
      "grad_norm": 0.07427552294515202,
      "learning_rate": 1.2000485606047837e-05,
      "loss": 1.0254,
      "step": 408
    },
    {
      "epoch": 4.1841432225063935,
      "grad_norm": 0.07918066825847282,
      "learning_rate": 1.1698589719728911e-05,
      "loss": 1.025,
      "step": 409
    },
    {
      "epoch": 4.194373401534527,
      "grad_norm": 0.0745101976937812,
      "learning_rate": 1.1400239962644294e-05,
      "loss": 1.011,
      "step": 410
    },
    {
      "epoch": 4.20460358056266,
      "grad_norm": 0.07635541094438233,
      "learning_rate": 1.1105451824763933e-05,
      "loss": 1.0064,
      "step": 411
    },
    {
      "epoch": 4.2148337595907925,
      "grad_norm": 0.07409705314572482,
      "learning_rate": 1.0814240611142765e-05,
      "loss": 1.0049,
      "step": 412
    },
    {
      "epoch": 4.225063938618926,
      "grad_norm": 0.07875185704588439,
      "learning_rate": 1.0526621441125946e-05,
      "loss": 1.0039,
      "step": 413
    },
    {
      "epoch": 4.235294117647059,
      "grad_norm": 0.07576883769832815,
      "learning_rate": 1.0242609247563924e-05,
      "loss": 1.0204,
      "step": 414
    },
    {
      "epoch": 4.245524296675192,
      "grad_norm": 0.07808157523983991,
      "learning_rate": 9.962218776037234e-06,
      "loss": 1.0178,
      "step": 415
    },
    {
      "epoch": 4.255754475703325,
      "grad_norm": 0.0706208132598418,
      "learning_rate": 9.68546458409077e-06,
      "loss": 1.0144,
      "step": 416
    },
    {
      "epoch": 4.265984654731458,
      "grad_norm": 0.0816397013567019,
      "learning_rate": 9.41236104047806e-06,
      "loss": 1.0008,
      "step": 417
    },
    {
      "epoch": 4.276214833759591,
      "grad_norm": 0.07454955996130302,
      "learning_rate": 9.14292232441528e-06,
      "loss": 1.0102,
      "step": 418
    },
    {
      "epoch": 4.286445012787723,
      "grad_norm": 0.0715010447940489,
      "learning_rate": 8.877162424845012e-06,
      "loss": 0.997,
      "step": 419
    },
    {
      "epoch": 4.296675191815857,
      "grad_norm": 0.07752947453647589,
      "learning_rate": 8.615095139710044e-06,
      "loss": 1.0204,
      "step": 420
    },
    {
      "epoch": 4.30690537084399,
      "grad_norm": 0.07999927748669688,
      "learning_rate": 8.356734075236858e-06,
      "loss": 1.0286,
      "step": 421
    },
    {
      "epoch": 4.3171355498721224,
      "grad_norm": 0.07470822337555567,
      "learning_rate": 8.102092645229392e-06,
      "loss": 0.9999,
      "step": 422
    },
    {
      "epoch": 4.327365728900256,
      "grad_norm": 0.06986020359409462,
      "learning_rate": 7.8511840703725e-06,
      "loss": 1.0157,
      "step": 423
    },
    {
      "epoch": 4.337595907928389,
      "grad_norm": 0.07546615544304432,
      "learning_rate": 7.604021377545518e-06,
      "loss": 1.0177,
      "step": 424
    },
    {
      "epoch": 4.3478260869565215,
      "grad_norm": 0.07303012390571562,
      "learning_rate": 7.36061739914601e-06,
      "loss": 1.0053,
      "step": 425
    },
    {
      "epoch": 4.358056265984655,
      "grad_norm": 0.07983971604823199,
      "learning_rate": 7.120984772423507e-06,
      "loss": 1.0116,
      "step": 426
    },
    {
      "epoch": 4.368286445012788,
      "grad_norm": 0.07412392301940252,
      "learning_rate": 6.88513593882334e-06,
      "loss": 1.0157,
      "step": 427
    },
    {
      "epoch": 4.378516624040921,
      "grad_norm": 0.07540589914642663,
      "learning_rate": 6.653083143340748e-06,
      "loss": 1.0321,
      "step": 428
    },
    {
      "epoch": 4.388746803069053,
      "grad_norm": 0.0734695901154156,
      "learning_rate": 6.4248384338851146e-06,
      "loss": 1.0166,
      "step": 429
    },
    {
      "epoch": 4.398976982097187,
      "grad_norm": 0.07148917346240428,
      "learning_rate": 6.2004136606544515e-06,
      "loss": 1.0155,
      "step": 430
    },
    {
      "epoch": 4.40920716112532,
      "grad_norm": 0.06576015271509353,
      "learning_rate": 5.979820475520202e-06,
      "loss": 1.0268,
      "step": 431
    },
    {
      "epoch": 4.419437340153452,
      "grad_norm": 0.06602815168374586,
      "learning_rate": 5.763070331422151e-06,
      "loss": 1.0094,
      "step": 432
    },
    {
      "epoch": 4.429667519181586,
      "grad_norm": 0.06806818912226788,
      "learning_rate": 5.550174481773969e-06,
      "loss": 1.0117,
      "step": 433
    },
    {
      "epoch": 4.439897698209719,
      "grad_norm": 0.0654860198424932,
      "learning_rate": 5.341143979878851e-06,
      "loss": 1.024,
      "step": 434
    },
    {
      "epoch": 4.450127877237851,
      "grad_norm": 0.06347915481210827,
      "learning_rate": 5.135989678355664e-06,
      "loss": 1.0068,
      "step": 435
    },
    {
      "epoch": 4.460358056265985,
      "grad_norm": 0.06673007006524383,
      "learning_rate": 4.934722228575481e-06,
      "loss": 1.0144,
      "step": 436
    },
    {
      "epoch": 4.470588235294118,
      "grad_norm": 0.06276490466658945,
      "learning_rate": 4.7373520801085705e-06,
      "loss": 1.0149,
      "step": 437
    },
    {
      "epoch": 4.4808184143222505,
      "grad_norm": 0.06462800206504453,
      "learning_rate": 4.543889480181944e-06,
      "loss": 1.0209,
      "step": 438
    },
    {
      "epoch": 4.491048593350383,
      "grad_norm": 0.06638702716213918,
      "learning_rate": 4.354344473147194e-06,
      "loss": 1.0229,
      "step": 439
    },
    {
      "epoch": 4.501278772378517,
      "grad_norm": 0.06786708229119566,
      "learning_rate": 4.1687268999591164e-06,
      "loss": 1.0093,
      "step": 440
    },
    {
      "epoch": 4.5115089514066495,
      "grad_norm": 0.06344929556504937,
      "learning_rate": 3.98704639766474e-06,
      "loss": 1.0227,
      "step": 441
    },
    {
      "epoch": 4.521739130434782,
      "grad_norm": 0.061257597771025976,
      "learning_rate": 3.809312398903e-06,
      "loss": 1.0206,
      "step": 442
    },
    {
      "epoch": 4.531969309462916,
      "grad_norm": 0.06207018671567817,
      "learning_rate": 3.6355341314149216e-06,
      "loss": 1.0061,
      "step": 443
    },
    {
      "epoch": 4.542199488491049,
      "grad_norm": 0.0697177559048247,
      "learning_rate": 3.465720617564676e-06,
      "loss": 1.0001,
      "step": 444
    },
    {
      "epoch": 4.552429667519181,
      "grad_norm": 0.06909570953584229,
      "learning_rate": 3.299880673871023e-06,
      "loss": 1.0179,
      "step": 445
    },
    {
      "epoch": 4.562659846547315,
      "grad_norm": 0.0660073639280078,
      "learning_rate": 3.138022910549632e-06,
      "loss": 1.0261,
      "step": 446
    },
    {
      "epoch": 4.572890025575448,
      "grad_norm": 0.061208022344978574,
      "learning_rate": 2.980155731066017e-06,
      "loss": 0.9983,
      "step": 447
    },
    {
      "epoch": 4.58312020460358,
      "grad_norm": 0.05925952496152496,
      "learning_rate": 2.8262873316992556e-06,
      "loss": 1.0232,
      "step": 448
    },
    {
      "epoch": 4.593350383631714,
      "grad_norm": 0.06402981619720209,
      "learning_rate": 2.676425701116463e-06,
      "loss": 1.0065,
      "step": 449
    },
    {
      "epoch": 4.603580562659847,
      "grad_norm": 0.058782758741933706,
      "learning_rate": 2.530578619957993e-06,
      "loss": 1.0117,
      "step": 450
    },
    {
      "epoch": 4.6138107416879794,
      "grad_norm": 0.05836448954212346,
      "learning_rate": 2.3887536604334784e-06,
      "loss": 0.9904,
      "step": 451
    },
    {
      "epoch": 4.624040920716112,
      "grad_norm": 0.05866978691158495,
      "learning_rate": 2.2509581859287576e-06,
      "loss": 1.018,
      "step": 452
    },
    {
      "epoch": 4.634271099744246,
      "grad_norm": 0.05897537411295091,
      "learning_rate": 2.117199350623462e-06,
      "loss": 1.0224,
      "step": 453
    },
    {
      "epoch": 4.6445012787723785,
      "grad_norm": 0.0576826219117585,
      "learning_rate": 1.987484099119712e-06,
      "loss": 1.0256,
      "step": 454
    },
    {
      "epoch": 4.654731457800511,
      "grad_norm": 0.058074308794142736,
      "learning_rate": 1.8618191660814356e-06,
      "loss": 1.0126,
      "step": 455
    },
    {
      "epoch": 4.664961636828645,
      "grad_norm": 0.057718766623449665,
      "learning_rate": 1.7402110758847834e-06,
      "loss": 1.0064,
      "step": 456
    },
    {
      "epoch": 4.675191815856778,
      "grad_norm": 0.055268606367167225,
      "learning_rate": 1.6226661422794033e-06,
      "loss": 1.0015,
      "step": 457
    },
    {
      "epoch": 4.68542199488491,
      "grad_norm": 0.06120460420865381,
      "learning_rate": 1.5091904680605862e-06,
      "loss": 1.0195,
      "step": 458
    },
    {
      "epoch": 4.695652173913043,
      "grad_norm": 0.057198583474872174,
      "learning_rate": 1.3997899447524277e-06,
      "loss": 1.0313,
      "step": 459
    },
    {
      "epoch": 4.705882352941177,
      "grad_norm": 0.054125320926830395,
      "learning_rate": 1.294470252302009e-06,
      "loss": 1.0074,
      "step": 460
    },
    {
      "epoch": 4.716112531969309,
      "grad_norm": 0.05767289138418469,
      "learning_rate": 1.193236858784408e-06,
      "loss": 1.0073,
      "step": 461
    },
    {
      "epoch": 4.726342710997442,
      "grad_norm": 0.058374045513090105,
      "learning_rate": 1.0960950201188524e-06,
      "loss": 1.0217,
      "step": 462
    },
    {
      "epoch": 4.736572890025576,
      "grad_norm": 0.05554209835620826,
      "learning_rate": 1.003049779795866e-06,
      "loss": 1.0167,
      "step": 463
    },
    {
      "epoch": 4.746803069053708,
      "grad_norm": 0.06184311865928885,
      "learning_rate": 9.141059686153419e-07,
      "loss": 1.0207,
      "step": 464
    },
    {
      "epoch": 4.757033248081841,
      "grad_norm": 0.056079764830628674,
      "learning_rate": 8.292682044358114e-07,
      "loss": 1.0169,
      "step": 465
    },
    {
      "epoch": 4.767263427109975,
      "grad_norm": 0.05903827365785868,
      "learning_rate": 7.485408919346171e-07,
      "loss": 1.0276,
      "step": 466
    },
    {
      "epoch": 4.7774936061381075,
      "grad_norm": 0.05665451559441624,
      "learning_rate": 6.719282223793056e-07,
      "loss": 1.0108,
      "step": 467
    },
    {
      "epoch": 4.78772378516624,
      "grad_norm": 0.05541913708216884,
      "learning_rate": 5.994341734099429e-07,
      "loss": 1.0213,
      "step": 468
    },
    {
      "epoch": 4.797953964194374,
      "grad_norm": 0.0531727158843823,
      "learning_rate": 5.310625088326671e-07,
      "loss": 0.9962,
      "step": 469
    },
    {
      "epoch": 4.8081841432225065,
      "grad_norm": 0.056474278323680495,
      "learning_rate": 4.6681677842421724e-07,
      "loss": 1.0079,
      "step": 470
    },
    {
      "epoch": 4.818414322250639,
      "grad_norm": 0.05529573865410167,
      "learning_rate": 4.067003177476991e-07,
      "loss": 1.0025,
      "step": 471
    },
    {
      "epoch": 4.828644501278772,
      "grad_norm": 0.055929732769464856,
      "learning_rate": 3.507162479793369e-07,
      "loss": 1.0173,
      "step": 472
    },
    {
      "epoch": 4.838874680306906,
      "grad_norm": 0.0546056253906124,
      "learning_rate": 2.9886747574646936e-07,
      "loss": 1.0001,
      "step": 473
    },
    {
      "epoch": 4.849104859335038,
      "grad_norm": 0.05682769159587846,
      "learning_rate": 2.511566929766396e-07,
      "loss": 1.0062,
      "step": 474
    },
    {
      "epoch": 4.859335038363171,
      "grad_norm": 0.05365397819606409,
      "learning_rate": 2.075863767577957e-07,
      "loss": 1.0195,
      "step": 475
    },
    {
      "epoch": 4.869565217391305,
      "grad_norm": 0.05538514973032816,
      "learning_rate": 1.681587892097536e-07,
      "loss": 1.0159,
      "step": 476
    },
    {
      "epoch": 4.879795396419437,
      "grad_norm": 0.05391134086338163,
      "learning_rate": 1.3287597736667323e-07,
      "loss": 1.0233,
      "step": 477
    },
    {
      "epoch": 4.89002557544757,
      "grad_norm": 0.054055617154126184,
      "learning_rate": 1.0173977307082361e-07,
      "loss": 1.0188,
      "step": 478
    },
    {
      "epoch": 4.900255754475703,
      "grad_norm": 0.05372397484026782,
      "learning_rate": 7.475179287748547e-08,
      "loss": 1.0235,
      "step": 479
    },
    {
      "epoch": 4.910485933503836,
      "grad_norm": 0.054794275718811285,
      "learning_rate": 5.191343797096515e-08,
      "loss": 1.0018,
      "step": 480
    },
    {
      "epoch": 4.920716112531969,
      "grad_norm": 0.05350058384849903,
      "learning_rate": 3.322589409190613e-08,
      "loss": 1.009,
      "step": 481
    },
    {
      "epoch": 4.930946291560103,
      "grad_norm": 0.05480646007195536,
      "learning_rate": 1.8690131475711527e-08,
      "loss": 1.0273,
      "step": 482
    },
    {
      "epoch": 4.9411764705882355,
      "grad_norm": 0.05416557096319236,
      "learning_rate": 8.306904802148907e-09,
      "loss": 1.0322,
      "step": 483
    },
    {
      "epoch": 4.951406649616368,
      "grad_norm": 0.053492193014158126,
      "learning_rate": 2.07675315618161e-09,
      "loss": 1.0035,
      "step": 484
    },
    {
      "epoch": 4.961636828644501,
      "grad_norm": 0.055105663696246775,
      "learning_rate": 0.0,
      "loss": 1.0124,
      "step": 485
    },
    {
      "epoch": 4.961636828644501,
      "step": 485,
      "total_flos": 7065760181780480.0,
      "train_loss": 1.075610858509221,
      "train_runtime": 69151.8205,
      "train_samples_per_second": 7.23,
      "train_steps_per_second": 0.007
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 485,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 7065760181780480.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}