{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.982343499197432,
  "eval_steps": 500,
  "global_step": 385,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.012841091492776886,
      "grad_norm": 0.40566644072532654,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.2239,
      "step": 1
    },
    {
      "epoch": 0.025682182985553772,
      "grad_norm": 0.5796589851379395,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.2932,
      "step": 2
    },
    {
      "epoch": 0.038523274478330656,
      "grad_norm": 0.620160698890686,
      "learning_rate": 5e-05,
      "loss": 0.3059,
      "step": 3
    },
    {
      "epoch": 0.051364365971107544,
      "grad_norm": 0.4742761254310608,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.2751,
      "step": 4
    },
    {
      "epoch": 0.06420545746388442,
      "grad_norm": 0.39537593722343445,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.2436,
      "step": 5
    },
    {
      "epoch": 0.07704654895666131,
      "grad_norm": 0.33797481656074524,
      "learning_rate": 0.0001,
      "loss": 0.2592,
      "step": 6
    },
    {
      "epoch": 0.0898876404494382,
      "grad_norm": 0.3221365511417389,
      "learning_rate": 0.00011666666666666668,
      "loss": 0.2416,
      "step": 7
    },
    {
      "epoch": 0.10272873194221509,
      "grad_norm": 0.28052330017089844,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.2092,
      "step": 8
    },
    {
      "epoch": 0.11556982343499198,
      "grad_norm": 0.29556986689567566,
      "learning_rate": 0.00015000000000000001,
      "loss": 0.2056,
      "step": 9
    },
    {
      "epoch": 0.12841091492776885,
      "grad_norm": 0.30732303857803345,
      "learning_rate": 0.0001666666666666667,
      "loss": 0.1879,
      "step": 10
    },
    {
      "epoch": 0.14125200642054575,
      "grad_norm": 0.30247950553894043,
      "learning_rate": 0.00018333333333333334,
      "loss": 0.1839,
      "step": 11
    },
    {
      "epoch": 0.15409309791332262,
      "grad_norm": 0.3433546721935272,
      "learning_rate": 0.0002,
      "loss": 0.1908,
      "step": 12
    },
    {
      "epoch": 0.16693418940609953,
      "grad_norm": 0.19485895335674286,
      "learning_rate": 0.00019999645309530845,
      "loss": 0.1743,
      "step": 13
    },
    {
      "epoch": 0.1797752808988764,
      "grad_norm": 0.18382272124290466,
      "learning_rate": 0.00019998581263284444,
      "loss": 0.1843,
      "step": 14
    },
    {
      "epoch": 0.1926163723916533,
      "grad_norm": 0.1688140630722046,
      "learning_rate": 0.00019996807936742208,
      "loss": 0.1765,
      "step": 15
    },
    {
      "epoch": 0.20545746388443017,
      "grad_norm": 0.1830463707447052,
      "learning_rate": 0.00019994325455700542,
      "loss": 0.1922,
      "step": 16
    },
    {
      "epoch": 0.21829855537720708,
      "grad_norm": 0.22317859530448914,
      "learning_rate": 0.00019991133996261922,
      "loss": 0.2321,
      "step": 17
    },
    {
      "epoch": 0.23113964686998395,
      "grad_norm": 0.17708703875541687,
      "learning_rate": 0.00019987233784822395,
      "loss": 0.1965,
      "step": 18
    },
    {
      "epoch": 0.24398073836276082,
      "grad_norm": 0.18069303035736084,
      "learning_rate": 0.00019982625098055525,
      "loss": 0.1589,
      "step": 19
    },
    {
      "epoch": 0.2568218298555377,
      "grad_norm": 0.2211710810661316,
      "learning_rate": 0.0001997730826289277,
      "loss": 0.1806,
      "step": 20
    },
    {
      "epoch": 0.2696629213483146,
      "grad_norm": 0.2257445603609085,
      "learning_rate": 0.00019971283656500276,
      "loss": 0.1807,
      "step": 21
    },
    {
      "epoch": 0.2825040128410915,
      "grad_norm": 0.22210289537906647,
      "learning_rate": 0.00019964551706252144,
      "loss": 0.1869,
      "step": 22
    },
    {
      "epoch": 0.2953451043338684,
      "grad_norm": 0.19516178965568542,
      "learning_rate": 0.00019957112889700085,
      "loss": 0.1646,
      "step": 23
    },
    {
      "epoch": 0.30818619582664525,
      "grad_norm": 0.22516900300979614,
      "learning_rate": 0.00019948967734539571,
      "loss": 0.164,
      "step": 24
    },
    {
      "epoch": 0.32102728731942215,
      "grad_norm": 0.2650158107280731,
      "learning_rate": 0.0001994011681857238,
      "loss": 0.1733,
      "step": 25
    },
    {
      "epoch": 0.33386837881219905,
      "grad_norm": 0.16395625472068787,
      "learning_rate": 0.00019930560769665617,
      "loss": 0.1448,
      "step": 26
    },
    {
      "epoch": 0.3467094703049759,
      "grad_norm": 0.21180006861686707,
      "learning_rate": 0.00019920300265707184,
      "loss": 0.1933,
      "step": 27
    },
    {
      "epoch": 0.3595505617977528,
      "grad_norm": 0.17538852989673615,
      "learning_rate": 0.0001990933603455767,
      "loss": 0.1767,
      "step": 28
    },
    {
      "epoch": 0.3723916532905297,
      "grad_norm": 0.19017687439918518,
      "learning_rate": 0.00019897668853998725,
      "loss": 0.2061,
      "step": 29
    },
    {
      "epoch": 0.3852327447833066,
      "grad_norm": 0.1627688705921173,
      "learning_rate": 0.0001988529955167791,
      "loss": 0.165,
      "step": 30
    },
    {
      "epoch": 0.39807383627608345,
      "grad_norm": 0.23488560318946838,
      "learning_rate": 0.00019872229005049946,
      "loss": 0.1807,
      "step": 31
    },
    {
      "epoch": 0.41091492776886035,
      "grad_norm": 0.16365641355514526,
      "learning_rate": 0.00019858458141314503,
      "loss": 0.1606,
      "step": 32
    },
    {
      "epoch": 0.42375601926163725,
      "grad_norm": 0.1977553814649582,
      "learning_rate": 0.00019843987937350396,
      "loss": 0.1885,
      "step": 33
    },
    {
      "epoch": 0.43659711075441415,
      "grad_norm": 0.24794891476631165,
      "learning_rate": 0.00019828819419646316,
      "loss": 0.1695,
      "step": 34
    },
    {
      "epoch": 0.449438202247191,
      "grad_norm": 0.19939230382442474,
      "learning_rate": 0.00019812953664228,
      "loss": 0.1516,
      "step": 35
    },
    {
      "epoch": 0.4622792937399679,
      "grad_norm": 0.2248964011669159,
      "learning_rate": 0.00019796391796581887,
      "loss": 0.1509,
      "step": 36
    },
    {
      "epoch": 0.4751203852327448,
      "grad_norm": 0.2671429514884949,
      "learning_rate": 0.000197791349915753,
      "loss": 0.1936,
      "step": 37
    },
    {
      "epoch": 0.48796147672552165,
      "grad_norm": 0.1820429414510727,
      "learning_rate": 0.00019761184473373095,
      "loss": 0.1297,
      "step": 38
    },
    {
      "epoch": 0.5008025682182986,
      "grad_norm": 0.15517914295196533,
      "learning_rate": 0.00019742541515350813,
      "loss": 0.1664,
      "step": 39
    },
    {
      "epoch": 0.5136436597110754,
      "grad_norm": 0.1475019007921219,
      "learning_rate": 0.00019723207440004362,
      "loss": 0.1606,
      "step": 40
    },
    {
      "epoch": 0.5264847512038523,
      "grad_norm": 0.1478549689054489,
      "learning_rate": 0.0001970318361885619,
      "loss": 0.1629,
      "step": 41
    },
    {
      "epoch": 0.5393258426966292,
      "grad_norm": 0.1522587686777115,
      "learning_rate": 0.00019682471472358003,
      "loss": 0.1854,
      "step": 42
    },
    {
      "epoch": 0.5521669341894061,
      "grad_norm": 0.16658446192741394,
      "learning_rate": 0.00019661072469789992,
      "loss": 0.16,
      "step": 43
    },
    {
      "epoch": 0.565008025682183,
      "grad_norm": 0.1563337743282318,
      "learning_rate": 0.0001963898812915661,
      "loss": 0.1646,
      "step": 44
    },
    {
      "epoch": 0.5778491171749599,
      "grad_norm": 0.18028098344802856,
      "learning_rate": 0.00019616220017078882,
      "loss": 0.1679,
      "step": 45
    },
    {
      "epoch": 0.5906902086677368,
      "grad_norm": 0.17881852388381958,
      "learning_rate": 0.00019592769748683287,
      "loss": 0.1567,
      "step": 46
    },
    {
      "epoch": 0.6035313001605136,
      "grad_norm": 0.18338599801063538,
      "learning_rate": 0.00019568638987487155,
      "loss": 0.1475,
      "step": 47
    },
    {
      "epoch": 0.6163723916532905,
      "grad_norm": 0.19950391352176666,
      "learning_rate": 0.0001954382944528069,
      "loss": 0.1604,
      "step": 48
    },
    {
      "epoch": 0.6292134831460674,
      "grad_norm": 0.2285371720790863,
      "learning_rate": 0.00019518342882005532,
      "loss": 0.1463,
      "step": 49
    },
    {
      "epoch": 0.6420545746388443,
      "grad_norm": 0.26193344593048096,
      "learning_rate": 0.00019492181105629886,
      "loss": 0.1649,
      "step": 50
    },
    {
      "epoch": 0.6548956661316212,
      "grad_norm": 0.14862482249736786,
      "learning_rate": 0.00019465345972020313,
      "loss": 0.1303,
      "step": 51
    },
    {
      "epoch": 0.6677367576243981,
      "grad_norm": 0.14517804980278015,
      "learning_rate": 0.00019437839384810028,
      "loss": 0.161,
      "step": 52
    },
    {
      "epoch": 0.680577849117175,
      "grad_norm": 0.14767995476722717,
      "learning_rate": 0.00019409663295263902,
      "loss": 0.162,
      "step": 53
    },
    {
      "epoch": 0.6934189406099518,
      "grad_norm": 0.15824687480926514,
      "learning_rate": 0.00019380819702140016,
      "loss": 0.1555,
      "step": 54
    },
    {
      "epoch": 0.7062600321027287,
      "grad_norm": 0.1738673746585846,
      "learning_rate": 0.00019351310651547884,
      "loss": 0.1807,
      "step": 55
    },
    {
      "epoch": 0.7191011235955056,
      "grad_norm": 0.1739288568496704,
      "learning_rate": 0.00019321138236803311,
      "loss": 0.1824,
      "step": 56
    },
    {
      "epoch": 0.7319422150882825,
      "grad_norm": 0.1934700608253479,
      "learning_rate": 0.0001929030459827988,
      "loss": 0.1986,
      "step": 57
    },
    {
      "epoch": 0.7447833065810594,
      "grad_norm": 0.18333880603313446,
      "learning_rate": 0.00019258811923257137,
      "loss": 0.1798,
      "step": 58
    },
    {
      "epoch": 0.7576243980738363,
      "grad_norm": 0.17110952734947205,
      "learning_rate": 0.00019226662445765417,
      "loss": 0.1279,
      "step": 59
    },
    {
      "epoch": 0.7704654895666132,
      "grad_norm": 0.18288837373256683,
      "learning_rate": 0.0001919385844642737,
      "loss": 0.1372,
      "step": 60
    },
    {
      "epoch": 0.78330658105939,
      "grad_norm": 0.21783368289470673,
      "learning_rate": 0.0001916040225229618,
      "loss": 0.1669,
      "step": 61
    },
    {
      "epoch": 0.7961476725521669,
      "grad_norm": 0.24256236851215363,
      "learning_rate": 0.00019126296236690485,
      "loss": 0.1504,
      "step": 62
    },
    {
      "epoch": 0.8089887640449438,
      "grad_norm": 0.16231147944927216,
      "learning_rate": 0.00019091542819026024,
      "loss": 0.1243,
      "step": 63
    },
    {
      "epoch": 0.8218298555377207,
      "grad_norm": 0.14093080163002014,
      "learning_rate": 0.0001905614446464399,
      "loss": 0.1547,
      "step": 64
    },
    {
      "epoch": 0.8346709470304976,
      "grad_norm": 0.1700276881456375,
      "learning_rate": 0.00019020103684636177,
      "loss": 0.1762,
      "step": 65
    },
    {
      "epoch": 0.8475120385232745,
      "grad_norm": 0.15964765846729279,
      "learning_rate": 0.00018983423035666817,
      "loss": 0.1593,
      "step": 66
    },
    {
      "epoch": 0.8603531300160514,
      "grad_norm": 0.14215022325515747,
      "learning_rate": 0.0001894610511979123,
      "loss": 0.1454,
      "step": 67
    },
    {
      "epoch": 0.8731942215088283,
      "grad_norm": 0.1703556478023529,
      "learning_rate": 0.00018908152584271227,
      "loss": 0.1681,
      "step": 68
    },
    {
      "epoch": 0.8860353130016051,
      "grad_norm": 0.15497173368930817,
      "learning_rate": 0.00018869568121387343,
      "loss": 0.1503,
      "step": 69
    },
    {
      "epoch": 0.898876404494382,
      "grad_norm": 0.15470866858959198,
      "learning_rate": 0.00018830354468247817,
      "loss": 0.1417,
      "step": 70
    },
    {
      "epoch": 0.9117174959871589,
      "grad_norm": 0.17999379336833954,
      "learning_rate": 0.00018790514406594465,
      "loss": 0.1601,
      "step": 71
    },
    {
      "epoch": 0.9245585874799358,
      "grad_norm": 0.15672680735588074,
      "learning_rate": 0.00018750050762605312,
      "loss": 0.1282,
      "step": 72
    },
    {
      "epoch": 0.9373996789727127,
      "grad_norm": 0.21854081749916077,
      "learning_rate": 0.0001870896640669413,
      "loss": 0.1523,
      "step": 73
    },
    {
      "epoch": 0.9502407704654896,
      "grad_norm": 0.19732460379600525,
      "learning_rate": 0.00018667264253306823,
      "loss": 0.1554,
      "step": 74
    },
    {
      "epoch": 0.9630818619582665,
      "grad_norm": 0.24896161258220673,
      "learning_rate": 0.00018624947260714652,
      "loss": 0.1747,
      "step": 75
    },
    {
      "epoch": 0.9759229534510433,
      "grad_norm": 0.13258929550647736,
      "learning_rate": 0.0001858201843080441,
      "loss": 0.1484,
      "step": 76
    },
    {
      "epoch": 0.9887640449438202,
      "grad_norm": 0.17648810148239136,
      "learning_rate": 0.00018538480808865464,
      "loss": 0.1701,
      "step": 77
    },
    {
      "epoch": 1.0112359550561798,
      "grad_norm": 0.5716731548309326,
      "learning_rate": 0.00018494337483373726,
      "loss": 0.248,
      "step": 78
    },
    {
      "epoch": 1.0240770465489566,
      "grad_norm": 0.21971853077411652,
      "learning_rate": 0.00018449591585772553,
      "loss": 0.1502,
      "step": 79
    },
    {
      "epoch": 1.0369181380417336,
      "grad_norm": 0.20206806063652039,
      "learning_rate": 0.00018404246290250638,
      "loss": 0.1179,
      "step": 80
    },
    {
      "epoch": 1.0497592295345104,
      "grad_norm": 0.4210855960845947,
      "learning_rate": 0.000183583048135168,
      "loss": 0.1414,
      "step": 81
    },
    {
      "epoch": 1.0626003210272874,
      "grad_norm": 0.2643897831439972,
      "learning_rate": 0.00018311770414571835,
      "loss": 0.1228,
      "step": 82
    },
    {
      "epoch": 1.0754414125200642,
      "grad_norm": 0.27934709191322327,
      "learning_rate": 0.000182646463944773,
      "loss": 0.1244,
      "step": 83
    },
    {
      "epoch": 1.088282504012841,
      "grad_norm": 0.2761859893798828,
      "learning_rate": 0.00018216936096121348,
      "loss": 0.1373,
      "step": 84
    },
    {
      "epoch": 1.101123595505618,
      "grad_norm": 0.3699813187122345,
      "learning_rate": 0.00018168642903981607,
      "loss": 0.123,
      "step": 85
    },
    {
      "epoch": 1.1139646869983948,
      "grad_norm": 0.38410046696662903,
      "learning_rate": 0.00018119770243885065,
      "loss": 0.126,
      "step": 86
    },
    {
      "epoch": 1.1268057784911718,
      "grad_norm": 0.36278554797172546,
      "learning_rate": 0.0001807032158276508,
      "loss": 0.1081,
      "step": 87
    },
    {
      "epoch": 1.1396468699839486,
      "grad_norm": 0.4462052583694458,
      "learning_rate": 0.00018020300428415407,
      "loss": 0.1134,
      "step": 88
    },
    {
      "epoch": 1.1524879614767256,
      "grad_norm": 0.4787779152393341,
      "learning_rate": 0.00017969710329241386,
      "loss": 0.1165,
      "step": 89
    },
    {
      "epoch": 1.1653290529695024,
      "grad_norm": 0.34396493434906006,
      "learning_rate": 0.00017918554874008226,
      "loss": 0.1239,
      "step": 90
    },
    {
      "epoch": 1.1781701444622792,
      "grad_norm": 0.2252732515335083,
      "learning_rate": 0.00017866837691586404,
      "loss": 0.1598,
      "step": 91
    },
    {
      "epoch": 1.1910112359550562,
      "grad_norm": 0.28776365518569946,
      "learning_rate": 0.00017814562450694266,
      "loss": 0.1315,
      "step": 92
    },
    {
      "epoch": 1.203852327447833,
      "grad_norm": 0.2533372938632965,
      "learning_rate": 0.00017761732859637746,
      "loss": 0.1319,
      "step": 93
    },
    {
      "epoch": 1.21669341894061,
      "grad_norm": 0.258912593126297,
      "learning_rate": 0.0001770835266604734,
      "loss": 0.1492,
      "step": 94
    },
    {
      "epoch": 1.2295345104333868,
      "grad_norm": 0.2596625089645386,
      "learning_rate": 0.0001765442565661222,
      "loss": 0.1362,
      "step": 95
    },
    {
      "epoch": 1.2423756019261638,
      "grad_norm": 0.3291197717189789,
      "learning_rate": 0.00017599955656811653,
      "loss": 0.1325,
      "step": 96
    },
    {
      "epoch": 1.2552166934189406,
      "grad_norm": 0.3218838572502136,
      "learning_rate": 0.0001754494653064359,
      "loss": 0.1297,
      "step": 97
    },
    {
      "epoch": 1.2680577849117176,
      "grad_norm": 0.4448782801628113,
      "learning_rate": 0.00017489402180350582,
      "loss": 0.1291,
      "step": 98
    },
    {
      "epoch": 1.2808988764044944,
      "grad_norm": 0.33252349495887756,
      "learning_rate": 0.00017433326546142968,
      "loss": 0.1123,
      "step": 99
    },
    {
      "epoch": 1.2937399678972712,
      "grad_norm": 0.4413713216781616,
      "learning_rate": 0.00017376723605919345,
      "loss": 0.124,
      "step": 100
    },
    {
      "epoch": 1.3065810593900482,
      "grad_norm": 0.407705157995224,
      "learning_rate": 0.00017319597374984395,
      "loss": 0.1194,
      "step": 101
    },
    {
      "epoch": 1.319422150882825,
      "grad_norm": 0.4282243251800537,
      "learning_rate": 0.00017261951905764056,
      "loss": 0.1281,
      "step": 102
    },
    {
      "epoch": 1.332263242375602,
      "grad_norm": 0.30847808718681335,
      "learning_rate": 0.00017203791287518028,
      "loss": 0.127,
      "step": 103
    },
    {
      "epoch": 1.3451043338683788,
      "grad_norm": 0.3466252386569977,
      "learning_rate": 0.00017145119646049705,
      "loss": 0.1737,
      "step": 104
    },
    {
      "epoch": 1.3579454253611556,
      "grad_norm": 0.409016489982605,
      "learning_rate": 0.00017085941143413496,
      "loss": 0.1345,
      "step": 105
    },
    {
      "epoch": 1.3707865168539326,
      "grad_norm": 0.3255477547645569,
      "learning_rate": 0.0001702625997761957,
      "loss": 0.1621,
      "step": 106
    },
    {
      "epoch": 1.3836276083467094,
      "grad_norm": 0.2995210587978363,
      "learning_rate": 0.00016966080382336074,
      "loss": 0.1752,
      "step": 107
    },
    {
      "epoch": 1.3964686998394864,
      "grad_norm": 0.3129432797431946,
      "learning_rate": 0.00016905406626588777,
      "loss": 0.1763,
      "step": 108
    },
    {
      "epoch": 1.4093097913322632,
      "grad_norm": 0.31171512603759766,
      "learning_rate": 0.00016844243014458262,
      "loss": 0.1328,
      "step": 109
    },
    {
      "epoch": 1.4221508828250402,
      "grad_norm": 0.29522988200187683,
      "learning_rate": 0.00016782593884774586,
      "loss": 0.1394,
      "step": 110
    },
    {
      "epoch": 1.434991974317817,
      "grad_norm": 0.31598663330078125,
      "learning_rate": 0.0001672046361080949,
      "loss": 0.1351,
      "step": 111
    },
    {
      "epoch": 1.447833065810594,
      "grad_norm": 0.3824458718299866,
      "learning_rate": 0.00016657856599966182,
      "loss": 0.1334,
      "step": 112
    },
    {
      "epoch": 1.4606741573033708,
      "grad_norm": 0.39797738194465637,
      "learning_rate": 0.0001659477729346667,
      "loss": 0.1208,
      "step": 113
    },
    {
      "epoch": 1.4735152487961476,
      "grad_norm": 0.4944222867488861,
      "learning_rate": 0.0001653123016603672,
      "loss": 0.1392,
      "step": 114
    },
    {
      "epoch": 1.4863563402889246,
      "grad_norm": 0.30534660816192627,
      "learning_rate": 0.0001646721972558842,
      "loss": 0.1243,
      "step": 115
    },
    {
      "epoch": 1.4991974317817014,
      "grad_norm": 0.33155277371406555,
      "learning_rate": 0.00016402750512900397,
      "loss": 0.1633,
      "step": 116
    },
    {
      "epoch": 1.5120385232744784,
      "grad_norm": 0.343932569026947,
      "learning_rate": 0.0001633782710129571,
      "loss": 0.1701,
      "step": 117
    },
    {
      "epoch": 1.5248796147672552,
      "grad_norm": 0.27937018871307373,
      "learning_rate": 0.00016272454096317432,
      "loss": 0.1549,
      "step": 118
    },
    {
      "epoch": 1.537720706260032,
      "grad_norm": 0.35195839405059814,
      "learning_rate": 0.00016206636135401913,
      "loss": 0.1805,
      "step": 119
    },
    {
      "epoch": 1.550561797752809,
      "grad_norm": 0.3309479057788849,
      "learning_rate": 0.00016140377887549843,
      "loss": 0.138,
      "step": 120
    },
    {
      "epoch": 1.563402889245586,
      "grad_norm": 0.2956474721431732,
      "learning_rate": 0.0001607368405299503,
      "loss": 0.1484,
      "step": 121
    },
    {
      "epoch": 1.5762439807383628,
      "grad_norm": 0.300729364156723,
      "learning_rate": 0.00016006559362870964,
      "loss": 0.1577,
      "step": 122
    },
    {
      "epoch": 1.5890850722311396,
      "grad_norm": 0.35056933760643005,
      "learning_rate": 0.00015939008578875214,
      "loss": 0.1341,
      "step": 123
    },
    {
      "epoch": 1.6019261637239164,
      "grad_norm": 0.3127368688583374,
      "learning_rate": 0.00015871036492931632,
      "loss": 0.1245,
      "step": 124
    },
    {
      "epoch": 1.6147672552166934,
      "grad_norm": 0.31962379813194275,
      "learning_rate": 0.00015802647926850424,
      "loss": 0.1269,
      "step": 125
    },
    {
      "epoch": 1.6276083467094704,
      "grad_norm": 0.30289211869239807,
      "learning_rate": 0.00015733847731986113,
      "loss": 0.1189,
      "step": 126
    },
    {
      "epoch": 1.6404494382022472,
      "grad_norm": 0.42945238947868347,
      "learning_rate": 0.00015664640788893376,
      "loss": 0.1386,
      "step": 127
    },
    {
      "epoch": 1.653290529695024,
      "grad_norm": 0.2242669314146042,
      "learning_rate": 0.0001559503200698084,
      "loss": 0.1136,
      "step": 128
    },
    {
      "epoch": 1.666131621187801,
      "grad_norm": 0.382952481508255,
      "learning_rate": 0.00015525026324162805,
      "loss": 0.1661,
      "step": 129
    },
    {
      "epoch": 1.6789727126805778,
      "grad_norm": 0.31802457571029663,
      "learning_rate": 0.00015454628706508962,
      "loss": 0.1715,
      "step": 130
    },
    {
      "epoch": 1.6918138041733548,
      "grad_norm": 0.31699663400650024,
      "learning_rate": 0.00015383844147892126,
      "loss": 0.1454,
      "step": 131
    },
    {
      "epoch": 1.7046548956661316,
      "grad_norm": 0.3550918400287628,
      "learning_rate": 0.00015312677669633952,
      "loss": 0.1454,
      "step": 132
    },
    {
      "epoch": 1.7174959871589084,
      "grad_norm": 0.41878822445869446,
      "learning_rate": 0.00015241134320148752,
      "loss": 0.1565,
      "step": 133
    },
    {
      "epoch": 1.7303370786516854,
      "grad_norm": 0.2940831780433655,
      "learning_rate": 0.00015169219174585372,
      "loss": 0.1405,
      "step": 134
    },
    {
      "epoch": 1.7431781701444624,
      "grad_norm": 0.40646418929100037,
      "learning_rate": 0.00015096937334467151,
      "loss": 0.148,
      "step": 135
    },
    {
      "epoch": 1.7560192616372392,
      "grad_norm": 0.296655535697937,
      "learning_rate": 0.00015024293927330047,
      "loss": 0.1325,
      "step": 136
    },
    {
      "epoch": 1.768860353130016,
      "grad_norm": 0.2861349582672119,
      "learning_rate": 0.00014951294106358887,
      "loss": 0.1137,
      "step": 137
    },
    {
      "epoch": 1.7817014446227928,
      "grad_norm": 0.35022491216659546,
      "learning_rate": 0.00014877943050021827,
      "loss": 0.1437,
      "step": 138
    },
    {
      "epoch": 1.7945425361155698,
      "grad_norm": 0.39922255277633667,
      "learning_rate": 0.0001480424596170298,
      "loss": 0.1271,
      "step": 139
    },
    {
      "epoch": 1.8073836276083468,
      "grad_norm": 0.21837952733039856,
      "learning_rate": 0.00014730208069333313,
      "loss": 0.1044,
      "step": 140
    },
    {
      "epoch": 1.8202247191011236,
      "grad_norm": 0.28346794843673706,
      "learning_rate": 0.00014655834625019787,
      "loss": 0.1535,
      "step": 141
    },
    {
      "epoch": 1.8330658105939004,
      "grad_norm": 0.3031107783317566,
      "learning_rate": 0.00014581130904672764,
      "loss": 0.1509,
      "step": 142
    },
    {
      "epoch": 1.8459069020866774,
      "grad_norm": 0.288959801197052,
      "learning_rate": 0.00014506102207631773,
      "loss": 0.1565,
      "step": 143
    },
    {
      "epoch": 1.8587479935794544,
      "grad_norm": 0.26415252685546875,
      "learning_rate": 0.00014430753856289565,
      "loss": 0.1542,
      "step": 144
    },
    {
      "epoch": 1.8715890850722312,
      "grad_norm": 0.29712775349617004,
      "learning_rate": 0.0001435509119571456,
      "loss": 0.1442,
      "step": 145
    },
    {
      "epoch": 1.884430176565008,
      "grad_norm": 0.3248273730278015,
      "learning_rate": 0.00014279119593271666,
      "loss": 0.142,
      "step": 146
    },
    {
      "epoch": 1.8972712680577848,
      "grad_norm": 0.34400874376296997,
      "learning_rate": 0.00014202844438241546,
      "loss": 0.1495,
      "step": 147
    },
    {
      "epoch": 1.9101123595505618,
      "grad_norm": 0.6406733989715576,
      "learning_rate": 0.00014126271141438315,
      "loss": 0.1178,
      "step": 148
    },
    {
      "epoch": 1.9229534510433388,
      "grad_norm": 0.26794880628585815,
      "learning_rate": 0.00014049405134825677,
      "loss": 0.1219,
      "step": 149
    },
    {
      "epoch": 1.9357945425361156,
      "grad_norm": 0.35011592507362366,
      "learning_rate": 0.00013972251871131625,
      "loss": 0.1417,
      "step": 150
    },
    {
      "epoch": 1.9486356340288924,
      "grad_norm": 0.4136745035648346,
      "learning_rate": 0.0001389481682346162,
      "loss": 0.1319,
      "step": 151
    },
    {
      "epoch": 1.9614767255216692,
      "grad_norm": 0.4271896779537201,
      "learning_rate": 0.00013817105484910334,
      "loss": 0.1226,
      "step": 152
    },
    {
      "epoch": 1.9743178170144462,
      "grad_norm": 0.26580625772476196,
      "learning_rate": 0.00013739123368171994,
      "loss": 0.1477,
      "step": 153
    },
    {
      "epoch": 1.9871589085072232,
      "grad_norm": 0.36320167779922485,
      "learning_rate": 0.00013660876005149318,
      "loss": 0.1446,
      "step": 154
    },
    {
      "epoch": 2.009630818619583,
      "grad_norm": 0.556969404220581,
      "learning_rate": 0.00013582368946561083,
      "loss": 0.2174,
      "step": 155
    },
    {
      "epoch": 2.0224719101123596,
      "grad_norm": 0.29500341415405273,
      "learning_rate": 0.00013503607761548384,
      "loss": 0.1428,
      "step": 156
    },
    {
      "epoch": 2.0353130016051364,
      "grad_norm": 0.3006628155708313,
      "learning_rate": 0.00013424598037279544,
      "loss": 0.1346,
      "step": 157
    },
    {
      "epoch": 2.048154093097913,
      "grad_norm": 0.3429463803768158,
      "learning_rate": 0.00013345345378553805,
      "loss": 0.1418,
      "step": 158
    },
    {
      "epoch": 2.0609951845906904,
      "grad_norm": 0.31199735403060913,
      "learning_rate": 0.000132658554074037,
      "loss": 0.1335,
      "step": 159
    },
    {
      "epoch": 2.073836276083467,
      "grad_norm": 0.35404714941978455,
      "learning_rate": 0.00013186133762696267,
      "loss": 0.1295,
      "step": 160
    },
    {
      "epoch": 2.086677367576244,
      "grad_norm": 0.2951603829860687,
      "learning_rate": 0.00013106186099733018,
      "loss": 0.1309,
      "step": 161
    },
    {
      "epoch": 2.099518459069021,
      "grad_norm": 0.32650235295295715,
      "learning_rate": 0.0001302601808984877,
      "loss": 0.1414,
      "step": 162
    },
    {
      "epoch": 2.1123595505617976,
      "grad_norm": 0.5657181143760681,
      "learning_rate": 0.0001294563542000933,
      "loss": 0.1141,
      "step": 163
    },
    {
      "epoch": 2.125200642054575,
      "grad_norm": 0.31900152564048767,
      "learning_rate": 0.0001286504379240807,
      "loss": 0.1105,
      "step": 164
    },
    {
      "epoch": 2.1380417335473516,
      "grad_norm": 0.3072754442691803,
      "learning_rate": 0.0001278424892406143,
      "loss": 0.118,
      "step": 165
    },
    {
      "epoch": 2.1508828250401284,
      "grad_norm": 0.36620378494262695,
      "learning_rate": 0.00012703256546403374,
      "loss": 0.1147,
      "step": 166
    },
    {
      "epoch": 2.163723916532905,
      "grad_norm": 0.35902220010757446,
      "learning_rate": 0.00012622072404878774,
      "loss": 0.0722,
      "step": 167
    },
    {
      "epoch": 2.176565008025682,
      "grad_norm": 0.3162679970264435,
      "learning_rate": 0.0001254070225853589,
      "loss": 0.1503,
      "step": 168
    },
    {
      "epoch": 2.189406099518459,
      "grad_norm": 0.2725870907306671,
      "learning_rate": 0.00012459151879617785,
      "loss": 0.1488,
      "step": 169
    },
    {
      "epoch": 2.202247191011236,
      "grad_norm": 0.28934144973754883,
      "learning_rate": 0.00012377427053152903,
      "loss": 0.1199,
      "step": 170
    },
    {
      "epoch": 2.215088282504013,
      "grad_norm": 0.23648247122764587,
      "learning_rate": 0.00012295533576544648,
      "loss": 0.1137,
      "step": 171
    },
    {
      "epoch": 2.2279293739967896,
      "grad_norm": 0.3106015622615814,
      "learning_rate": 0.00012213477259160146,
      "loss": 0.1433,
      "step": 172
    },
    {
      "epoch": 2.240770465489567,
      "grad_norm": 0.25510647892951965,
      "learning_rate": 0.00012131263921918143,
      "loss": 0.1107,
      "step": 173
    },
    {
      "epoch": 2.2536115569823436,
      "grad_norm": 0.33392271399497986,
      "learning_rate": 0.00012048899396876065,
      "loss": 0.1245,
      "step": 174
    },
    {
      "epoch": 2.2664526484751204,
      "grad_norm": 0.2912399172782898,
      "learning_rate": 0.00011966389526816322,
      "loss": 0.1124,
      "step": 175
    },
    {
      "epoch": 2.279293739967897,
      "grad_norm": 0.30920150876045227,
      "learning_rate": 0.00011883740164831818,
      "loss": 0.0881,
      "step": 176
    },
    {
      "epoch": 2.292134831460674,
      "grad_norm": 0.3264816701412201,
      "learning_rate": 0.00011800957173910748,
      "loss": 0.1108,
      "step": 177
    },
    {
      "epoch": 2.304975922953451,
      "grad_norm": 0.33401617407798767,
      "learning_rate": 0.00011718046426520689,
      "loss": 0.1088,
      "step": 178
    },
    {
      "epoch": 2.317817014446228,
      "grad_norm": 0.3508036434650421,
      "learning_rate": 0.00011635013804192015,
      "loss": 0.1157,
      "step": 179
    },
    {
      "epoch": 2.330658105939005,
      "grad_norm": 0.31395018100738525,
      "learning_rate": 0.00011551865197100686,
      "loss": 0.1019,
      "step": 180
    },
    {
      "epoch": 2.3434991974317816,
      "grad_norm": 0.27874499559402466,
      "learning_rate": 0.00011468606503650394,
      "loss": 0.1243,
      "step": 181
    },
    {
      "epoch": 2.3563402889245584,
      "grad_norm": 0.24277067184448242,
      "learning_rate": 0.00011385243630054144,
      "loss": 0.1447,
      "step": 182
    },
    {
      "epoch": 2.3691813804173356,
      "grad_norm": 0.2734696567058563,
      "learning_rate": 0.00011301782489915287,
      "loss": 0.1438,
      "step": 183
    },
    {
      "epoch": 2.3820224719101124,
      "grad_norm": 0.2731485068798065,
      "learning_rate": 0.00011218229003808012,
      "loss": 0.1431,
      "step": 184
    },
    {
      "epoch": 2.394863563402889,
      "grad_norm": 0.47293299436569214,
      "learning_rate": 0.00011134589098857356,
      "loss": 0.1287,
      "step": 185
    },
    {
      "epoch": 2.407704654895666,
      "grad_norm": 0.3162679076194763,
      "learning_rate": 0.00011050868708318747,
      "loss": 0.1217,
      "step": 186
    },
    {
      "epoch": 2.420545746388443,
      "grad_norm": 0.39970093965530396,
      "learning_rate": 0.00010967073771157099,
      "loss": 0.1352,
      "step": 187
    },
    {
      "epoch": 2.43338683788122,
      "grad_norm": 0.2743336856365204,
      "learning_rate": 0.00010883210231625534,
      "loss": 0.0977,
      "step": 188
    },
    {
      "epoch": 2.446227929373997,
      "grad_norm": 0.37246617674827576,
      "learning_rate": 0.00010799284038843686,
      "loss": 0.1047,
      "step": 189
    },
    {
      "epoch": 2.4590690208667736,
      "grad_norm": 0.3695964813232422,
      "learning_rate": 0.00010715301146375694,
      "loss": 0.1127,
      "step": 190
    },
    {
      "epoch": 2.4719101123595504,
      "grad_norm": 0.35048359632492065,
      "learning_rate": 0.00010631267511807861,
      "loss": 0.1153,
      "step": 191
    },
    {
      "epoch": 2.4847512038523276,
      "grad_norm": 0.20369504392147064,
      "learning_rate": 0.0001054718909632604,
      "loss": 0.0828,
      "step": 192
    },
    {
      "epoch": 2.4975922953451044,
      "grad_norm": 0.3476305902004242,
      "learning_rate": 0.00010463071864292764,
      "loss": 0.1384,
      "step": 193
    },
    {
      "epoch": 2.510433386837881,
      "grad_norm": 0.2885417640209198,
      "learning_rate": 0.00010378921782824128,
      "loss": 0.1496,
      "step": 194
    },
    {
      "epoch": 2.523274478330658,
      "grad_norm": 0.2671290338039398,
      "learning_rate": 0.00010294744821366504,
      "loss": 0.14,
      "step": 195
    },
    {
      "epoch": 2.5361155698234352,
      "grad_norm": 0.30776816606521606,
      "learning_rate": 0.0001021054695127309,
      "loss": 0.1442,
      "step": 196
    },
    {
      "epoch": 2.548956661316212,
      "grad_norm": 0.2914685904979706,
      "learning_rate": 0.00010126334145380288,
      "loss": 0.1321,
      "step": 197
    },
    {
      "epoch": 2.561797752808989,
      "grad_norm": 0.23030787706375122,
      "learning_rate": 0.00010042112377584028,
      "loss": 0.1177,
      "step": 198
    },
    {
      "epoch": 2.5746388443017656,
      "grad_norm": 0.2737199068069458,
      "learning_rate": 9.957887622415975e-05,
      "loss": 0.1198,
      "step": 199
    },
    {
      "epoch": 2.5874799357945424,
      "grad_norm": 0.26640573143959045,
      "learning_rate": 9.873665854619715e-05,
      "loss": 0.1084,
      "step": 200
    },
    {
      "epoch": 2.600321027287319,
      "grad_norm": 0.3201727569103241,
      "learning_rate": 9.789453048726912e-05,
      "loss": 0.1022,
      "step": 201
    },
    {
      "epoch": 2.6131621187800964,
      "grad_norm": 0.262690007686615,
      "learning_rate": 9.705255178633497e-05,
      "loss": 0.094,
      "step": 202
    },
    {
      "epoch": 2.626003210272873,
      "grad_norm": 0.43916091322898865,
      "learning_rate": 9.621078217175876e-05,
      "loss": 0.1059,
      "step": 203
    },
    {
      "epoch": 2.63884430176565,
      "grad_norm": 0.3432372212409973,
      "learning_rate": 9.53692813570724e-05,
      "loss": 0.1142,
      "step": 204
    },
    {
      "epoch": 2.6516853932584272,
      "grad_norm": 0.23425506055355072,
      "learning_rate": 9.452810903673963e-05,
      "loss": 0.1118,
      "step": 205
    },
    {
      "epoch": 2.664526484751204,
      "grad_norm": 0.2430526316165924,
      "learning_rate": 9.368732488192143e-05,
      "loss": 0.0987,
      "step": 206
    },
    {
      "epoch": 2.677367576243981,
      "grad_norm": 0.28097087144851685,
      "learning_rate": 9.28469885362431e-05,
      "loss": 0.1198,
      "step": 207
    },
    {
      "epoch": 2.6902086677367576,
      "grad_norm": 0.2345268875360489,
      "learning_rate": 9.200715961156317e-05,
      "loss": 0.1115,
      "step": 208
    },
    {
      "epoch": 2.7030497592295344,
      "grad_norm": 0.4230293035507202,
      "learning_rate": 9.116789768374467e-05,
      "loss": 0.1497,
      "step": 209
    },
    {
      "epoch": 2.715890850722311,
      "grad_norm": 0.2744080424308777,
      "learning_rate": 9.032926228842902e-05,
      "loss": 0.1296,
      "step": 210
    },
    {
      "epoch": 2.7287319422150884,
      "grad_norm": 0.3323458135128021,
      "learning_rate": 8.949131291681257e-05,
      "loss": 0.143,
      "step": 211
    },
    {
      "epoch": 2.741573033707865,
      "grad_norm": 0.2832281291484833,
      "learning_rate": 8.865410901142645e-05,
      "loss": 0.1154,
      "step": 212
    },
    {
      "epoch": 2.754414125200642,
      "grad_norm": 0.3109094798564911,
      "learning_rate": 8.781770996191992e-05,
      "loss": 0.1146,
      "step": 213
    },
    {
      "epoch": 2.767255216693419,
      "grad_norm": 0.27764928340911865,
      "learning_rate": 8.698217510084717e-05,
      "loss": 0.0992,
      "step": 214
    },
    {
      "epoch": 2.780096308186196,
      "grad_norm": 0.3414210379123688,
      "learning_rate": 8.614756369945856e-05,
      "loss": 0.0996,
      "step": 215
    },
    {
      "epoch": 2.792937399678973,
      "grad_norm": 0.276713490486145,
      "learning_rate": 8.531393496349606e-05,
      "loss": 0.0948,
      "step": 216
    },
    {
      "epoch": 2.8057784911717496,
      "grad_norm": 0.24623610079288483,
      "learning_rate": 8.448134802899314e-05,
      "loss": 0.0819,
      "step": 217
    },
    {
      "epoch": 2.8186195826645264,
      "grad_norm": 0.21359184384346008,
      "learning_rate": 8.364986195807986e-05,
      "loss": 0.1337,
      "step": 218
    },
    {
      "epoch": 2.831460674157303,
      "grad_norm": 0.28975728154182434,
      "learning_rate": 8.281953573479315e-05,
      "loss": 0.1421,
      "step": 219
    },
    {
      "epoch": 2.8443017656500804,
      "grad_norm": 0.25927087664604187,
      "learning_rate": 8.199042826089252e-05,
      "loss": 0.1162,
      "step": 220
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.2766461968421936,
      "learning_rate": 8.116259835168183e-05,
      "loss": 0.1,
      "step": 221
    },
    {
      "epoch": 2.869983948635634,
      "grad_norm": 0.28379619121551514,
      "learning_rate": 8.033610473183678e-05,
      "loss": 0.1001,
      "step": 222
    },
    {
      "epoch": 2.882825040128411,
      "grad_norm": 0.2919027805328369,
      "learning_rate": 7.951100603123936e-05,
      "loss": 0.1081,
      "step": 223
    },
    {
      "epoch": 2.895666131621188,
      "grad_norm": 0.26831895112991333,
      "learning_rate": 7.868736078081859e-05,
      "loss": 0.1092,
      "step": 224
    },
    {
      "epoch": 2.908507223113965,
      "grad_norm": 0.2470114827156067,
      "learning_rate": 7.786522740839852e-05,
      "loss": 0.1069,
      "step": 225
    },
    {
      "epoch": 2.9213483146067416,
      "grad_norm": 0.27561888098716736,
      "learning_rate": 7.704466423455353e-05,
      "loss": 0.0981,
      "step": 226
    },
    {
      "epoch": 2.9341894060995184,
      "grad_norm": 0.26214268803596497,
      "learning_rate": 7.622572946847098e-05,
      "loss": 0.0933,
      "step": 227
    },
    {
      "epoch": 2.947030497592295,
      "grad_norm": 0.3250983655452728,
      "learning_rate": 7.540848120382214e-05,
      "loss": 0.1005,
      "step": 228
    },
    {
      "epoch": 2.959871589085072,
      "grad_norm": 0.3818821609020233,
      "learning_rate": 7.459297741464115e-05,
      "loss": 0.0977,
      "step": 229
    },
    {
      "epoch": 2.972712680577849,
      "grad_norm": 0.1819436401128769,
      "learning_rate": 7.37792759512123e-05,
      "loss": 0.1218,
      "step": 230
    },
    {
      "epoch": 2.985553772070626,
      "grad_norm": 0.28734883666038513,
      "learning_rate": 7.296743453596631e-05,
      "loss": 0.1154,
      "step": 231
    },
    {
      "epoch": 3.0080256821829856,
      "grad_norm": 0.5932813882827759,
      "learning_rate": 7.215751075938571e-05,
      "loss": 0.2066,
      "step": 232
    },
    {
      "epoch": 3.0208667736757624,
      "grad_norm": 0.19985750317573547,
      "learning_rate": 7.134956207591936e-05,
      "loss": 0.1257,
      "step": 233
    },
    {
      "epoch": 3.033707865168539,
      "grad_norm": 0.2326143980026245,
      "learning_rate": 7.054364579990677e-05,
      "loss": 0.1019,
      "step": 234
    },
    {
      "epoch": 3.0465489566613164,
      "grad_norm": 0.24029302597045898,
      "learning_rate": 6.973981910151234e-05,
      "loss": 0.1013,
      "step": 235
    },
    {
      "epoch": 3.059390048154093,
      "grad_norm": 0.19977405667304993,
      "learning_rate": 6.893813900266987e-05,
      "loss": 0.087,
      "step": 236
    },
    {
      "epoch": 3.07223113964687,
      "grad_norm": 0.2730342447757721,
      "learning_rate": 6.813866237303735e-05,
      "loss": 0.0991,
      "step": 237
    },
    {
      "epoch": 3.085072231139647,
      "grad_norm": 0.2400568425655365,
      "learning_rate": 6.734144592596303e-05,
      "loss": 0.1135,
      "step": 238
    },
    {
      "epoch": 3.0979133226324236,
      "grad_norm": 0.22795532643795013,
      "learning_rate": 6.654654621446202e-05,
      "loss": 0.1163,
      "step": 239
    },
    {
      "epoch": 3.110754414125201,
      "grad_norm": 0.2850149869918823,
      "learning_rate": 6.57540196272046e-05,
      "loss": 0.1007,
      "step": 240
    },
    {
      "epoch": 3.1235955056179776,
      "grad_norm": 0.2573602497577667,
      "learning_rate": 6.496392238451621e-05,
      "loss": 0.0706,
      "step": 241
    },
    {
      "epoch": 3.1364365971107544,
      "grad_norm": 0.27073922753334045,
      "learning_rate": 6.417631053438917e-05,
      "loss": 0.0842,
      "step": 242
    },
    {
      "epoch": 3.149277688603531,
      "grad_norm": 0.24778737127780914,
      "learning_rate": 6.339123994850684e-05,
      "loss": 0.0822,
      "step": 243
    },
    {
      "epoch": 3.162118780096308,
      "grad_norm": 0.24561458826065063,
      "learning_rate": 6.26087663182801e-05,
      "loss": 0.0441,
      "step": 244
    },
    {
      "epoch": 3.174959871589085,
      "grad_norm": 0.27637338638305664,
      "learning_rate": 6.182894515089672e-05,
      "loss": 0.1313,
      "step": 245
    },
    {
      "epoch": 3.187800963081862,
      "grad_norm": 0.2668153941631317,
      "learning_rate": 6.105183176538385e-05,
      "loss": 0.1261,
      "step": 246
    },
    {
      "epoch": 3.200642054574639,
      "grad_norm": 0.2336243838071823,
      "learning_rate": 6.0277481288683765e-05,
      "loss": 0.0968,
      "step": 247
    },
    {
      "epoch": 3.2134831460674156,
      "grad_norm": 0.2731676995754242,
      "learning_rate": 5.950594865174325e-05,
      "loss": 0.1,
      "step": 248
    },
    {
      "epoch": 3.226324237560193,
      "grad_norm": 0.19752222299575806,
      "learning_rate": 5.8737288585616867e-05,
      "loss": 0.1018,
      "step": 249
    },
    {
      "epoch": 3.2391653290529696,
      "grad_norm": 0.274467796087265,
      "learning_rate": 5.797155561758454e-05,
      "loss": 0.1193,
      "step": 250
    },
    {
      "epoch": 3.2520064205457464,
      "grad_norm": 0.2718445062637329,
      "learning_rate": 5.7208804067283396e-05,
      "loss": 0.0881,
      "step": 251
    },
    {
      "epoch": 3.264847512038523,
      "grad_norm": 0.2557855248451233,
      "learning_rate": 5.644908804285447e-05,
      "loss": 0.0833,
      "step": 252
    },
    {
      "epoch": 3.2776886035313,
      "grad_norm": 0.2935195863246918,
      "learning_rate": 5.5692461437104335e-05,
      "loss": 0.0881,
      "step": 253
    },
    {
      "epoch": 3.290529695024077,
      "grad_norm": 0.2287595272064209,
      "learning_rate": 5.493897792368228e-05,
      "loss": 0.0688,
      "step": 254
    },
    {
      "epoch": 3.303370786516854,
      "grad_norm": 0.2637540102005005,
      "learning_rate": 5.418869095327237e-05,
      "loss": 0.0665,
      "step": 255
    },
    {
      "epoch": 3.316211878009631,
      "grad_norm": 0.3496418595314026,
      "learning_rate": 5.344165374980217e-05,
      "loss": 0.0901,
      "step": 256
    },
    {
      "epoch": 3.3290529695024076,
      "grad_norm": 0.1895870864391327,
      "learning_rate": 5.269791930666688e-05,
      "loss": 0.0885,
      "step": 257
    },
    {
      "epoch": 3.341894060995185,
      "grad_norm": 0.2253326177597046,
      "learning_rate": 5.195754038297023e-05,
      "loss": 0.1087,
      "step": 258
    },
    {
      "epoch": 3.3547351524879616,
      "grad_norm": 0.18178820610046387,
      "learning_rate": 5.122056949978176e-05,
      "loss": 0.1029,
      "step": 259
    },
    {
      "epoch": 3.3675762439807384,
      "grad_norm": 0.26113295555114746,
      "learning_rate": 5.0487058936411144e-05,
      "loss": 0.1091,
      "step": 260
    },
    {
      "epoch": 3.380417335473515,
      "grad_norm": 0.2430870234966278,
      "learning_rate": 4.975706072669958e-05,
      "loss": 0.1085,
      "step": 261
    },
    {
      "epoch": 3.393258426966292,
      "grad_norm": 0.2758089005947113,
      "learning_rate": 4.9030626655328516e-05,
      "loss": 0.0981,
      "step": 262
    },
    {
      "epoch": 3.4060995184590688,
      "grad_norm": 0.26935628056526184,
      "learning_rate": 4.8307808254146266e-05,
      "loss": 0.0873,
      "step": 263
    },
    {
      "epoch": 3.418940609951846,
      "grad_norm": 0.2657538652420044,
      "learning_rate": 4.758865679851247e-05,
      "loss": 0.0843,
      "step": 264
    },
    {
      "epoch": 3.431781701444623,
      "grad_norm": 0.24748681485652924,
      "learning_rate": 4.68732233036605e-05,
      "loss": 0.0798,
      "step": 265
    },
    {
      "epoch": 3.4446227929373996,
      "grad_norm": 0.456748366355896,
      "learning_rate": 4.616155852107877e-05,
      "loss": 0.0851,
      "step": 266
    },
    {
      "epoch": 3.4574638844301764,
      "grad_norm": 0.23406369984149933,
      "learning_rate": 4.54537129349104e-05,
      "loss": 0.0831,
      "step": 267
    },
    {
      "epoch": 3.4703049759229536,
      "grad_norm": 0.24098779261112213,
      "learning_rate": 4.4749736758372e-05,
      "loss": 0.081,
      "step": 268
    },
    {
      "epoch": 3.4831460674157304,
      "grad_norm": 0.13690468668937683,
      "learning_rate": 4.404967993019162e-05,
      "loss": 0.0439,
      "step": 269
    },
    {
      "epoch": 3.495987158908507,
      "grad_norm": 0.1986464560031891,
      "learning_rate": 4.335359211106624e-05,
      "loss": 0.1084,
      "step": 270
    },
    {
      "epoch": 3.508828250401284,
      "grad_norm": 0.2006080001592636,
      "learning_rate": 4.26615226801389e-05,
      "loss": 0.1235,
      "step": 271
    },
    {
      "epoch": 3.521669341894061,
      "grad_norm": 0.2104363888502121,
      "learning_rate": 4.1973520731495765e-05,
      "loss": 0.1134,
      "step": 272
    },
    {
      "epoch": 3.534510433386838,
      "grad_norm": 0.24415647983551025,
      "learning_rate": 4.128963507068371e-05,
      "loss": 0.1146,
      "step": 273
    },
    {
      "epoch": 3.547351524879615,
      "grad_norm": 0.2873344421386719,
      "learning_rate": 4.0609914211247866e-05,
      "loss": 0.0897,
      "step": 274
    },
    {
      "epoch": 3.5601926163723916,
      "grad_norm": 0.19503550231456757,
      "learning_rate": 3.993440637129036e-05,
      "loss": 0.0896,
      "step": 275
    },
    {
      "epoch": 3.5730337078651684,
      "grad_norm": 0.27025559544563293,
      "learning_rate": 3.926315947004971e-05,
      "loss": 0.0979,
      "step": 276
    },
    {
      "epoch": 3.5858747993579456,
      "grad_norm": 0.3529897630214691,
      "learning_rate": 3.859622112450158e-05,
      "loss": 0.0784,
      "step": 277
    },
    {
      "epoch": 3.5987158908507224,
      "grad_norm": 0.31557080149650574,
      "learning_rate": 3.7933638645980905e-05,
      "loss": 0.0921,
      "step": 278
    },
    {
      "epoch": 3.611556982343499,
      "grad_norm": 0.25294047594070435,
      "learning_rate": 3.727545903682572e-05,
      "loss": 0.086,
      "step": 279
    },
    {
      "epoch": 3.624398073836276,
      "grad_norm": 0.20190520584583282,
      "learning_rate": 3.6621728987042905e-05,
      "loss": 0.0605,
      "step": 280
    },
    {
      "epoch": 3.637239165329053,
      "grad_norm": 0.2746667265892029,
      "learning_rate": 3.5972494870996063e-05,
      "loss": 0.077,
      "step": 281
    },
    {
      "epoch": 3.65008025682183,
      "grad_norm": 0.23425573110580444,
      "learning_rate": 3.532780274411581e-05,
      "loss": 0.0991,
      "step": 282
    },
    {
      "epoch": 3.662921348314607,
      "grad_norm": 0.15118825435638428,
      "learning_rate": 3.468769833963279e-05,
      "loss": 0.1051,
      "step": 283
    },
    {
      "epoch": 3.6757624398073836,
      "grad_norm": 0.23835481703281403,
      "learning_rate": 3.40522270653333e-05,
      "loss": 0.0939,
      "step": 284
    },
    {
      "epoch": 3.6886035313001604,
      "grad_norm": 0.17009741067886353,
      "learning_rate": 3.3421434000338194e-05,
      "loss": 0.085,
      "step": 285
    },
    {
      "epoch": 3.7014446227929376,
      "grad_norm": 0.23825782537460327,
      "learning_rate": 3.279536389190512e-05,
      "loss": 0.0982,
      "step": 286
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.21432097256183624,
      "learning_rate": 3.2174061152254166e-05,
      "loss": 0.079,
      "step": 287
    },
    {
      "epoch": 3.727126805778491,
      "grad_norm": 0.21313104033470154,
      "learning_rate": 3.1557569855417394e-05,
      "loss": 0.0852,
      "step": 288
    },
    {
      "epoch": 3.739967897271268,
      "grad_norm": 0.28463664650917053,
      "learning_rate": 3.094593373411224e-05,
      "loss": 0.0883,
      "step": 289
    },
    {
      "epoch": 3.752808988764045,
      "grad_norm": 0.24580109119415283,
      "learning_rate": 3.0339196176639283e-05,
      "loss": 0.0771,
      "step": 290
    },
    {
      "epoch": 3.7656500802568216,
      "grad_norm": 0.24096901714801788,
      "learning_rate": 2.973740022380428e-05,
      "loss": 0.0749,
      "step": 291
    },
    {
      "epoch": 3.778491171749599,
      "grad_norm": 0.25022003054618835,
      "learning_rate": 2.9140588565865077e-05,
      "loss": 0.0748,
      "step": 292
    },
    {
      "epoch": 3.7913322632423756,
      "grad_norm": 0.2789011597633362,
      "learning_rate": 2.8548803539502987e-05,
      "loss": 0.0717,
      "step": 293
    },
    {
      "epoch": 3.8041733547351524,
      "grad_norm": 0.1361372172832489,
      "learning_rate": 2.7962087124819757e-05,
      "loss": 0.0431,
      "step": 294
    },
    {
      "epoch": 3.8170144462279296,
      "grad_norm": 0.22512777149677277,
      "learning_rate": 2.7380480942359477e-05,
      "loss": 0.1221,
      "step": 295
    },
    {
      "epoch": 3.8298555377207064,
      "grad_norm": 0.2254047840833664,
      "learning_rate": 2.6804026250156077e-05,
      "loss": 0.1062,
      "step": 296
    },
    {
      "epoch": 3.842696629213483,
      "grad_norm": 0.27256065607070923,
      "learning_rate": 2.6232763940806606e-05,
      "loss": 0.1023,
      "step": 297
    },
    {
      "epoch": 3.85553772070626,
      "grad_norm": 0.2014126181602478,
      "learning_rate": 2.5666734538570337e-05,
      "loss": 0.0999,
      "step": 298
    },
    {
      "epoch": 3.868378812199037,
      "grad_norm": 0.27657970786094666,
      "learning_rate": 2.5105978196494184e-05,
      "loss": 0.0859,
      "step": 299
    },
    {
      "epoch": 3.8812199036918136,
      "grad_norm": 0.287130743265152,
      "learning_rate": 2.455053469356413e-05,
      "loss": 0.0946,
      "step": 300
    },
    {
      "epoch": 3.894060995184591,
      "grad_norm": 0.28707414865493774,
      "learning_rate": 2.400044343188349e-05,
      "loss": 0.0848,
      "step": 301
    },
    {
      "epoch": 3.9069020866773676,
      "grad_norm": 0.20366519689559937,
      "learning_rate": 2.3455743433877808e-05,
      "loss": 0.0774,
      "step": 302
    },
    {
      "epoch": 3.9197431781701444,
      "grad_norm": 0.2152596265077591,
      "learning_rate": 2.2916473339526644e-05,
      "loss": 0.0715,
      "step": 303
    },
    {
      "epoch": 3.932584269662921,
      "grad_norm": 0.26020580530166626,
      "learning_rate": 2.238267140362257e-05,
      "loss": 0.0721,
      "step": 304
    },
    {
      "epoch": 3.9454253611556984,
      "grad_norm": 0.3326077461242676,
      "learning_rate": 2.1854375493057388e-05,
      "loss": 0.0757,
      "step": 305
    },
    {
      "epoch": 3.958266452648475,
      "grad_norm": 0.26648905873298645,
      "learning_rate": 2.1331623084135976e-05,
      "loss": 0.0832,
      "step": 306
    },
    {
      "epoch": 3.971107544141252,
      "grad_norm": 0.21082034707069397,
      "learning_rate": 2.0814451259917767e-05,
      "loss": 0.0825,
      "step": 307
    },
    {
      "epoch": 3.983948635634029,
      "grad_norm": 0.19309554994106293,
      "learning_rate": 2.0302896707586127e-05,
      "loss": 0.0906,
      "step": 308
    },
    {
      "epoch": 4.006420545746389,
      "grad_norm": 0.35690435767173767,
      "learning_rate": 1.9796995715845955e-05,
      "loss": 0.1062,
      "step": 309
    },
    {
      "epoch": 4.019261637239166,
      "grad_norm": 0.1724165827035904,
      "learning_rate": 1.9296784172349223e-05,
      "loss": 0.1117,
      "step": 310
    },
    {
      "epoch": 4.032102728731942,
      "grad_norm": 0.17502835392951965,
      "learning_rate": 1.8802297561149363e-05,
      "loss": 0.1141,
      "step": 311
    },
    {
      "epoch": 4.044943820224719,
      "grad_norm": 0.21144062280654907,
      "learning_rate": 1.8313570960183977e-05,
      "loss": 0.0925,
      "step": 312
    },
    {
      "epoch": 4.057784911717496,
      "grad_norm": 0.25978124141693115,
      "learning_rate": 1.7830639038786555e-05,
      "loss": 0.0865,
      "step": 313
    },
    {
      "epoch": 4.070626003210273,
      "grad_norm": 0.20121395587921143,
      "learning_rate": 1.7353536055227047e-05,
      "loss": 0.0658,
      "step": 314
    },
    {
      "epoch": 4.08346709470305,
      "grad_norm": 0.3032616376876831,
      "learning_rate": 1.688229585428167e-05,
      "loss": 0.0741,
      "step": 315
    },
    {
      "epoch": 4.096308186195826,
      "grad_norm": 0.19196993112564087,
      "learning_rate": 1.6416951864832008e-05,
      "loss": 0.069,
      "step": 316
    },
    {
      "epoch": 4.109149277688603,
      "grad_norm": 0.2227143496274948,
      "learning_rate": 1.5957537097493637e-05,
      "loss": 0.064,
      "step": 317
    },
    {
      "epoch": 4.121990369181381,
      "grad_norm": 0.2821972072124481,
      "learning_rate": 1.5504084142274477e-05,
      "loss": 0.0563,
      "step": 318
    },
    {
      "epoch": 4.134831460674158,
      "grad_norm": 0.23713919520378113,
      "learning_rate": 1.5056625166262772e-05,
      "loss": 0.0657,
      "step": 319
    },
    {
      "epoch": 4.147672552166934,
      "grad_norm": 0.18978498876094818,
      "learning_rate": 1.4615191911345371e-05,
      "loss": 0.0609,
      "step": 320
    },
    {
      "epoch": 4.160513643659711,
      "grad_norm": 0.12965190410614014,
      "learning_rate": 1.4179815691955923e-05,
      "loss": 0.0341,
      "step": 321
    },
    {
      "epoch": 4.173354735152488,
      "grad_norm": 0.1908600926399231,
      "learning_rate": 1.3750527392853519e-05,
      "loss": 0.0838,
      "step": 322
    },
    {
      "epoch": 4.186195826645265,
      "grad_norm": 0.22569157183170319,
      "learning_rate": 1.3327357466931811e-05,
      "loss": 0.0833,
      "step": 323
    },
    {
      "epoch": 4.199036918138042,
      "grad_norm": 0.20586968958377838,
      "learning_rate": 1.2910335933058714e-05,
      "loss": 0.0921,
      "step": 324
    },
    {
      "epoch": 4.211878009630818,
      "grad_norm": 0.22846491634845734,
      "learning_rate": 1.2499492373946919e-05,
      "loss": 0.0935,
      "step": 325
    },
    {
      "epoch": 4.224719101123595,
      "grad_norm": 0.2915958762168884,
      "learning_rate": 1.2094855934055361e-05,
      "loss": 0.0931,
      "step": 326
    },
    {
      "epoch": 4.237560192616373,
      "grad_norm": 0.22078849375247955,
      "learning_rate": 1.1696455317521826e-05,
      "loss": 0.0873,
      "step": 327
    },
    {
      "epoch": 4.25040128410915,
      "grad_norm": 0.2387520968914032,
      "learning_rate": 1.1304318786126588e-05,
      "loss": 0.0819,
      "step": 328
    },
    {
      "epoch": 4.263242375601926,
      "grad_norm": 0.23125675320625305,
      "learning_rate": 1.0918474157287728e-05,
      "loss": 0.0713,
      "step": 329
    },
    {
      "epoch": 4.276083467094703,
      "grad_norm": 0.25590938329696655,
      "learning_rate": 1.0538948802087722e-05,
      "loss": 0.0734,
      "step": 330
    },
    {
      "epoch": 4.28892455858748,
      "grad_norm": 0.23935070633888245,
      "learning_rate": 1.0165769643331836e-05,
      "loss": 0.0677,
      "step": 331
    },
    {
      "epoch": 4.301765650080257,
      "grad_norm": 0.18608956038951874,
      "learning_rate": 9.79896315363823e-06,
      "loss": 0.0541,
      "step": 332
    },
    {
      "epoch": 4.314606741573034,
      "grad_norm": 0.3140433132648468,
      "learning_rate": 9.438555353560107e-06,
      "loss": 0.0586,
      "step": 333
    },
    {
      "epoch": 4.32744783306581,
      "grad_norm": 0.31043872237205505,
      "learning_rate": 9.084571809739806e-06,
      "loss": 0.0726,
      "step": 334
    },
    {
      "epoch": 4.340288924558587,
      "grad_norm": 0.19324089586734772,
      "learning_rate": 8.737037633095168e-06,
      "loss": 0.082,
      "step": 335
    },
    {
      "epoch": 4.353130016051364,
      "grad_norm": 0.2725972831249237,
      "learning_rate": 8.395977477038198e-06,
      "loss": 0.1139,
      "step": 336
    },
    {
      "epoch": 4.365971107544142,
      "grad_norm": 0.16910400986671448,
      "learning_rate": 8.061415535726303e-06,
      "loss": 0.0728,
      "step": 337
    },
    {
      "epoch": 4.378812199036918,
      "grad_norm": 0.22665587067604065,
      "learning_rate": 7.733375542345832e-06,
      "loss": 0.096,
      "step": 338
    },
    {
      "epoch": 4.391653290529695,
      "grad_norm": 0.26520606875419617,
      "learning_rate": 7.411880767428636e-06,
      "loss": 0.0902,
      "step": 339
    },
    {
      "epoch": 4.404494382022472,
      "grad_norm": 0.19831718504428864,
      "learning_rate": 7.096954017201207e-06,
      "loss": 0.0816,
      "step": 340
    },
    {
      "epoch": 4.417335473515249,
      "grad_norm": 0.20937596261501312,
      "learning_rate": 6.788617631966909e-06,
      "loss": 0.065,
      "step": 341
    },
    {
      "epoch": 4.430176565008026,
      "grad_norm": 0.284106582403183,
      "learning_rate": 6.486893484521162e-06,
      "loss": 0.0653,
      "step": 342
    },
    {
      "epoch": 4.443017656500802,
      "grad_norm": 0.23955251276493073,
      "learning_rate": 6.19180297859987e-06,
      "loss": 0.0483,
      "step": 343
    },
    {
      "epoch": 4.455858747993579,
      "grad_norm": 0.2952927052974701,
      "learning_rate": 5.9033670473610085e-06,
      "loss": 0.0719,
      "step": 344
    },
    {
      "epoch": 4.468699839486356,
      "grad_norm": 0.2724378705024719,
      "learning_rate": 5.621606151899716e-06,
      "loss": 0.0544,
      "step": 345
    },
    {
      "epoch": 4.481540930979134,
      "grad_norm": 0.17572782933712006,
      "learning_rate": 5.346540279796886e-06,
      "loss": 0.0446,
      "step": 346
    },
    {
      "epoch": 4.49438202247191,
      "grad_norm": 0.22200757265090942,
      "learning_rate": 5.078188943701123e-06,
      "loss": 0.1018,
      "step": 347
    },
    {
      "epoch": 4.507223113964687,
      "grad_norm": 0.2331586480140686,
      "learning_rate": 4.816571179944706e-06,
      "loss": 0.0991,
      "step": 348
    },
    {
      "epoch": 4.520064205457464,
      "grad_norm": 0.18275317549705505,
      "learning_rate": 4.561705547193096e-06,
      "loss": 0.0768,
      "step": 349
    },
    {
      "epoch": 4.532905296950241,
      "grad_norm": 0.1845731884241104,
      "learning_rate": 4.313610125128464e-06,
      "loss": 0.0922,
      "step": 350
    },
    {
      "epoch": 4.545746388443018,
      "grad_norm": 0.2531459331512451,
      "learning_rate": 4.072302513167148e-06,
      "loss": 0.0731,
      "step": 351
    },
    {
      "epoch": 4.558587479935794,
      "grad_norm": 0.17735856771469116,
      "learning_rate": 3.837799829211164e-06,
      "loss": 0.0664,
      "step": 352
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.1611146628856659,
      "learning_rate": 3.610118708433907e-06,
      "loss": 0.0681,
      "step": 353
    },
    {
      "epoch": 4.584269662921348,
      "grad_norm": 0.2285057157278061,
      "learning_rate": 3.389275302100081e-06,
      "loss": 0.0685,
      "step": 354
    },
    {
      "epoch": 4.597110754414125,
      "grad_norm": 0.21249911189079285,
      "learning_rate": 3.1752852764199812e-06,
      "loss": 0.0642,
      "step": 355
    },
    {
      "epoch": 4.609951845906902,
      "grad_norm": 0.27474352717399597,
      "learning_rate": 2.9681638114381184e-06,
      "loss": 0.0629,
      "step": 356
    },
    {
      "epoch": 4.622792937399679,
      "grad_norm": 0.23277732729911804,
      "learning_rate": 2.767925599956411e-06,
      "loss": 0.0533,
      "step": 357
    },
    {
      "epoch": 4.635634028892456,
      "grad_norm": 0.34809577465057373,
      "learning_rate": 2.57458484649189e-06,
      "loss": 0.0653,
      "step": 358
    },
    {
      "epoch": 4.648475120385233,
      "grad_norm": 0.19076129794120789,
      "learning_rate": 2.3881552662690765e-06,
      "loss": 0.0871,
      "step": 359
    },
    {
      "epoch": 4.66131621187801,
      "grad_norm": 0.15909138321876526,
      "learning_rate": 2.2086500842470214e-06,
      "loss": 0.0811,
      "step": 360
    },
    {
      "epoch": 4.674157303370786,
      "grad_norm": 0.1952279508113861,
      "learning_rate": 2.0360820341811635e-06,
      "loss": 0.0826,
      "step": 361
    },
    {
      "epoch": 4.686998394863563,
      "grad_norm": 0.14884842932224274,
      "learning_rate": 1.8704633577200204e-06,
      "loss": 0.0582,
      "step": 362
    },
    {
      "epoch": 4.69983948635634,
      "grad_norm": 0.2566874027252197,
      "learning_rate": 1.7118058035368368e-06,
      "loss": 0.0891,
      "step": 363
    },
    {
      "epoch": 4.712680577849117,
      "grad_norm": 0.23755532503128052,
      "learning_rate": 1.5601206264960644e-06,
      "loss": 0.0904,
      "step": 364
    },
    {
      "epoch": 4.725521669341894,
      "grad_norm": 0.2856099009513855,
      "learning_rate": 1.4154185868550106e-06,
      "loss": 0.0754,
      "step": 365
    },
    {
      "epoch": 4.738362760834671,
      "grad_norm": 0.24481110274791718,
      "learning_rate": 1.2777099495005495e-06,
      "loss": 0.0777,
      "step": 366
    },
    {
      "epoch": 4.751203852327448,
      "grad_norm": 0.2835271656513214,
      "learning_rate": 1.147004483220926e-06,
      "loss": 0.0688,
      "step": 367
    },
    {
      "epoch": 4.764044943820225,
      "grad_norm": 0.26856350898742676,
      "learning_rate": 1.0233114600127547e-06,
      "loss": 0.0639,
      "step": 368
    },
    {
      "epoch": 4.776886035313002,
      "grad_norm": 0.2596482038497925,
      "learning_rate": 9.066396544233468e-07,
      "loss": 0.0557,
      "step": 369
    },
    {
      "epoch": 4.789727126805778,
      "grad_norm": 0.25499358773231506,
      "learning_rate": 7.969973429281741e-07,
      "loss": 0.0536,
      "step": 370
    },
    {
      "epoch": 4.802568218298555,
      "grad_norm": 0.1461191475391388,
      "learning_rate": 6.94392303343816e-07,
      "loss": 0.0412,
      "step": 371
    },
    {
      "epoch": 4.815409309791332,
      "grad_norm": 0.32968461513519287,
      "learning_rate": 5.988318142762128e-07,
      "loss": 0.1382,
      "step": 372
    },
    {
      "epoch": 4.828250401284109,
      "grad_norm": 0.19888003170490265,
      "learning_rate": 5.103226546043005e-07,
      "loss": 0.0932,
      "step": 373
    },
    {
      "epoch": 4.841091492776886,
      "grad_norm": 0.22171039879322052,
      "learning_rate": 4.28871102999151e-07,
      "loss": 0.0748,
      "step": 374
    },
    {
      "epoch": 4.853932584269663,
      "grad_norm": 0.18117375671863556,
      "learning_rate": 3.54482937478573e-07,
      "loss": 0.0805,
      "step": 375
    },
    {
      "epoch": 4.86677367576244,
      "grad_norm": 0.18704640865325928,
      "learning_rate": 2.871634349972285e-07,
      "loss": 0.0959,
      "step": 376
    },
    {
      "epoch": 4.879614767255217,
      "grad_norm": 0.2359827756881714,
      "learning_rate": 2.2691737107231048e-07,
      "loss": 0.0779,
      "step": 377
    },
    {
      "epoch": 4.892455858747994,
      "grad_norm": 0.19910109043121338,
      "learning_rate": 1.7374901944474663e-07,
      "loss": 0.0737,
      "step": 378
    },
    {
      "epoch": 4.90529695024077,
      "grad_norm": 0.2202347368001938,
      "learning_rate": 1.2766215177605345e-07,
      "loss": 0.0775,
      "step": 379
    },
    {
      "epoch": 4.918138041733547,
      "grad_norm": 0.2741416096687317,
      "learning_rate": 8.866003738078332e-08,
      "loss": 0.0816,
      "step": 380
    },
    {
      "epoch": 4.930979133226324,
      "grad_norm": 0.19433164596557617,
      "learning_rate": 5.674544299457685e-08,
      "loss": 0.0703,
      "step": 381
    },
    {
      "epoch": 4.943820224719101,
      "grad_norm": 0.24093866348266602,
      "learning_rate": 3.192063257793088e-08,
      "loss": 0.0588,
      "step": 382
    },
    {
      "epoch": 4.956661316211878,
      "grad_norm": 0.3715647757053375,
      "learning_rate": 1.4187367155582555e-08,
      "loss": 0.0546,
      "step": 383
    },
    {
      "epoch": 4.969502407704655,
      "grad_norm": 0.26011672616004944,
      "learning_rate": 3.546904691564823e-09,
      "loss": 0.0893,
      "step": 384
    },
    {
      "epoch": 4.982343499197432,
      "grad_norm": 0.19426091015338898,
      "learning_rate": 0.0,
      "loss": 0.0736,
      "step": 385
    }
  ],
  "logging_steps": 1,
  "max_steps": 385,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2073448843153244e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}