{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 474,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004232804232804233,
      "grad_norm": 43.72717354138711,
      "learning_rate": 0.0,
      "loss": 1.6554,
      "step": 1
    },
    {
      "epoch": 0.008465608465608466,
      "grad_norm": 45.20956747041741,
      "learning_rate": 2.0833333333333333e-07,
      "loss": 1.6613,
      "step": 2
    },
    {
      "epoch": 0.012698412698412698,
      "grad_norm": 43.537183963199226,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 1.5809,
      "step": 3
    },
    {
      "epoch": 0.016931216931216932,
      "grad_norm": 44.16788560508001,
      "learning_rate": 6.25e-07,
      "loss": 1.5904,
      "step": 4
    },
    {
      "epoch": 0.021164021164021163,
      "grad_norm": 46.192802343401844,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.7731,
      "step": 5
    },
    {
      "epoch": 0.025396825396825397,
      "grad_norm": 41.283893260891254,
      "learning_rate": 1.0416666666666667e-06,
      "loss": 1.3879,
      "step": 6
    },
    {
      "epoch": 0.02962962962962963,
      "grad_norm": 41.031222170070116,
      "learning_rate": 1.25e-06,
      "loss": 1.379,
      "step": 7
    },
    {
      "epoch": 0.033862433862433865,
      "grad_norm": 36.224662378650926,
      "learning_rate": 1.4583333333333335e-06,
      "loss": 1.2278,
      "step": 8
    },
    {
      "epoch": 0.0380952380952381,
      "grad_norm": 34.90626982037512,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.1593,
      "step": 9
    },
    {
      "epoch": 0.042328042328042326,
      "grad_norm": 33.784429793622,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 1.0654,
      "step": 10
    },
    {
      "epoch": 0.04656084656084656,
      "grad_norm": 15.61619058454679,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 0.6672,
      "step": 11
    },
    {
      "epoch": 0.050793650793650794,
      "grad_norm": 16.519814951576077,
      "learning_rate": 2.2916666666666666e-06,
      "loss": 0.7885,
      "step": 12
    },
    {
      "epoch": 0.05502645502645503,
      "grad_norm": 13.171010801124316,
      "learning_rate": 2.5e-06,
      "loss": 0.71,
      "step": 13
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 10.014105616857696,
      "learning_rate": 2.7083333333333334e-06,
      "loss": 0.5411,
      "step": 14
    },
    {
      "epoch": 0.06349206349206349,
      "grad_norm": 5.737900509202217,
      "learning_rate": 2.916666666666667e-06,
      "loss": 0.4377,
      "step": 15
    },
    {
      "epoch": 0.06772486772486773,
      "grad_norm": 5.771623449687533,
      "learning_rate": 3.125e-06,
      "loss": 0.5075,
      "step": 16
    },
    {
      "epoch": 0.07195767195767196,
      "grad_norm": 5.175193297939713,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.4612,
      "step": 17
    },
    {
      "epoch": 0.0761904761904762,
      "grad_norm": 4.172409316471575,
      "learning_rate": 3.5416666666666673e-06,
      "loss": 0.4509,
      "step": 18
    },
    {
      "epoch": 0.08042328042328042,
      "grad_norm": 3.0944559263366096,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.4225,
      "step": 19
    },
    {
      "epoch": 0.08465608465608465,
      "grad_norm": 2.9291701549743854,
      "learning_rate": 3.958333333333333e-06,
      "loss": 0.4555,
      "step": 20
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 4.3100971907175305,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.3314,
      "step": 21
    },
    {
      "epoch": 0.09312169312169312,
      "grad_norm": 3.283513340198027,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 0.3266,
      "step": 22
    },
    {
      "epoch": 0.09735449735449736,
      "grad_norm": 2.581455776581018,
      "learning_rate": 4.583333333333333e-06,
      "loss": 0.3687,
      "step": 23
    },
    {
      "epoch": 0.10158730158730159,
      "grad_norm": 2.9154599797832703,
      "learning_rate": 4.791666666666668e-06,
      "loss": 0.3703,
      "step": 24
    },
    {
      "epoch": 0.10582010582010581,
      "grad_norm": 3.1934063229246834,
      "learning_rate": 5e-06,
      "loss": 0.3491,
      "step": 25
    },
    {
      "epoch": 0.11005291005291006,
      "grad_norm": 2.959988508176276,
      "learning_rate": 5.208333333333334e-06,
      "loss": 0.3126,
      "step": 26
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 2.507296053358847,
      "learning_rate": 5.416666666666667e-06,
      "loss": 0.3653,
      "step": 27
    },
    {
      "epoch": 0.11851851851851852,
      "grad_norm": 2.1595052302076323,
      "learning_rate": 5.625e-06,
      "loss": 0.2977,
      "step": 28
    },
    {
      "epoch": 0.12275132275132275,
      "grad_norm": 2.55785433661827,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.3175,
      "step": 29
    },
    {
      "epoch": 0.12698412698412698,
      "grad_norm": 2.461287213627725,
      "learning_rate": 6.041666666666667e-06,
      "loss": 0.2476,
      "step": 30
    },
    {
      "epoch": 0.1312169312169312,
      "grad_norm": 2.7466343812327136,
      "learning_rate": 6.25e-06,
      "loss": 0.2898,
      "step": 31
    },
    {
      "epoch": 0.13544973544973546,
      "grad_norm": 2.6506565239032276,
      "learning_rate": 6.458333333333334e-06,
      "loss": 0.2668,
      "step": 32
    },
    {
      "epoch": 0.13968253968253969,
      "grad_norm": 2.445226241184337,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2737,
      "step": 33
    },
    {
      "epoch": 0.1439153439153439,
      "grad_norm": 1.7323266471999959,
      "learning_rate": 6.875e-06,
      "loss": 0.2488,
      "step": 34
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 2.8002485372037227,
      "learning_rate": 7.083333333333335e-06,
      "loss": 0.2553,
      "step": 35
    },
    {
      "epoch": 0.1523809523809524,
      "grad_norm": 2.5749122065394436,
      "learning_rate": 7.291666666666667e-06,
      "loss": 0.1856,
      "step": 36
    },
    {
      "epoch": 0.15661375661375662,
      "grad_norm": 2.0107150492116785,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.2574,
      "step": 37
    },
    {
      "epoch": 0.16084656084656085,
      "grad_norm": 2.07586369622188,
      "learning_rate": 7.708333333333334e-06,
      "loss": 0.2342,
      "step": 38
    },
    {
      "epoch": 0.16507936507936508,
      "grad_norm": 2.650702366970764,
      "learning_rate": 7.916666666666667e-06,
      "loss": 0.2513,
      "step": 39
    },
    {
      "epoch": 0.1693121693121693,
      "grad_norm": 2.4643245144932457,
      "learning_rate": 8.125000000000001e-06,
      "loss": 0.2118,
      "step": 40
    },
    {
      "epoch": 0.17354497354497356,
      "grad_norm": 2.2583893699241404,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.2598,
      "step": 41
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 1.9649988674548964,
      "learning_rate": 8.541666666666666e-06,
      "loss": 0.2121,
      "step": 42
    },
    {
      "epoch": 0.182010582010582,
      "grad_norm": 2.2512170588798477,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.1935,
      "step": 43
    },
    {
      "epoch": 0.18624338624338624,
      "grad_norm": 2.428750936867218,
      "learning_rate": 8.958333333333334e-06,
      "loss": 0.2479,
      "step": 44
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 2.170136828036097,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.2397,
      "step": 45
    },
    {
      "epoch": 0.19470899470899472,
      "grad_norm": 2.0585076637804103,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.2467,
      "step": 46
    },
    {
      "epoch": 0.19894179894179895,
      "grad_norm": 3.5115386946858522,
      "learning_rate": 9.583333333333335e-06,
      "loss": 0.1458,
      "step": 47
    },
    {
      "epoch": 0.20317460317460317,
      "grad_norm": 2.4060311560165157,
      "learning_rate": 9.791666666666666e-06,
      "loss": 0.2102,
      "step": 48
    },
    {
      "epoch": 0.2074074074074074,
      "grad_norm": 1.8201349561390552,
      "learning_rate": 1e-05,
      "loss": 0.1991,
      "step": 49
    },
    {
      "epoch": 0.21164021164021163,
      "grad_norm": 1.60595437087767,
      "learning_rate": 9.999969538288953e-06,
      "loss": 0.1734,
      "step": 50
    },
    {
      "epoch": 0.21587301587301588,
      "grad_norm": 1.6521270217539161,
      "learning_rate": 9.999878153526974e-06,
      "loss": 0.1832,
      "step": 51
    },
    {
      "epoch": 0.2201058201058201,
      "grad_norm": 2.3590338162527713,
      "learning_rate": 9.999725846827562e-06,
      "loss": 0.2099,
      "step": 52
    },
    {
      "epoch": 0.22433862433862434,
      "grad_norm": 1.9526628599541012,
      "learning_rate": 9.999512620046523e-06,
      "loss": 0.2356,
      "step": 53
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 1.7003027985463,
      "learning_rate": 9.999238475781957e-06,
      "loss": 0.1916,
      "step": 54
    },
    {
      "epoch": 0.2328042328042328,
      "grad_norm": 2.197253252430344,
      "learning_rate": 9.998903417374228e-06,
      "loss": 0.17,
      "step": 55
    },
    {
      "epoch": 0.23703703703703705,
      "grad_norm": 2.392152390417205,
      "learning_rate": 9.998507448905917e-06,
      "loss": 0.1976,
      "step": 56
    },
    {
      "epoch": 0.24126984126984127,
      "grad_norm": 1.6988210931642431,
      "learning_rate": 9.998050575201772e-06,
      "loss": 0.1848,
      "step": 57
    },
    {
      "epoch": 0.2455026455026455,
      "grad_norm": 1.4811689327728754,
      "learning_rate": 9.997532801828659e-06,
      "loss": 0.1685,
      "step": 58
    },
    {
      "epoch": 0.24973544973544973,
      "grad_norm": 2.3716195293830635,
      "learning_rate": 9.99695413509548e-06,
      "loss": 0.2242,
      "step": 59
    },
    {
      "epoch": 0.25396825396825395,
      "grad_norm": 2.078219190553916,
      "learning_rate": 9.996314582053106e-06,
      "loss": 0.1641,
      "step": 60
    },
    {
      "epoch": 0.2582010582010582,
      "grad_norm": 2.945016339798954,
      "learning_rate": 9.995614150494293e-06,
      "loss": 0.1991,
      "step": 61
    },
    {
      "epoch": 0.2624338624338624,
      "grad_norm": 1.7177524445891639,
      "learning_rate": 9.994852848953574e-06,
      "loss": 0.1764,
      "step": 62
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 1.4250330727718503,
      "learning_rate": 9.994030686707171e-06,
      "loss": 0.143,
      "step": 63
    },
    {
      "epoch": 0.2708994708994709,
      "grad_norm": 1.3236123948431169,
      "learning_rate": 9.993147673772869e-06,
      "loss": 0.1352,
      "step": 64
    },
    {
      "epoch": 0.2751322751322751,
      "grad_norm": 1.4471704995242132,
      "learning_rate": 9.992203820909906e-06,
      "loss": 0.1687,
      "step": 65
    },
    {
      "epoch": 0.27936507936507937,
      "grad_norm": 1.898245103472612,
      "learning_rate": 9.991199139618828e-06,
      "loss": 0.1909,
      "step": 66
    },
    {
      "epoch": 0.28359788359788357,
      "grad_norm": 1.4090770167093434,
      "learning_rate": 9.990133642141359e-06,
      "loss": 0.1742,
      "step": 67
    },
    {
      "epoch": 0.2878306878306878,
      "grad_norm": 1.7072856786986559,
      "learning_rate": 9.989007341460251e-06,
      "loss": 0.1714,
      "step": 68
    },
    {
      "epoch": 0.2920634920634921,
      "grad_norm": 1.811382538600758,
      "learning_rate": 9.987820251299121e-06,
      "loss": 0.1739,
      "step": 69
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 1.3512054929719852,
      "learning_rate": 9.98657238612229e-06,
      "loss": 0.1712,
      "step": 70
    },
    {
      "epoch": 0.30052910052910053,
      "grad_norm": 1.3933604620853732,
      "learning_rate": 9.985263761134602e-06,
      "loss": 0.159,
      "step": 71
    },
    {
      "epoch": 0.3047619047619048,
      "grad_norm": 7.4550791659417275,
      "learning_rate": 9.983894392281237e-06,
      "loss": 0.1697,
      "step": 72
    },
    {
      "epoch": 0.308994708994709,
      "grad_norm": 1.7482631074271633,
      "learning_rate": 9.982464296247523e-06,
      "loss": 0.2084,
      "step": 73
    },
    {
      "epoch": 0.31322751322751324,
      "grad_norm": 1.4279579591947398,
      "learning_rate": 9.980973490458728e-06,
      "loss": 0.1146,
      "step": 74
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 1.752202038891908,
      "learning_rate": 9.979421993079853e-06,
      "loss": 0.2068,
      "step": 75
    },
    {
      "epoch": 0.3216931216931217,
      "grad_norm": 1.3196365121929348,
      "learning_rate": 9.9778098230154e-06,
      "loss": 0.1529,
      "step": 76
    },
    {
      "epoch": 0.32592592592592595,
      "grad_norm": 1.4588919246797958,
      "learning_rate": 9.976136999909156e-06,
      "loss": 0.1447,
      "step": 77
    },
    {
      "epoch": 0.33015873015873015,
      "grad_norm": 1.534245227167607,
      "learning_rate": 9.974403544143942e-06,
      "loss": 0.1407,
      "step": 78
    },
    {
      "epoch": 0.3343915343915344,
      "grad_norm": 1.2573367552977501,
      "learning_rate": 9.972609476841368e-06,
      "loss": 0.128,
      "step": 79
    },
    {
      "epoch": 0.3386243386243386,
      "grad_norm": 1.4699487437745995,
      "learning_rate": 9.970754819861577e-06,
      "loss": 0.1399,
      "step": 80
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 1.9752095010628823,
      "learning_rate": 9.968839595802982e-06,
      "loss": 0.1996,
      "step": 81
    },
    {
      "epoch": 0.3470899470899471,
      "grad_norm": 1.36801818550152,
      "learning_rate": 9.966863828001982e-06,
      "loss": 0.1524,
      "step": 82
    },
    {
      "epoch": 0.3513227513227513,
      "grad_norm": 1.4453966569682166,
      "learning_rate": 9.964827540532685e-06,
      "loss": 0.1635,
      "step": 83
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 1.3826856785129076,
      "learning_rate": 9.962730758206612e-06,
      "loss": 0.1363,
      "step": 84
    },
    {
      "epoch": 0.35978835978835977,
      "grad_norm": 1.8672803480166524,
      "learning_rate": 9.960573506572391e-06,
      "loss": 0.1987,
      "step": 85
    },
    {
      "epoch": 0.364021164021164,
      "grad_norm": 1.6365447793184111,
      "learning_rate": 9.958355811915452e-06,
      "loss": 0.1642,
      "step": 86
    },
    {
      "epoch": 0.3682539682539683,
      "grad_norm": 1.7338201367733481,
      "learning_rate": 9.95607770125771e-06,
      "loss": 0.1709,
      "step": 87
    },
    {
      "epoch": 0.3724867724867725,
      "grad_norm": 1.299464880095663,
      "learning_rate": 9.953739202357219e-06,
      "loss": 0.1323,
      "step": 88
    },
    {
      "epoch": 0.37671957671957673,
      "grad_norm": 1.5426139181892822,
      "learning_rate": 9.951340343707852e-06,
      "loss": 0.15,
      "step": 89
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 1.885701142257975,
      "learning_rate": 9.948881154538946e-06,
      "loss": 0.1644,
      "step": 90
    },
    {
      "epoch": 0.3851851851851852,
      "grad_norm": 1.4591100615878896,
      "learning_rate": 9.946361664814942e-06,
      "loss": 0.1537,
      "step": 91
    },
    {
      "epoch": 0.38941798941798944,
      "grad_norm": 1.1799285767193204,
      "learning_rate": 9.94378190523503e-06,
      "loss": 0.1154,
      "step": 92
    },
    {
      "epoch": 0.39365079365079364,
      "grad_norm": 1.216634736766946,
      "learning_rate": 9.941141907232766e-06,
      "loss": 0.1199,
      "step": 93
    },
    {
      "epoch": 0.3978835978835979,
      "grad_norm": 2.829211639773649,
      "learning_rate": 9.938441702975689e-06,
      "loss": 0.1745,
      "step": 94
    },
    {
      "epoch": 0.4021164021164021,
      "grad_norm": 1.284371191947075,
      "learning_rate": 9.93568132536494e-06,
      "loss": 0.1221,
      "step": 95
    },
    {
      "epoch": 0.40634920634920635,
      "grad_norm": 1.5092414625197825,
      "learning_rate": 9.932860808034847e-06,
      "loss": 0.1551,
      "step": 96
    },
    {
      "epoch": 0.4105820105820106,
      "grad_norm": 1.428323617465579,
      "learning_rate": 9.929980185352525e-06,
      "loss": 0.1233,
      "step": 97
    },
    {
      "epoch": 0.4148148148148148,
      "grad_norm": 1.4247177035890213,
      "learning_rate": 9.927039492417452e-06,
      "loss": 0.1487,
      "step": 98
    },
    {
      "epoch": 0.41904761904761906,
      "grad_norm": 1.4478986915245,
      "learning_rate": 9.924038765061042e-06,
      "loss": 0.1372,
      "step": 99
    },
    {
      "epoch": 0.42328042328042326,
      "grad_norm": 1.700062917316063,
      "learning_rate": 9.92097803984621e-06,
      "loss": 0.1476,
      "step": 100
    },
    {
      "epoch": 0.4275132275132275,
      "grad_norm": 1.4235599018999308,
      "learning_rate": 9.91785735406693e-06,
      "loss": 0.1348,
      "step": 101
    },
    {
      "epoch": 0.43174603174603177,
      "grad_norm": 1.5944688415988961,
      "learning_rate": 9.914676745747772e-06,
      "loss": 0.1321,
      "step": 102
    },
    {
      "epoch": 0.43597883597883597,
      "grad_norm": 1.7126018443966151,
      "learning_rate": 9.911436253643445e-06,
      "loss": 0.1617,
      "step": 103
    },
    {
      "epoch": 0.4402116402116402,
      "grad_norm": 1.5173736456674343,
      "learning_rate": 9.908135917238321e-06,
      "loss": 0.1555,
      "step": 104
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 1.643979121514149,
      "learning_rate": 9.904775776745959e-06,
      "loss": 0.1357,
      "step": 105
    },
    {
      "epoch": 0.4486772486772487,
      "grad_norm": 1.6123058130098344,
      "learning_rate": 9.901355873108611e-06,
      "loss": 0.1611,
      "step": 106
    },
    {
      "epoch": 0.45291005291005293,
      "grad_norm": 1.36764926534103,
      "learning_rate": 9.89787624799672e-06,
      "loss": 0.1369,
      "step": 107
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 1.5312605310786775,
      "learning_rate": 9.894336943808426e-06,
      "loss": 0.1613,
      "step": 108
    },
    {
      "epoch": 0.4613756613756614,
      "grad_norm": 1.48194042128105,
      "learning_rate": 9.890738003669029e-06,
      "loss": 0.1364,
      "step": 109
    },
    {
      "epoch": 0.4656084656084656,
      "grad_norm": 2.3838744962487963,
      "learning_rate": 9.887079471430481e-06,
      "loss": 0.1469,
      "step": 110
    },
    {
      "epoch": 0.46984126984126984,
      "grad_norm": 1.865768936599346,
      "learning_rate": 9.883361391670841e-06,
      "loss": 0.0879,
      "step": 111
    },
    {
      "epoch": 0.4740740740740741,
      "grad_norm": 1.9344837800965173,
      "learning_rate": 9.879583809693737e-06,
      "loss": 0.1396,
      "step": 112
    },
    {
      "epoch": 0.4783068783068783,
      "grad_norm": 1.4564213566749669,
      "learning_rate": 9.875746771527817e-06,
      "loss": 0.1655,
      "step": 113
    },
    {
      "epoch": 0.48253968253968255,
      "grad_norm": 1.416412960445032,
      "learning_rate": 9.871850323926178e-06,
      "loss": 0.1601,
      "step": 114
    },
    {
      "epoch": 0.48677248677248675,
      "grad_norm": 1.2470195823746264,
      "learning_rate": 9.867894514365802e-06,
      "loss": 0.1124,
      "step": 115
    },
    {
      "epoch": 0.491005291005291,
      "grad_norm": 1.6737155479454644,
      "learning_rate": 9.863879391046985e-06,
      "loss": 0.1204,
      "step": 116
    },
    {
      "epoch": 0.49523809523809526,
      "grad_norm": 1.267448596747368,
      "learning_rate": 9.859805002892733e-06,
      "loss": 0.1067,
      "step": 117
    },
    {
      "epoch": 0.49947089947089945,
      "grad_norm": 1.35223310268365,
      "learning_rate": 9.85567139954818e-06,
      "loss": 0.1329,
      "step": 118
    },
    {
      "epoch": 0.5037037037037037,
      "grad_norm": 1.3410113455783514,
      "learning_rate": 9.851478631379982e-06,
      "loss": 0.1324,
      "step": 119
    },
    {
      "epoch": 0.5079365079365079,
      "grad_norm": 1.5555421463772552,
      "learning_rate": 9.847226749475696e-06,
      "loss": 0.1075,
      "step": 120
    },
    {
      "epoch": 0.5121693121693122,
      "grad_norm": 1.425042236100241,
      "learning_rate": 9.842915805643156e-06,
      "loss": 0.1314,
      "step": 121
    },
    {
      "epoch": 0.5164021164021164,
      "grad_norm": 2.1003863770949844,
      "learning_rate": 9.838545852409857e-06,
      "loss": 0.166,
      "step": 122
    },
    {
      "epoch": 0.5206349206349207,
      "grad_norm": 1.0792059058292323,
      "learning_rate": 9.834116943022299e-06,
      "loss": 0.0989,
      "step": 123
    },
    {
      "epoch": 0.5248677248677248,
      "grad_norm": 1.5340606066613167,
      "learning_rate": 9.829629131445342e-06,
      "loss": 0.1125,
      "step": 124
    },
    {
      "epoch": 0.5291005291005291,
      "grad_norm": 1.8203980636061152,
      "learning_rate": 9.825082472361558e-06,
      "loss": 0.1393,
      "step": 125
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 1.8653391457662798,
      "learning_rate": 9.82047702117055e-06,
      "loss": 0.1781,
      "step": 126
    },
    {
      "epoch": 0.5375661375661376,
      "grad_norm": 1.868686455952389,
      "learning_rate": 9.815812833988292e-06,
      "loss": 0.1511,
      "step": 127
    },
    {
      "epoch": 0.5417989417989418,
      "grad_norm": 3.4516742714903956,
      "learning_rate": 9.811089967646427e-06,
      "loss": 0.182,
      "step": 128
    },
    {
      "epoch": 0.546031746031746,
      "grad_norm": 1.302285639983537,
      "learning_rate": 9.806308479691595e-06,
      "loss": 0.1362,
      "step": 129
    },
    {
      "epoch": 0.5502645502645502,
      "grad_norm": 1.3554292892825641,
      "learning_rate": 9.801468428384716e-06,
      "loss": 0.1183,
      "step": 130
    },
    {
      "epoch": 0.5544973544973545,
      "grad_norm": 1.5569558889730544,
      "learning_rate": 9.796569872700287e-06,
      "loss": 0.115,
      "step": 131
    },
    {
      "epoch": 0.5587301587301587,
      "grad_norm": 1.4128905574376962,
      "learning_rate": 9.791612872325667e-06,
      "loss": 0.1424,
      "step": 132
    },
    {
      "epoch": 0.562962962962963,
      "grad_norm": 1.5703621512026078,
      "learning_rate": 9.786597487660336e-06,
      "loss": 0.1426,
      "step": 133
    },
    {
      "epoch": 0.5671957671957671,
      "grad_norm": 1.1134755961640572,
      "learning_rate": 9.781523779815178e-06,
      "loss": 0.1148,
      "step": 134
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 1.6279717460336087,
      "learning_rate": 9.776391810611719e-06,
      "loss": 0.1432,
      "step": 135
    },
    {
      "epoch": 0.5756613756613757,
      "grad_norm": 1.1710680368668092,
      "learning_rate": 9.771201642581384e-06,
      "loss": 0.1112,
      "step": 136
    },
    {
      "epoch": 0.5798941798941799,
      "grad_norm": 1.468212722664273,
      "learning_rate": 9.765953338964736e-06,
      "loss": 0.1467,
      "step": 137
    },
    {
      "epoch": 0.5841269841269842,
      "grad_norm": 1.452581142800487,
      "learning_rate": 9.760646963710694e-06,
      "loss": 0.1686,
      "step": 138
    },
    {
      "epoch": 0.5883597883597883,
      "grad_norm": 1.000120927347695,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.0845,
      "step": 139
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 1.3458988878420348,
      "learning_rate": 9.749860257623262e-06,
      "loss": 0.1351,
      "step": 140
    },
    {
      "epoch": 0.5968253968253968,
      "grad_norm": 1.2838054499563394,
      "learning_rate": 9.744380058222483e-06,
      "loss": 0.1296,
      "step": 141
    },
    {
      "epoch": 0.6010582010582011,
      "grad_norm": 1.302987625581627,
      "learning_rate": 9.73884205004793e-06,
      "loss": 0.1163,
      "step": 142
    },
    {
      "epoch": 0.6052910052910053,
      "grad_norm": 1.3357457319423431,
      "learning_rate": 9.733246300578482e-06,
      "loss": 0.1345,
      "step": 143
    },
    {
      "epoch": 0.6095238095238096,
      "grad_norm": 1.530634255987009,
      "learning_rate": 9.727592877996585e-06,
      "loss": 0.1248,
      "step": 144
    },
    {
      "epoch": 0.6137566137566137,
      "grad_norm": 1.55621140705952,
      "learning_rate": 9.721881851187406e-06,
      "loss": 0.155,
      "step": 145
    },
    {
      "epoch": 0.617989417989418,
      "grad_norm": 1.5076666049322645,
      "learning_rate": 9.716113289738005e-06,
      "loss": 0.1262,
      "step": 146
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 3.424577642100365,
      "learning_rate": 9.710287263936485e-06,
      "loss": 0.1012,
      "step": 147
    },
    {
      "epoch": 0.6264550264550265,
      "grad_norm": 1.467957894333247,
      "learning_rate": 9.704403844771128e-06,
      "loss": 0.119,
      "step": 148
    },
    {
      "epoch": 0.6306878306878307,
      "grad_norm": 1.9287811168902935,
      "learning_rate": 9.698463103929542e-06,
      "loss": 0.1635,
      "step": 149
    },
    {
      "epoch": 0.6349206349206349,
      "grad_norm": 1.5672777610226705,
      "learning_rate": 9.69246511379778e-06,
      "loss": 0.1281,
      "step": 150
    },
    {
      "epoch": 0.6391534391534391,
      "grad_norm": 1.2674787426967342,
      "learning_rate": 9.68640994745946e-06,
      "loss": 0.1114,
      "step": 151
    },
    {
      "epoch": 0.6433862433862434,
      "grad_norm": 1.461753783787954,
      "learning_rate": 9.680297678694867e-06,
      "loss": 0.1439,
      "step": 152
    },
    {
      "epoch": 0.6476190476190476,
      "grad_norm": 1.5617305093987728,
      "learning_rate": 9.674128381980073e-06,
      "loss": 0.1308,
      "step": 153
    },
    {
      "epoch": 0.6518518518518519,
      "grad_norm": 1.5817050995954458,
      "learning_rate": 9.667902132486009e-06,
      "loss": 0.1244,
      "step": 154
    },
    {
      "epoch": 0.656084656084656,
      "grad_norm": 1.0297735518000437,
      "learning_rate": 9.661619006077562e-06,
      "loss": 0.0944,
      "step": 155
    },
    {
      "epoch": 0.6603174603174603,
      "grad_norm": 1.1490229757653174,
      "learning_rate": 9.655279079312643e-06,
      "loss": 0.1071,
      "step": 156
    },
    {
      "epoch": 0.6645502645502646,
      "grad_norm": 1.3820142441126761,
      "learning_rate": 9.648882429441258e-06,
      "loss": 0.1302,
      "step": 157
    },
    {
      "epoch": 0.6687830687830688,
      "grad_norm": 1.3066591757009012,
      "learning_rate": 9.642429134404568e-06,
      "loss": 0.1443,
      "step": 158
    },
    {
      "epoch": 0.6730158730158731,
      "grad_norm": 1.0248040135390748,
      "learning_rate": 9.635919272833938e-06,
      "loss": 0.0812,
      "step": 159
    },
    {
      "epoch": 0.6772486772486772,
      "grad_norm": 1.253548823921826,
      "learning_rate": 9.629352924049975e-06,
      "loss": 0.135,
      "step": 160
    },
    {
      "epoch": 0.6814814814814815,
      "grad_norm": 1.321779244061287,
      "learning_rate": 9.622730168061568e-06,
      "loss": 0.1236,
      "step": 161
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 1.0767529959934636,
      "learning_rate": 9.616051085564905e-06,
      "loss": 0.0846,
      "step": 162
    },
    {
      "epoch": 0.68994708994709,
      "grad_norm": 1.2506064320593104,
      "learning_rate": 9.609315757942504e-06,
      "loss": 0.1048,
      "step": 163
    },
    {
      "epoch": 0.6941798941798942,
      "grad_norm": 1.0490674529865702,
      "learning_rate": 9.602524267262202e-06,
      "loss": 0.1043,
      "step": 164
    },
    {
      "epoch": 0.6984126984126984,
      "grad_norm": 1.1274540124608232,
      "learning_rate": 9.595676696276173e-06,
      "loss": 0.1067,
      "step": 165
    },
    {
      "epoch": 0.7026455026455026,
      "grad_norm": 1.2376351263786731,
      "learning_rate": 9.588773128419907e-06,
      "loss": 0.1278,
      "step": 166
    },
    {
      "epoch": 0.7068783068783069,
      "grad_norm": 1.5158325756912805,
      "learning_rate": 9.581813647811199e-06,
      "loss": 0.1197,
      "step": 167
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 1.9368463333626136,
      "learning_rate": 9.574798339249124e-06,
      "loss": 0.108,
      "step": 168
    },
    {
      "epoch": 0.7153439153439154,
      "grad_norm": 1.3604192116787752,
      "learning_rate": 9.567727288213005e-06,
      "loss": 0.1131,
      "step": 169
    },
    {
      "epoch": 0.7195767195767195,
      "grad_norm": 2.3852673421250494,
      "learning_rate": 9.560600580861366e-06,
      "loss": 0.1146,
      "step": 170
    },
    {
      "epoch": 0.7238095238095238,
      "grad_norm": 1.3447438337348436,
      "learning_rate": 9.553418304030886e-06,
      "loss": 0.1114,
      "step": 171
    },
    {
      "epoch": 0.728042328042328,
      "grad_norm": 1.459045833803003,
      "learning_rate": 9.546180545235344e-06,
      "loss": 0.1467,
      "step": 172
    },
    {
      "epoch": 0.7322751322751323,
      "grad_norm": 1.6596975324831713,
      "learning_rate": 9.538887392664544e-06,
      "loss": 0.1069,
      "step": 173
    },
    {
      "epoch": 0.7365079365079366,
      "grad_norm": 1.3125689917057628,
      "learning_rate": 9.531538935183252e-06,
      "loss": 0.1155,
      "step": 174
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 1.1254939749629098,
      "learning_rate": 9.524135262330098e-06,
      "loss": 0.1014,
      "step": 175
    },
    {
      "epoch": 0.744973544973545,
      "grad_norm": 1.1850635605555897,
      "learning_rate": 9.516676464316505e-06,
      "loss": 0.0973,
      "step": 176
    },
    {
      "epoch": 0.7492063492063492,
      "grad_norm": 1.7947316330954919,
      "learning_rate": 9.50916263202557e-06,
      "loss": 0.1398,
      "step": 177
    },
    {
      "epoch": 0.7534391534391535,
      "grad_norm": 1.5534071768101145,
      "learning_rate": 9.501593857010968e-06,
      "loss": 0.1124,
      "step": 178
    },
    {
      "epoch": 0.7576719576719577,
      "grad_norm": 1.333104024444634,
      "learning_rate": 9.493970231495836e-06,
      "loss": 0.1172,
      "step": 179
    },
    {
      "epoch": 0.7619047619047619,
      "grad_norm": 1.4850034964893315,
      "learning_rate": 9.486291848371642e-06,
      "loss": 0.0939,
      "step": 180
    },
    {
      "epoch": 0.7661375661375661,
      "grad_norm": 1.657420935625152,
      "learning_rate": 9.478558801197065e-06,
      "loss": 0.0969,
      "step": 181
    },
    {
      "epoch": 0.7703703703703704,
      "grad_norm": 1.2670523544038959,
      "learning_rate": 9.470771184196842e-06,
      "loss": 0.0999,
      "step": 182
    },
    {
      "epoch": 0.7746031746031746,
      "grad_norm": 1.3320429466871224,
      "learning_rate": 9.46292909226063e-06,
      "loss": 0.148,
      "step": 183
    },
    {
      "epoch": 0.7788359788359789,
      "grad_norm": 1.2935520644229133,
      "learning_rate": 9.45503262094184e-06,
      "loss": 0.1137,
      "step": 184
    },
    {
      "epoch": 0.783068783068783,
      "grad_norm": 1.2311993159949326,
      "learning_rate": 9.44708186645649e-06,
      "loss": 0.1219,
      "step": 185
    },
    {
      "epoch": 0.7873015873015873,
      "grad_norm": 1.324888001044985,
      "learning_rate": 9.439076925682006e-06,
      "loss": 0.1016,
      "step": 186
    },
    {
      "epoch": 0.7915343915343915,
      "grad_norm": 0.9753462251818058,
      "learning_rate": 9.431017896156074e-06,
      "loss": 0.0939,
      "step": 187
    },
    {
      "epoch": 0.7957671957671958,
      "grad_norm": 1.162015032545452,
      "learning_rate": 9.42290487607542e-06,
      "loss": 0.1008,
      "step": 188
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.339272642650438,
      "learning_rate": 9.414737964294636e-06,
      "loss": 0.1047,
      "step": 189
    },
    {
      "epoch": 0.8042328042328042,
      "grad_norm": 1.7752016458700282,
      "learning_rate": 9.406517260324962e-06,
      "loss": 0.0836,
      "step": 190
    },
    {
      "epoch": 0.8084656084656084,
      "grad_norm": 1.2132508053585902,
      "learning_rate": 9.398242864333084e-06,
      "loss": 0.0893,
      "step": 191
    },
    {
      "epoch": 0.8126984126984127,
      "grad_norm": 1.0549315304944495,
      "learning_rate": 9.389914877139903e-06,
      "loss": 0.0958,
      "step": 192
    },
    {
      "epoch": 0.816931216931217,
      "grad_norm": 1.5784101564965254,
      "learning_rate": 9.381533400219319e-06,
      "loss": 0.0825,
      "step": 193
    },
    {
      "epoch": 0.8211640211640212,
      "grad_norm": 1.2660572838595667,
      "learning_rate": 9.37309853569698e-06,
      "loss": 0.1021,
      "step": 194
    },
    {
      "epoch": 0.8253968253968254,
      "grad_norm": 1.3321730339276188,
      "learning_rate": 9.364610386349048e-06,
      "loss": 0.0993,
      "step": 195
    },
    {
      "epoch": 0.8296296296296296,
      "grad_norm": 1.2635879093213929,
      "learning_rate": 9.356069055600949e-06,
      "loss": 0.0983,
      "step": 196
    },
    {
      "epoch": 0.8338624338624339,
      "grad_norm": 1.4546715685072167,
      "learning_rate": 9.347474647526095e-06,
      "loss": 0.0928,
      "step": 197
    },
    {
      "epoch": 0.8380952380952381,
      "grad_norm": 1.02322670489793,
      "learning_rate": 9.338827266844643e-06,
      "loss": 0.0772,
      "step": 198
    },
    {
      "epoch": 0.8423280423280424,
      "grad_norm": 1.3325182305302485,
      "learning_rate": 9.330127018922195e-06,
      "loss": 0.0987,
      "step": 199
    },
    {
      "epoch": 0.8465608465608465,
      "grad_norm": 1.403989810243998,
      "learning_rate": 9.321374009768525e-06,
      "loss": 0.0958,
      "step": 200
    },
    {
      "epoch": 0.8507936507936508,
      "grad_norm": 1.2529973130678513,
      "learning_rate": 9.312568346036288e-06,
      "loss": 0.0922,
      "step": 201
    },
    {
      "epoch": 0.855026455026455,
      "grad_norm": 1.339942832189008,
      "learning_rate": 9.30371013501972e-06,
      "loss": 0.0847,
      "step": 202
    },
    {
      "epoch": 0.8592592592592593,
      "grad_norm": 1.9827932325299167,
      "learning_rate": 9.294799484653323e-06,
      "loss": 0.0941,
      "step": 203
    },
    {
      "epoch": 0.8634920634920635,
      "grad_norm": 1.1213092752857448,
      "learning_rate": 9.285836503510562e-06,
      "loss": 0.0784,
      "step": 204
    },
    {
      "epoch": 0.8677248677248677,
      "grad_norm": 1.022053544019045,
      "learning_rate": 9.276821300802535e-06,
      "loss": 0.0826,
      "step": 205
    },
    {
      "epoch": 0.8719576719576719,
      "grad_norm": 1.2437584729757036,
      "learning_rate": 9.267753986376638e-06,
      "loss": 0.0988,
      "step": 206
    },
    {
      "epoch": 0.8761904761904762,
      "grad_norm": 1.4060299126574083,
      "learning_rate": 9.25863467071524e-06,
      "loss": 0.0963,
      "step": 207
    },
    {
      "epoch": 0.8804232804232804,
      "grad_norm": 1.1657287674213732,
      "learning_rate": 9.24946346493432e-06,
      "loss": 0.0855,
      "step": 208
    },
    {
      "epoch": 0.8846560846560847,
      "grad_norm": 1.1071987425032126,
      "learning_rate": 9.24024048078213e-06,
      "loss": 0.0828,
      "step": 209
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 1.376476091762823,
      "learning_rate": 9.230965830637821e-06,
      "loss": 0.1064,
      "step": 210
    },
    {
      "epoch": 0.8931216931216931,
      "grad_norm": 1.3361508869710212,
      "learning_rate": 9.221639627510076e-06,
      "loss": 0.1173,
      "step": 211
    },
    {
      "epoch": 0.8973544973544973,
      "grad_norm": 1.3580696473911005,
      "learning_rate": 9.21226198503574e-06,
      "loss": 0.0949,
      "step": 212
    },
    {
      "epoch": 0.9015873015873016,
      "grad_norm": 1.2783463166289852,
      "learning_rate": 9.202833017478421e-06,
      "loss": 0.0786,
      "step": 213
    },
    {
      "epoch": 0.9058201058201059,
      "grad_norm": 1.1862774845425663,
      "learning_rate": 9.193352839727122e-06,
      "loss": 0.0752,
      "step": 214
    },
    {
      "epoch": 0.91005291005291,
      "grad_norm": 1.0330987153646436,
      "learning_rate": 9.18382156729481e-06,
      "loss": 0.0731,
      "step": 215
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 1.2227259879389292,
      "learning_rate": 9.174239316317034e-06,
      "loss": 0.0852,
      "step": 216
    },
    {
      "epoch": 0.9185185185185185,
      "grad_norm": 1.338767605192061,
      "learning_rate": 9.164606203550498e-06,
      "loss": 0.1114,
      "step": 217
    },
    {
      "epoch": 0.9227513227513228,
      "grad_norm": 1.1914966704194456,
      "learning_rate": 9.154922346371641e-06,
      "loss": 0.0792,
      "step": 218
    },
    {
      "epoch": 0.926984126984127,
      "grad_norm": 1.2077161006608368,
      "learning_rate": 9.145187862775208e-06,
      "loss": 0.0937,
      "step": 219
    },
    {
      "epoch": 0.9312169312169312,
      "grad_norm": 3.455070672821183,
      "learning_rate": 9.13540287137281e-06,
      "loss": 0.1029,
      "step": 220
    },
    {
      "epoch": 0.9354497354497354,
      "grad_norm": 1.3029805107614403,
      "learning_rate": 9.125567491391476e-06,
      "loss": 0.0868,
      "step": 221
    },
    {
      "epoch": 0.9396825396825397,
      "grad_norm": 1.1230441398539233,
      "learning_rate": 9.115681842672211e-06,
      "loss": 0.0792,
      "step": 222
    },
    {
      "epoch": 0.9439153439153439,
      "grad_norm": 1.0274770873560906,
      "learning_rate": 9.10574604566852e-06,
      "loss": 0.0828,
      "step": 223
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 1.1591065147913875,
      "learning_rate": 9.09576022144496e-06,
      "loss": 0.0868,
      "step": 224
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 1.3488128549102674,
      "learning_rate": 9.085724491675642e-06,
      "loss": 0.106,
      "step": 225
    },
    {
      "epoch": 0.9566137566137566,
      "grad_norm": 1.2276156196511196,
      "learning_rate": 9.07563897864277e-06,
      "loss": 0.1007,
      "step": 226
    },
    {
      "epoch": 0.9608465608465608,
      "grad_norm": 1.0373897808991692,
      "learning_rate": 9.065503805235139e-06,
      "loss": 0.0757,
      "step": 227
    },
    {
      "epoch": 0.9650793650793651,
      "grad_norm": 1.5642701263049024,
      "learning_rate": 9.055319094946633e-06,
      "loss": 0.1113,
      "step": 228
    },
    {
      "epoch": 0.9693121693121693,
      "grad_norm": 1.3821470998948688,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.1066,
      "step": 229
    },
    {
      "epoch": 0.9735449735449735,
      "grad_norm": 1.2849131411511094,
      "learning_rate": 9.03480156071901e-06,
      "loss": 0.0996,
      "step": 230
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 1.4937027707955488,
      "learning_rate": 9.02446898677957e-06,
      "loss": 0.1066,
      "step": 231
    },
    {
      "epoch": 0.982010582010582,
      "grad_norm": 0.9718589878629968,
      "learning_rate": 9.014087375955574e-06,
      "loss": 0.0695,
      "step": 232
    },
    {
      "epoch": 0.9862433862433863,
      "grad_norm": 1.0615875981137657,
      "learning_rate": 9.003656854743667e-06,
      "loss": 0.0741,
      "step": 233
    },
    {
      "epoch": 0.9904761904761905,
      "grad_norm": 1.0154034408078385,
      "learning_rate": 8.993177550236464e-06,
      "loss": 0.085,
      "step": 234
    },
    {
      "epoch": 0.9947089947089947,
      "grad_norm": 1.0590350628316874,
      "learning_rate": 8.982649590120982e-06,
      "loss": 0.0666,
      "step": 235
    },
    {
      "epoch": 0.9989417989417989,
      "grad_norm": 1.009502950887791,
      "learning_rate": 8.972073102677091e-06,
      "loss": 0.0667,
      "step": 236
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.5207179670384026,
      "learning_rate": 8.961448216775955e-06,
      "loss": 0.0719,
      "step": 237
    },
    {
      "epoch": 1.0042328042328041,
      "grad_norm": 1.0439827565118902,
      "learning_rate": 8.950775061878453e-06,
      "loss": 0.0617,
      "step": 238
    },
    {
      "epoch": 1.0084656084656085,
      "grad_norm": 1.0127555558104633,
      "learning_rate": 8.94005376803361e-06,
      "loss": 0.0604,
      "step": 239
    },
    {
      "epoch": 1.0126984126984127,
      "grad_norm": 1.0868468474919644,
      "learning_rate": 8.92928446587701e-06,
      "loss": 0.0593,
      "step": 240
    },
    {
      "epoch": 1.016931216931217,
      "grad_norm": 1.1138192394576714,
      "learning_rate": 8.9184672866292e-06,
      "loss": 0.066,
      "step": 241
    },
    {
      "epoch": 1.0211640211640212,
      "grad_norm": 1.0568629136248378,
      "learning_rate": 8.907602362094094e-06,
      "loss": 0.0521,
      "step": 242
    },
    {
      "epoch": 1.0253968253968253,
      "grad_norm": 1.3202632756778692,
      "learning_rate": 8.896689824657371e-06,
      "loss": 0.0737,
      "step": 243
    },
    {
      "epoch": 1.0296296296296297,
      "grad_norm": 1.094241539250586,
      "learning_rate": 8.885729807284855e-06,
      "loss": 0.0577,
      "step": 244
    },
    {
      "epoch": 1.0338624338624338,
      "grad_norm": 0.7772743020562026,
      "learning_rate": 8.874722443520898e-06,
      "loss": 0.0414,
      "step": 245
    },
    {
      "epoch": 1.0380952380952382,
      "grad_norm": 1.0784304127588566,
      "learning_rate": 8.863667867486756e-06,
      "loss": 0.0601,
      "step": 246
    },
    {
      "epoch": 1.0423280423280423,
      "grad_norm": 1.0115870407215455,
      "learning_rate": 8.852566213878947e-06,
      "loss": 0.0672,
      "step": 247
    },
    {
      "epoch": 1.0465608465608465,
      "grad_norm": 0.9393680447714831,
      "learning_rate": 8.841417617967618e-06,
      "loss": 0.0562,
      "step": 248
    },
    {
      "epoch": 1.0507936507936508,
      "grad_norm": 1.000103639091894,
      "learning_rate": 8.83022221559489e-06,
      "loss": 0.0505,
      "step": 249
    },
    {
      "epoch": 1.055026455026455,
      "grad_norm": 0.8530141568102377,
      "learning_rate": 8.818980143173212e-06,
      "loss": 0.0445,
      "step": 250
    },
    {
      "epoch": 1.0592592592592593,
      "grad_norm": 0.9092503604322939,
      "learning_rate": 8.807691537683685e-06,
      "loss": 0.0377,
      "step": 251
    },
    {
      "epoch": 1.0634920634920635,
      "grad_norm": 1.3616532372700458,
      "learning_rate": 8.796356536674404e-06,
      "loss": 0.0447,
      "step": 252
    },
    {
      "epoch": 1.0677248677248676,
      "grad_norm": 0.9961095762490313,
      "learning_rate": 8.784975278258783e-06,
      "loss": 0.0422,
      "step": 253
    },
    {
      "epoch": 1.071957671957672,
      "grad_norm": 1.417792501888075,
      "learning_rate": 8.773547901113862e-06,
      "loss": 0.0685,
      "step": 254
    },
    {
      "epoch": 1.0761904761904761,
      "grad_norm": 1.109045645130657,
      "learning_rate": 8.762074544478622e-06,
      "loss": 0.057,
      "step": 255
    },
    {
      "epoch": 1.0804232804232805,
      "grad_norm": 1.1085366846759674,
      "learning_rate": 8.750555348152299e-06,
      "loss": 0.0618,
      "step": 256
    },
    {
      "epoch": 1.0846560846560847,
      "grad_norm": 0.9675368242876587,
      "learning_rate": 8.73899045249266e-06,
      "loss": 0.0698,
      "step": 257
    },
    {
      "epoch": 1.0888888888888888,
      "grad_norm": 0.8969933415528727,
      "learning_rate": 8.727379998414311e-06,
      "loss": 0.0499,
      "step": 258
    },
    {
      "epoch": 1.0931216931216932,
      "grad_norm": 0.9848664591981523,
      "learning_rate": 8.715724127386971e-06,
      "loss": 0.0664,
      "step": 259
    },
    {
      "epoch": 1.0973544973544973,
      "grad_norm": 0.9775428435297511,
      "learning_rate": 8.70402298143375e-06,
      "loss": 0.0467,
      "step": 260
    },
    {
      "epoch": 1.1015873015873017,
      "grad_norm": 0.8332721352250031,
      "learning_rate": 8.692276703129421e-06,
      "loss": 0.035,
      "step": 261
    },
    {
      "epoch": 1.1058201058201058,
      "grad_norm": 0.9472578567420649,
      "learning_rate": 8.680485435598674e-06,
      "loss": 0.0499,
      "step": 262
    },
    {
      "epoch": 1.11005291005291,
      "grad_norm": 0.8344760929082524,
      "learning_rate": 8.668649322514382e-06,
      "loss": 0.0473,
      "step": 263
    },
    {
      "epoch": 1.1142857142857143,
      "grad_norm": 1.0269303477988245,
      "learning_rate": 8.656768508095853e-06,
      "loss": 0.0453,
      "step": 264
    },
    {
      "epoch": 1.1185185185185185,
      "grad_norm": 1.1216585383229603,
      "learning_rate": 8.644843137107058e-06,
      "loss": 0.0445,
      "step": 265
    },
    {
      "epoch": 1.1227513227513228,
      "grad_norm": 1.1375626971477624,
      "learning_rate": 8.632873354854881e-06,
      "loss": 0.0609,
      "step": 266
    },
    {
      "epoch": 1.126984126984127,
      "grad_norm": 1.0980193157423086,
      "learning_rate": 8.620859307187339e-06,
      "loss": 0.0585,
      "step": 267
    },
    {
      "epoch": 1.1312169312169311,
      "grad_norm": 1.0923249482421242,
      "learning_rate": 8.608801140491811e-06,
      "loss": 0.0535,
      "step": 268
    },
    {
      "epoch": 1.1354497354497355,
      "grad_norm": 1.199920331072888,
      "learning_rate": 8.596699001693257e-06,
      "loss": 0.0574,
      "step": 269
    },
    {
      "epoch": 1.1396825396825396,
      "grad_norm": 1.3840691329312167,
      "learning_rate": 8.584553038252415e-06,
      "loss": 0.0603,
      "step": 270
    },
    {
      "epoch": 1.143915343915344,
      "grad_norm": 1.0976468243723367,
      "learning_rate": 8.572363398164017e-06,
      "loss": 0.0495,
      "step": 271
    },
    {
      "epoch": 1.1481481481481481,
      "grad_norm": 0.8361407917583945,
      "learning_rate": 8.560130229954985e-06,
      "loss": 0.0428,
      "step": 272
    },
    {
      "epoch": 1.1523809523809523,
      "grad_norm": 1.1399444462070143,
      "learning_rate": 8.547853682682605e-06,
      "loss": 0.0507,
      "step": 273
    },
    {
      "epoch": 1.1566137566137566,
      "grad_norm": 1.9349212107218514,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.0399,
      "step": 274
    },
    {
      "epoch": 1.1608465608465608,
      "grad_norm": 1.4185063496527812,
      "learning_rate": 8.523171049817974e-06,
      "loss": 0.0504,
      "step": 275
    },
    {
      "epoch": 1.1650793650793652,
      "grad_norm": 0.9431563961918448,
      "learning_rate": 8.510765264975813e-06,
      "loss": 0.045,
      "step": 276
    },
    {
      "epoch": 1.1693121693121693,
      "grad_norm": 0.9634549342361133,
      "learning_rate": 8.498316702566828e-06,
      "loss": 0.0433,
      "step": 277
    },
    {
      "epoch": 1.1735449735449737,
      "grad_norm": 0.8107013087955712,
      "learning_rate": 8.485825514272824e-06,
      "loss": 0.0396,
      "step": 278
    },
    {
      "epoch": 1.1777777777777778,
      "grad_norm": 1.2143724846797779,
      "learning_rate": 8.473291852294986e-06,
      "loss": 0.0503,
      "step": 279
    },
    {
      "epoch": 1.182010582010582,
      "grad_norm": 1.214634058606961,
      "learning_rate": 8.460715869352035e-06,
      "loss": 0.0492,
      "step": 280
    },
    {
      "epoch": 1.1862433862433863,
      "grad_norm": 1.5326647754225693,
      "learning_rate": 8.44809771867835e-06,
      "loss": 0.0697,
      "step": 281
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 1.0183387396190917,
      "learning_rate": 8.435437554022116e-06,
      "loss": 0.0506,
      "step": 282
    },
    {
      "epoch": 1.1947089947089946,
      "grad_norm": 0.9774623735237468,
      "learning_rate": 8.422735529643445e-06,
      "loss": 0.0478,
      "step": 283
    },
    {
      "epoch": 1.198941798941799,
      "grad_norm": 0.9151100643280784,
      "learning_rate": 8.409991800312493e-06,
      "loss": 0.0521,
      "step": 284
    },
    {
      "epoch": 1.2031746031746031,
      "grad_norm": 0.944795332484615,
      "learning_rate": 8.397206521307584e-06,
      "loss": 0.0462,
      "step": 285
    },
    {
      "epoch": 1.2074074074074075,
      "grad_norm": 0.8708767785679999,
      "learning_rate": 8.384379848413304e-06,
      "loss": 0.04,
      "step": 286
    },
    {
      "epoch": 1.2116402116402116,
      "grad_norm": 0.9060528905235375,
      "learning_rate": 8.371511937918616e-06,
      "loss": 0.0415,
      "step": 287
    },
    {
      "epoch": 1.215873015873016,
      "grad_norm": 0.9029101418348026,
      "learning_rate": 8.358602946614952e-06,
      "loss": 0.0457,
      "step": 288
    },
    {
      "epoch": 1.2201058201058201,
      "grad_norm": 1.1143501235536941,
      "learning_rate": 8.345653031794292e-06,
      "loss": 0.0492,
      "step": 289
    },
    {
      "epoch": 1.2243386243386243,
      "grad_norm": 0.6904563329955595,
      "learning_rate": 8.332662351247262e-06,
      "loss": 0.029,
      "step": 290
    },
    {
      "epoch": 1.2285714285714286,
      "grad_norm": 0.9806076212927931,
      "learning_rate": 8.319631063261209e-06,
      "loss": 0.0565,
      "step": 291
    },
    {
      "epoch": 1.2328042328042328,
      "grad_norm": 0.9637920964858679,
      "learning_rate": 8.30655932661826e-06,
      "loss": 0.0547,
      "step": 292
    },
    {
      "epoch": 1.237037037037037,
      "grad_norm": 0.9864706909349096,
      "learning_rate": 8.293447300593402e-06,
      "loss": 0.0485,
      "step": 293
    },
    {
      "epoch": 1.2412698412698413,
      "grad_norm": 1.0056521811983314,
      "learning_rate": 8.280295144952537e-06,
      "loss": 0.0455,
      "step": 294
    },
    {
      "epoch": 1.2455026455026454,
      "grad_norm": 2.7374704058965413,
      "learning_rate": 8.267103019950529e-06,
      "loss": 0.0402,
      "step": 295
    },
    {
      "epoch": 1.2497354497354498,
      "grad_norm": 1.0675471111168537,
      "learning_rate": 8.253871086329255e-06,
      "loss": 0.0444,
      "step": 296
    },
    {
      "epoch": 1.253968253968254,
      "grad_norm": 0.8741609150096519,
      "learning_rate": 8.240599505315656e-06,
      "loss": 0.034,
      "step": 297
    },
    {
      "epoch": 1.2582010582010583,
      "grad_norm": 1.0949757974574201,
      "learning_rate": 8.227288438619754e-06,
      "loss": 0.0413,
      "step": 298
    },
    {
      "epoch": 1.2624338624338625,
      "grad_norm": 1.3321265503517306,
      "learning_rate": 8.213938048432697e-06,
      "loss": 0.0432,
      "step": 299
    },
    {
      "epoch": 1.2666666666666666,
      "grad_norm": 1.16809688783282,
      "learning_rate": 8.200548497424779e-06,
      "loss": 0.0457,
      "step": 300
    },
    {
      "epoch": 1.270899470899471,
      "grad_norm": 1.2234242804842506,
      "learning_rate": 8.18711994874345e-06,
      "loss": 0.0351,
      "step": 301
    },
    {
      "epoch": 1.2751322751322751,
      "grad_norm": 0.8455313768916234,
      "learning_rate": 8.173652566011339e-06,
      "loss": 0.0305,
      "step": 302
    },
    {
      "epoch": 1.2793650793650793,
      "grad_norm": 1.264978471670017,
      "learning_rate": 8.160146513324256e-06,
      "loss": 0.0681,
      "step": 303
    },
    {
      "epoch": 1.2835978835978836,
      "grad_norm": 1.0191087783122954,
      "learning_rate": 8.146601955249187e-06,
      "loss": 0.0458,
      "step": 304
    },
    {
      "epoch": 1.2878306878306878,
      "grad_norm": 1.5364104175585032,
      "learning_rate": 8.133019056822303e-06,
      "loss": 0.0478,
      "step": 305
    },
    {
      "epoch": 1.2920634920634921,
      "grad_norm": 1.0186300363769811,
      "learning_rate": 8.119397983546932e-06,
      "loss": 0.0521,
      "step": 306
    },
    {
      "epoch": 1.2962962962962963,
      "grad_norm": 0.9695838159643356,
      "learning_rate": 8.105738901391553e-06,
      "loss": 0.043,
      "step": 307
    },
    {
      "epoch": 1.3005291005291006,
      "grad_norm": 1.1606845112374315,
      "learning_rate": 8.092041976787772e-06,
      "loss": 0.0391,
      "step": 308
    },
    {
      "epoch": 1.3047619047619048,
      "grad_norm": 0.9131906220785553,
      "learning_rate": 8.078307376628292e-06,
      "loss": 0.0503,
      "step": 309
    },
    {
      "epoch": 1.308994708994709,
      "grad_norm": 1.1213236223136132,
      "learning_rate": 8.064535268264883e-06,
      "loss": 0.0563,
      "step": 310
    },
    {
      "epoch": 1.3132275132275133,
      "grad_norm": 1.5919757523939648,
      "learning_rate": 8.05072581950634e-06,
      "loss": 0.0485,
      "step": 311
    },
    {
      "epoch": 1.3174603174603174,
      "grad_norm": 0.8068544658944983,
      "learning_rate": 8.036879198616434e-06,
      "loss": 0.036,
      "step": 312
    },
    {
      "epoch": 1.3216931216931216,
      "grad_norm": 1.32303739260173,
      "learning_rate": 8.022995574311876e-06,
      "loss": 0.0557,
      "step": 313
    },
    {
      "epoch": 1.325925925925926,
      "grad_norm": 1.6279390963252223,
      "learning_rate": 8.009075115760243e-06,
      "loss": 0.0553,
      "step": 314
    },
    {
      "epoch": 1.33015873015873,
      "grad_norm": 0.713797069897446,
      "learning_rate": 7.99511799257793e-06,
      "loss": 0.0397,
      "step": 315
    },
    {
      "epoch": 1.3343915343915345,
      "grad_norm": 0.8079576051295935,
      "learning_rate": 7.981124374828079e-06,
      "loss": 0.04,
      "step": 316
    },
    {
      "epoch": 1.3386243386243386,
      "grad_norm": 1.1435417675953259,
      "learning_rate": 7.967094433018508e-06,
      "loss": 0.0435,
      "step": 317
    },
    {
      "epoch": 1.342857142857143,
      "grad_norm": 1.143163112909474,
      "learning_rate": 7.953028338099628e-06,
      "loss": 0.0474,
      "step": 318
    },
    {
      "epoch": 1.3470899470899471,
      "grad_norm": 0.9269672035843429,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.0313,
      "step": 319
    },
    {
      "epoch": 1.3513227513227513,
      "grad_norm": 1.0466290884719476,
      "learning_rate": 7.92478837493608e-06,
      "loss": 0.0505,
      "step": 320
    },
    {
      "epoch": 1.3555555555555556,
      "grad_norm": 1.256639700693128,
      "learning_rate": 7.910614850786448e-06,
      "loss": 0.0438,
      "step": 321
    },
    {
      "epoch": 1.3597883597883598,
      "grad_norm": 0.802253728130833,
      "learning_rate": 7.896405861713393e-06,
      "loss": 0.0344,
      "step": 322
    },
    {
      "epoch": 1.364021164021164,
      "grad_norm": 1.0700812448503103,
      "learning_rate": 7.882161580848966e-06,
      "loss": 0.0516,
      "step": 323
    },
    {
      "epoch": 1.3682539682539683,
      "grad_norm": 1.0286894659033428,
      "learning_rate": 7.86788218175523e-06,
      "loss": 0.0485,
      "step": 324
    },
    {
      "epoch": 1.3724867724867724,
      "grad_norm": 1.032321094949915,
      "learning_rate": 7.85356783842216e-06,
      "loss": 0.0493,
      "step": 325
    },
    {
      "epoch": 1.3767195767195768,
      "grad_norm": 0.8860252372178579,
      "learning_rate": 7.839218725265507e-06,
      "loss": 0.0413,
      "step": 326
    },
    {
      "epoch": 1.380952380952381,
      "grad_norm": 0.9104230123199901,
      "learning_rate": 7.82483501712469e-06,
      "loss": 0.0484,
      "step": 327
    },
    {
      "epoch": 1.3851851851851853,
      "grad_norm": 0.9105868366799567,
      "learning_rate": 7.810416889260653e-06,
      "loss": 0.033,
      "step": 328
    },
    {
      "epoch": 1.3894179894179894,
      "grad_norm": 0.8874460617855023,
      "learning_rate": 7.795964517353734e-06,
      "loss": 0.0418,
      "step": 329
    },
    {
      "epoch": 1.3936507936507936,
      "grad_norm": 0.6979249575618531,
      "learning_rate": 7.781478077501526e-06,
      "loss": 0.0276,
      "step": 330
    },
    {
      "epoch": 1.397883597883598,
      "grad_norm": 1.1401863184171674,
      "learning_rate": 7.76695774621672e-06,
      "loss": 0.0487,
      "step": 331
    },
    {
      "epoch": 1.402116402116402,
      "grad_norm": 1.180237330609796,
      "learning_rate": 7.752403700424978e-06,
      "loss": 0.0481,
      "step": 332
    },
    {
      "epoch": 1.4063492063492062,
      "grad_norm": 0.971305148025189,
      "learning_rate": 7.737816117462752e-06,
      "loss": 0.0533,
      "step": 333
    },
    {
      "epoch": 1.4105820105820106,
      "grad_norm": 1.143064790256319,
      "learning_rate": 7.723195175075136e-06,
      "loss": 0.0481,
      "step": 334
    },
    {
      "epoch": 1.4148148148148147,
      "grad_norm": 1.3039942970009886,
      "learning_rate": 7.7085410514137e-06,
      "loss": 0.0464,
      "step": 335
    },
    {
      "epoch": 1.4190476190476191,
      "grad_norm": 0.7197106725379593,
      "learning_rate": 7.693853925034316e-06,
      "loss": 0.0294,
      "step": 336
    },
    {
      "epoch": 1.4232804232804233,
      "grad_norm": 0.7576529055995104,
      "learning_rate": 7.679133974894984e-06,
      "loss": 0.0363,
      "step": 337
    },
    {
      "epoch": 1.4275132275132276,
      "grad_norm": 1.0996527863122494,
      "learning_rate": 7.66438138035365e-06,
      "loss": 0.0511,
      "step": 338
    },
    {
      "epoch": 1.4317460317460318,
      "grad_norm": 1.2175807102371912,
      "learning_rate": 7.649596321166024e-06,
      "loss": 0.0521,
      "step": 339
    },
    {
      "epoch": 1.435978835978836,
      "grad_norm": 1.1677918243054364,
      "learning_rate": 7.634778977483389e-06,
      "loss": 0.053,
      "step": 340
    },
    {
      "epoch": 1.4402116402116403,
      "grad_norm": 1.3661180396603787,
      "learning_rate": 7.619929529850397e-06,
      "loss": 0.0541,
      "step": 341
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 1.134002971754773,
      "learning_rate": 7.605048159202884e-06,
      "loss": 0.0418,
      "step": 342
    },
    {
      "epoch": 1.4486772486772486,
      "grad_norm": 1.2740661399159647,
      "learning_rate": 7.590135046865652e-06,
      "loss": 0.0477,
      "step": 343
    },
    {
      "epoch": 1.452910052910053,
      "grad_norm": 0.9564673428495316,
      "learning_rate": 7.575190374550272e-06,
      "loss": 0.0424,
      "step": 344
    },
    {
      "epoch": 1.457142857142857,
      "grad_norm": 0.8592538896773881,
      "learning_rate": 7.560214324352858e-06,
      "loss": 0.0313,
      "step": 345
    },
    {
      "epoch": 1.4613756613756614,
      "grad_norm": 1.025295823546742,
      "learning_rate": 7.545207078751858e-06,
      "loss": 0.05,
      "step": 346
    },
    {
      "epoch": 1.4656084656084656,
      "grad_norm": 1.005206182718304,
      "learning_rate": 7.530168820605819e-06,
      "loss": 0.0366,
      "step": 347
    },
    {
      "epoch": 1.46984126984127,
      "grad_norm": 0.817943689379754,
      "learning_rate": 7.515099733151177e-06,
      "loss": 0.0348,
      "step": 348
    },
    {
      "epoch": 1.474074074074074,
      "grad_norm": 0.9487808741793253,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.0462,
      "step": 349
    },
    {
      "epoch": 1.4783068783068782,
      "grad_norm": 0.7940195414603822,
      "learning_rate": 7.484869805137778e-06,
      "loss": 0.0333,
      "step": 350
    },
    {
      "epoch": 1.4825396825396826,
      "grad_norm": 2.594857831175931,
      "learning_rate": 7.469709332921155e-06,
      "loss": 0.0377,
      "step": 351
    },
    {
      "epoch": 1.4867724867724867,
      "grad_norm": 0.6298189399907342,
      "learning_rate": 7.454518768075705e-06,
      "loss": 0.0298,
      "step": 352
    },
    {
      "epoch": 1.491005291005291,
      "grad_norm": 1.2746680413135476,
      "learning_rate": 7.4392982956936644e-06,
      "loss": 0.056,
      "step": 353
    },
    {
      "epoch": 1.4952380952380953,
      "grad_norm": 0.822040862171887,
      "learning_rate": 7.424048101231687e-06,
      "loss": 0.033,
      "step": 354
    },
    {
      "epoch": 1.4994708994708994,
      "grad_norm": 1.1242964464799274,
      "learning_rate": 7.408768370508577e-06,
      "loss": 0.0517,
      "step": 355
    },
    {
      "epoch": 1.5037037037037035,
      "grad_norm": 0.8701346706195384,
      "learning_rate": 7.393459289703035e-06,
      "loss": 0.0315,
      "step": 356
    },
    {
      "epoch": 1.507936507936508,
      "grad_norm": 0.9133640557604403,
      "learning_rate": 7.378121045351378e-06,
      "loss": 0.039,
      "step": 357
    },
    {
      "epoch": 1.5121693121693123,
      "grad_norm": 1.0021618381519257,
      "learning_rate": 7.362753824345271e-06,
      "loss": 0.0394,
      "step": 358
    },
    {
      "epoch": 1.5164021164021164,
      "grad_norm": 0.8996476991571326,
      "learning_rate": 7.347357813929455e-06,
      "loss": 0.0446,
      "step": 359
    },
    {
      "epoch": 1.5206349206349206,
      "grad_norm": 2.264884992718341,
      "learning_rate": 7.3319332016994575e-06,
      "loss": 0.0484,
      "step": 360
    },
    {
      "epoch": 1.524867724867725,
      "grad_norm": 1.066146344538063,
      "learning_rate": 7.31648017559931e-06,
      "loss": 0.0436,
      "step": 361
    },
    {
      "epoch": 1.529100529100529,
      "grad_norm": 0.9302808730777476,
      "learning_rate": 7.300998923919259e-06,
      "loss": 0.0328,
      "step": 362
    },
    {
      "epoch": 1.5333333333333332,
      "grad_norm": 0.9911847404144593,
      "learning_rate": 7.285489635293472e-06,
      "loss": 0.0423,
      "step": 363
    },
    {
      "epoch": 1.5375661375661376,
      "grad_norm": 1.0364471553258094,
      "learning_rate": 7.269952498697734e-06,
      "loss": 0.0395,
      "step": 364
    },
    {
      "epoch": 1.541798941798942,
      "grad_norm": 0.7662023272437689,
      "learning_rate": 7.254387703447154e-06,
      "loss": 0.0307,
      "step": 365
    },
    {
      "epoch": 1.5460317460317459,
      "grad_norm": 1.0377441871093285,
      "learning_rate": 7.238795439193849e-06,
      "loss": 0.0389,
      "step": 366
    },
    {
      "epoch": 1.5502645502645502,
      "grad_norm": 1.0484977562889242,
      "learning_rate": 7.223175895924638e-06,
      "loss": 0.0474,
      "step": 367
    },
    {
      "epoch": 1.5544973544973546,
      "grad_norm": 1.3320962267108385,
      "learning_rate": 7.207529263958727e-06,
      "loss": 0.0461,
      "step": 368
    },
    {
      "epoch": 1.5587301587301587,
      "grad_norm": 1.1981890101491062,
      "learning_rate": 7.191855733945388e-06,
      "loss": 0.046,
      "step": 369
    },
    {
      "epoch": 1.5629629629629629,
      "grad_norm": 1.0353677690689251,
      "learning_rate": 7.176155496861639e-06,
      "loss": 0.0467,
      "step": 370
    },
    {
      "epoch": 1.5671957671957673,
      "grad_norm": 0.9023734846152021,
      "learning_rate": 7.160428744009913e-06,
      "loss": 0.032,
      "step": 371
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 0.9795672938910132,
      "learning_rate": 7.1446756670157306e-06,
      "loss": 0.042,
      "step": 372
    },
    {
      "epoch": 1.5756613756613755,
      "grad_norm": 0.8949456456419879,
      "learning_rate": 7.128896457825364e-06,
      "loss": 0.034,
      "step": 373
    },
    {
      "epoch": 1.57989417989418,
      "grad_norm": 1.0979703990784655,
      "learning_rate": 7.113091308703498e-06,
      "loss": 0.0356,
      "step": 374
    },
    {
      "epoch": 1.5841269841269843,
      "grad_norm": 0.9108245034081557,
      "learning_rate": 7.0972604122308865e-06,
      "loss": 0.0378,
      "step": 375
    },
    {
      "epoch": 1.5883597883597882,
      "grad_norm": 0.9069830623011091,
      "learning_rate": 7.081403961302007e-06,
      "loss": 0.0382,
      "step": 376
    },
    {
      "epoch": 1.5925925925925926,
      "grad_norm": 0.9298727976375222,
      "learning_rate": 7.06552214912271e-06,
      "loss": 0.0381,
      "step": 377
    },
    {
      "epoch": 1.596825396825397,
      "grad_norm": 1.030708836778186,
      "learning_rate": 7.049615169207864e-06,
      "loss": 0.0369,
      "step": 378
    },
    {
      "epoch": 1.601058201058201,
      "grad_norm": 1.0309556409580762,
      "learning_rate": 7.033683215379002e-06,
      "loss": 0.0439,
      "step": 379
    },
    {
      "epoch": 1.6052910052910052,
      "grad_norm": 1.0407782701760113,
      "learning_rate": 7.0177264817619514e-06,
      "loss": 0.0421,
      "step": 380
    },
    {
      "epoch": 1.6095238095238096,
      "grad_norm": 1.3354904716977571,
      "learning_rate": 7.0017451627844765e-06,
      "loss": 0.0526,
      "step": 381
    },
    {
      "epoch": 1.6137566137566137,
      "grad_norm": 0.8800338194269767,
      "learning_rate": 6.985739453173903e-06,
      "loss": 0.0379,
      "step": 382
    },
    {
      "epoch": 1.6179894179894179,
      "grad_norm": 0.752228395119178,
      "learning_rate": 6.9697095479547564e-06,
      "loss": 0.0306,
      "step": 383
    },
    {
      "epoch": 1.6222222222222222,
      "grad_norm": 0.9413579645875557,
      "learning_rate": 6.953655642446368e-06,
      "loss": 0.0331,
      "step": 384
    },
    {
      "epoch": 1.6264550264550266,
      "grad_norm": 0.9570770027126146,
      "learning_rate": 6.9375779322605154e-06,
      "loss": 0.0387,
      "step": 385
    },
    {
      "epoch": 1.6306878306878307,
      "grad_norm": 0.8684490278106084,
      "learning_rate": 6.921476613299018e-06,
      "loss": 0.0355,
      "step": 386
    },
    {
      "epoch": 1.6349206349206349,
      "grad_norm": 1.0632953241977892,
      "learning_rate": 6.905351881751372e-06,
      "loss": 0.0454,
      "step": 387
    },
    {
      "epoch": 1.6391534391534393,
      "grad_norm": 1.15676003887316,
      "learning_rate": 6.889203934092337e-06,
      "loss": 0.0517,
      "step": 388
    },
    {
      "epoch": 1.6433862433862434,
      "grad_norm": 0.8832495191608842,
      "learning_rate": 6.873032967079562e-06,
      "loss": 0.0358,
      "step": 389
    },
    {
      "epoch": 1.6476190476190475,
      "grad_norm": 0.8891845587906834,
      "learning_rate": 6.856839177751175e-06,
      "loss": 0.0329,
      "step": 390
    },
    {
      "epoch": 1.651851851851852,
      "grad_norm": 0.833507703150214,
      "learning_rate": 6.840622763423391e-06,
      "loss": 0.0314,
      "step": 391
    },
    {
      "epoch": 1.656084656084656,
      "grad_norm": 0.8520667024026805,
      "learning_rate": 6.824383921688098e-06,
      "loss": 0.0336,
      "step": 392
    },
    {
      "epoch": 1.6603174603174602,
      "grad_norm": 0.9617042714964271,
      "learning_rate": 6.808122850410461e-06,
      "loss": 0.0317,
      "step": 393
    },
    {
      "epoch": 1.6645502645502646,
      "grad_norm": 0.9734363303715623,
      "learning_rate": 6.7918397477265e-06,
      "loss": 0.0483,
      "step": 394
    },
    {
      "epoch": 1.668783068783069,
      "grad_norm": 1.0180051756915056,
      "learning_rate": 6.775534812040686e-06,
      "loss": 0.0438,
      "step": 395
    },
    {
      "epoch": 1.673015873015873,
      "grad_norm": 0.9663036959934387,
      "learning_rate": 6.759208242023509e-06,
      "loss": 0.0411,
      "step": 396
    },
    {
      "epoch": 1.6772486772486772,
      "grad_norm": 0.6966551244304431,
      "learning_rate": 6.7428602366090764e-06,
      "loss": 0.0234,
      "step": 397
    },
    {
      "epoch": 1.6814814814814816,
      "grad_norm": 1.3759059743950008,
      "learning_rate": 6.7264909949926735e-06,
      "loss": 0.0291,
      "step": 398
    },
    {
      "epoch": 1.6857142857142857,
      "grad_norm": 0.7565996741340549,
      "learning_rate": 6.710100716628345e-06,
      "loss": 0.0224,
      "step": 399
    },
    {
      "epoch": 1.6899470899470899,
      "grad_norm": 1.1887100463923472,
      "learning_rate": 6.693689601226458e-06,
      "loss": 0.042,
      "step": 400
    },
    {
      "epoch": 1.6941798941798942,
      "grad_norm": 0.966051588155485,
      "learning_rate": 6.677257848751276e-06,
      "loss": 0.0378,
      "step": 401
    },
    {
      "epoch": 1.6984126984126984,
      "grad_norm": 0.8772916268098477,
      "learning_rate": 6.6608056594185166e-06,
      "loss": 0.0335,
      "step": 402
    },
    {
      "epoch": 1.7026455026455025,
      "grad_norm": 0.8372311631604785,
      "learning_rate": 6.644333233692917e-06,
      "loss": 0.0367,
      "step": 403
    },
    {
      "epoch": 1.7068783068783069,
      "grad_norm": 0.992879995777924,
      "learning_rate": 6.627840772285784e-06,
      "loss": 0.045,
      "step": 404
    },
    {
      "epoch": 1.7111111111111112,
      "grad_norm": 0.9366130180501334,
      "learning_rate": 6.611328476152557e-06,
      "loss": 0.0289,
      "step": 405
    },
    {
      "epoch": 1.7153439153439154,
      "grad_norm": 0.8915482740408233,
      "learning_rate": 6.594796546490351e-06,
      "loss": 0.0355,
      "step": 406
    },
    {
      "epoch": 1.7195767195767195,
      "grad_norm": 1.3583953457116698,
      "learning_rate": 6.578245184735513e-06,
      "loss": 0.0461,
      "step": 407
    },
    {
      "epoch": 1.723809523809524,
      "grad_norm": 0.8410154802690613,
      "learning_rate": 6.561674592561164e-06,
      "loss": 0.0352,
      "step": 408
    },
    {
      "epoch": 1.728042328042328,
      "grad_norm": 0.9907489907823347,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.037,
      "step": 409
    },
    {
      "epoch": 1.7322751322751322,
      "grad_norm": 0.7734556527152434,
      "learning_rate": 6.5284765248155295e-06,
      "loss": 0.0264,
      "step": 410
    },
    {
      "epoch": 1.7365079365079366,
      "grad_norm": 0.8517169774537486,
      "learning_rate": 6.5118494537522235e-06,
      "loss": 0.0336,
      "step": 411
    },
    {
      "epoch": 1.7407407407407407,
      "grad_norm": 1.0261903054202124,
      "learning_rate": 6.495203961280434e-06,
      "loss": 0.0435,
      "step": 412
    },
    {
      "epoch": 1.7449735449735448,
      "grad_norm": 0.8195140295058992,
      "learning_rate": 6.4785402502202345e-06,
      "loss": 0.0319,
      "step": 413
    },
    {
      "epoch": 1.7492063492063492,
      "grad_norm": 0.8064178126782999,
      "learning_rate": 6.461858523613684e-06,
      "loss": 0.0262,
      "step": 414
    },
    {
      "epoch": 1.7534391534391536,
      "grad_norm": 1.059308376312486,
      "learning_rate": 6.445158984722358e-06,
      "loss": 0.0323,
      "step": 415
    },
    {
      "epoch": 1.7576719576719577,
      "grad_norm": 0.7467600140287795,
      "learning_rate": 6.428441837024868e-06,
      "loss": 0.0274,
      "step": 416
    },
    {
      "epoch": 1.7619047619047619,
      "grad_norm": 0.8230132283566644,
      "learning_rate": 6.411707284214384e-06,
      "loss": 0.0236,
      "step": 417
    },
    {
      "epoch": 1.7661375661375662,
      "grad_norm": 0.787813130354089,
      "learning_rate": 6.3949555301961474e-06,
      "loss": 0.0355,
      "step": 418
    },
    {
      "epoch": 1.7703703703703704,
      "grad_norm": 1.451682463238513,
      "learning_rate": 6.378186779084996e-06,
      "loss": 0.0437,
      "step": 419
    },
    {
      "epoch": 1.7746031746031745,
      "grad_norm": 0.8747504739561074,
      "learning_rate": 6.361401235202872e-06,
      "loss": 0.0284,
      "step": 420
    },
    {
      "epoch": 1.7788359788359789,
      "grad_norm": 1.070054317826232,
      "learning_rate": 6.344599103076329e-06,
      "loss": 0.0491,
      "step": 421
    },
    {
      "epoch": 1.783068783068783,
      "grad_norm": 0.9839694781276292,
      "learning_rate": 6.327780587434045e-06,
      "loss": 0.0262,
      "step": 422
    },
    {
      "epoch": 1.7873015873015872,
      "grad_norm": 1.9302390005333394,
      "learning_rate": 6.310945893204324e-06,
      "loss": 0.0381,
      "step": 423
    },
    {
      "epoch": 1.7915343915343915,
      "grad_norm": 0.9588230652812939,
      "learning_rate": 6.294095225512604e-06,
      "loss": 0.0304,
      "step": 424
    },
    {
      "epoch": 1.795767195767196,
      "grad_norm": 0.947273510951518,
      "learning_rate": 6.277228789678953e-06,
      "loss": 0.0468,
      "step": 425
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.8800021402393514,
      "learning_rate": 6.26034679121557e-06,
      "loss": 0.0199,
      "step": 426
    },
    {
      "epoch": 1.8042328042328042,
      "grad_norm": 1.130364633548092,
      "learning_rate": 6.243449435824276e-06,
      "loss": 0.0319,
      "step": 427
    },
    {
      "epoch": 1.8084656084656086,
      "grad_norm": 1.1066604882510598,
      "learning_rate": 6.2265369293940135e-06,
      "loss": 0.0329,
      "step": 428
    },
    {
      "epoch": 1.8126984126984127,
      "grad_norm": 1.0550706667423015,
      "learning_rate": 6.209609477998339e-06,
      "loss": 0.0283,
      "step": 429
    },
    {
      "epoch": 1.8169312169312168,
      "grad_norm": 0.9811790647444768,
      "learning_rate": 6.192667287892905e-06,
      "loss": 0.0335,
      "step": 430
    },
    {
      "epoch": 1.8211640211640212,
      "grad_norm": 0.7827472464069767,
      "learning_rate": 6.17571056551295e-06,
      "loss": 0.0328,
      "step": 431
    },
    {
      "epoch": 1.8253968253968254,
      "grad_norm": 0.7344784176592902,
      "learning_rate": 6.158739517470786e-06,
      "loss": 0.0244,
      "step": 432
    },
    {
      "epoch": 1.8296296296296295,
      "grad_norm": 1.1298574892350588,
      "learning_rate": 6.141754350553279e-06,
      "loss": 0.041,
      "step": 433
    },
    {
      "epoch": 1.8338624338624339,
      "grad_norm": 0.748246425007082,
      "learning_rate": 6.124755271719326e-06,
      "loss": 0.0281,
      "step": 434
    },
    {
      "epoch": 1.8380952380952382,
      "grad_norm": 1.0435711016816667,
      "learning_rate": 6.107742488097338e-06,
      "loss": 0.0389,
      "step": 435
    },
    {
      "epoch": 1.8423280423280424,
      "grad_norm": 1.0041457638530626,
      "learning_rate": 6.090716206982714e-06,
      "loss": 0.0328,
      "step": 436
    },
    {
      "epoch": 1.8465608465608465,
      "grad_norm": 0.940959041938284,
      "learning_rate": 6.073676635835317e-06,
      "loss": 0.0274,
      "step": 437
    },
    {
      "epoch": 1.8507936507936509,
      "grad_norm": 0.995917260966997,
      "learning_rate": 6.056623982276945e-06,
      "loss": 0.0302,
      "step": 438
    },
    {
      "epoch": 1.855026455026455,
      "grad_norm": 0.6253437609792718,
      "learning_rate": 6.039558454088796e-06,
      "loss": 0.0275,
      "step": 439
    },
    {
      "epoch": 1.8592592592592592,
      "grad_norm": 0.7265389917244132,
      "learning_rate": 6.022480259208951e-06,
      "loss": 0.0227,
      "step": 440
    },
    {
      "epoch": 1.8634920634920635,
      "grad_norm": 1.0390428613384723,
      "learning_rate": 6.005389605729824e-06,
      "loss": 0.0372,
      "step": 441
    },
    {
      "epoch": 1.8677248677248677,
      "grad_norm": 0.8380178245455322,
      "learning_rate": 5.988286701895631e-06,
      "loss": 0.027,
      "step": 442
    },
    {
      "epoch": 1.8719576719576718,
      "grad_norm": 1.0320239189151428,
      "learning_rate": 5.97117175609986e-06,
      "loss": 0.0273,
      "step": 443
    },
    {
      "epoch": 1.8761904761904762,
      "grad_norm": 0.9065424646753207,
      "learning_rate": 5.954044976882725e-06,
      "loss": 0.0278,
      "step": 444
    },
    {
      "epoch": 1.8804232804232806,
      "grad_norm": 0.9178765588368991,
      "learning_rate": 5.936906572928625e-06,
      "loss": 0.0331,
      "step": 445
    },
    {
      "epoch": 1.8846560846560847,
      "grad_norm": 0.7960677399262612,
      "learning_rate": 5.919756753063601e-06,
      "loss": 0.0193,
      "step": 446
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 1.280013434964197,
      "learning_rate": 5.902595726252801e-06,
      "loss": 0.0411,
      "step": 447
    },
    {
      "epoch": 1.8931216931216932,
      "grad_norm": 1.1561219072627489,
      "learning_rate": 5.885423701597918e-06,
      "loss": 0.0243,
      "step": 448
    },
    {
      "epoch": 1.8973544973544973,
      "grad_norm": 1.0393388710304052,
      "learning_rate": 5.8682408883346535e-06,
      "loss": 0.0365,
      "step": 449
    },
    {
      "epoch": 1.9015873015873015,
      "grad_norm": 0.8819590478297261,
      "learning_rate": 5.851047495830163e-06,
      "loss": 0.0302,
      "step": 450
    },
    {
      "epoch": 1.9058201058201059,
      "grad_norm": 1.778211477880275,
      "learning_rate": 5.8338437335805124e-06,
      "loss": 0.0498,
      "step": 451
    },
    {
      "epoch": 1.91005291005291,
      "grad_norm": 0.6164685645460388,
      "learning_rate": 5.816629811208112e-06,
      "loss": 0.02,
      "step": 452
    },
    {
      "epoch": 1.9142857142857141,
      "grad_norm": 1.022308010686047,
      "learning_rate": 5.799405938459175e-06,
      "loss": 0.0388,
      "step": 453
    },
    {
      "epoch": 1.9185185185185185,
      "grad_norm": 0.8643023085266177,
      "learning_rate": 5.782172325201155e-06,
      "loss": 0.0233,
      "step": 454
    },
    {
      "epoch": 1.9227513227513229,
      "grad_norm": 1.0448789439073158,
      "learning_rate": 5.764929181420191e-06,
      "loss": 0.0366,
      "step": 455
    },
    {
      "epoch": 1.926984126984127,
      "grad_norm": 1.1136298371281064,
      "learning_rate": 5.747676717218549e-06,
      "loss": 0.0406,
      "step": 456
    },
    {
      "epoch": 1.9312169312169312,
      "grad_norm": 0.8054456826112729,
      "learning_rate": 5.730415142812059e-06,
      "loss": 0.0204,
      "step": 457
    },
    {
      "epoch": 1.9354497354497355,
      "grad_norm": 0.8704692982674614,
      "learning_rate": 5.7131446685275595e-06,
      "loss": 0.0228,
      "step": 458
    },
    {
      "epoch": 1.9396825396825397,
      "grad_norm": 0.7876406159977293,
      "learning_rate": 5.695865504800328e-06,
      "loss": 0.0272,
      "step": 459
    },
    {
      "epoch": 1.9439153439153438,
      "grad_norm": 0.9252100216116378,
      "learning_rate": 5.678577862171523e-06,
      "loss": 0.0237,
      "step": 460
    },
    {
      "epoch": 1.9481481481481482,
      "grad_norm": 0.6893854806674795,
      "learning_rate": 5.661281951285613e-06,
      "loss": 0.0206,
      "step": 461
    },
    {
      "epoch": 1.9523809523809523,
      "grad_norm": 0.7845843824207334,
      "learning_rate": 5.643977982887815e-06,
      "loss": 0.0312,
      "step": 462
    },
    {
      "epoch": 1.9566137566137565,
      "grad_norm": 0.8049990285785104,
      "learning_rate": 5.626666167821522e-06,
      "loss": 0.0292,
      "step": 463
    },
    {
      "epoch": 1.9608465608465608,
      "grad_norm": 1.148625600779929,
      "learning_rate": 5.609346717025738e-06,
      "loss": 0.0193,
      "step": 464
    },
    {
      "epoch": 1.9650793650793652,
      "grad_norm": 0.6413181984555196,
      "learning_rate": 5.592019841532507e-06,
      "loss": 0.0198,
      "step": 465
    },
    {
      "epoch": 1.9693121693121693,
      "grad_norm": 0.8580613271611244,
      "learning_rate": 5.5746857524643335e-06,
      "loss": 0.0219,
      "step": 466
    },
    {
      "epoch": 1.9735449735449735,
      "grad_norm": 0.9676515609676355,
      "learning_rate": 5.557344661031628e-06,
      "loss": 0.0326,
      "step": 467
    },
    {
      "epoch": 1.9777777777777779,
      "grad_norm": 0.9277086173806844,
      "learning_rate": 5.539996778530114e-06,
      "loss": 0.0331,
      "step": 468
    },
    {
      "epoch": 1.982010582010582,
      "grad_norm": 1.2919252511895343,
      "learning_rate": 5.522642316338268e-06,
      "loss": 0.0298,
      "step": 469
    },
    {
      "epoch": 1.9862433862433861,
      "grad_norm": 0.8640757560649422,
      "learning_rate": 5.505281485914732e-06,
      "loss": 0.0218,
      "step": 470
    },
    {
      "epoch": 1.9904761904761905,
      "grad_norm": 1.1249585693174393,
      "learning_rate": 5.487914498795748e-06,
      "loss": 0.0418,
      "step": 471
    },
    {
      "epoch": 1.9947089947089947,
      "grad_norm": 0.810724157613082,
      "learning_rate": 5.470541566592573e-06,
      "loss": 0.0294,
      "step": 472
    },
    {
      "epoch": 1.9989417989417988,
      "grad_norm": 0.8074300914872133,
      "learning_rate": 5.453162900988902e-06,
      "loss": 0.0221,
      "step": 473
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6149507180824806,
      "learning_rate": 5.435778713738292e-06,
      "loss": 0.0114,
      "step": 474
    }
  ],
  "logging_steps": 1,
  "max_steps": 948,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 120380300677120.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}