{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 216,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013888888888888888,
      "grad_norm": 32.20025306755104,
      "learning_rate": 0.0,
      "loss": 1.6489,
      "step": 1
    },
    {
      "epoch": 0.027777777777777776,
      "grad_norm": 31.757967894129138,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 1.5608,
      "step": 2
    },
    {
      "epoch": 0.041666666666666664,
      "grad_norm": 27.005592965716428,
      "learning_rate": 9.090909090909091e-07,
      "loss": 1.4318,
      "step": 3
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 25.876596085278493,
      "learning_rate": 1.3636363636363636e-06,
      "loss": 1.3459,
      "step": 4
    },
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 20.05325433673777,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 1.1116,
      "step": 5
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 17.2745705787659,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 1.0695,
      "step": 6
    },
    {
      "epoch": 0.09722222222222222,
      "grad_norm": 11.547654869877704,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 1.052,
      "step": 7
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 7.017914033099854,
      "learning_rate": 3.181818181818182e-06,
      "loss": 1.0261,
      "step": 8
    },
    {
      "epoch": 0.125,
      "grad_norm": 6.79190992209115,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.9326,
      "step": 9
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 6.623522507160447,
      "learning_rate": 4.0909090909090915e-06,
      "loss": 1.012,
      "step": 10
    },
    {
      "epoch": 0.1527777777777778,
      "grad_norm": 3.7533541005048554,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.809,
      "step": 11
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 3.843755827537282,
      "learning_rate": 5e-06,
      "loss": 0.8047,
      "step": 12
    },
    {
      "epoch": 0.18055555555555555,
      "grad_norm": 4.34940165800091,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.7493,
      "step": 13
    },
    {
      "epoch": 0.19444444444444445,
      "grad_norm": 2.6403288766736477,
      "learning_rate": 5.90909090909091e-06,
      "loss": 0.4881,
      "step": 14
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 2.1536362938927005,
      "learning_rate": 6.363636363636364e-06,
      "loss": 0.5604,
      "step": 15
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 2.943068061086033,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.6005,
      "step": 16
    },
    {
      "epoch": 0.2361111111111111,
      "grad_norm": 2.4040059583348214,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.4677,
      "step": 17
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.373878455942415,
      "learning_rate": 7.727272727272727e-06,
      "loss": 0.5283,
      "step": 18
    },
    {
      "epoch": 0.2638888888888889,
      "grad_norm": 2.5632160548134713,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.4489,
      "step": 19
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 2.5874585166377924,
      "learning_rate": 8.636363636363637e-06,
      "loss": 0.6012,
      "step": 20
    },
    {
      "epoch": 0.2916666666666667,
      "grad_norm": 2.347060535429832,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.4817,
      "step": 21
    },
    {
      "epoch": 0.3055555555555556,
      "grad_norm": 3.103950712452586,
      "learning_rate": 9.545454545454547e-06,
      "loss": 0.5688,
      "step": 22
    },
    {
      "epoch": 0.3194444444444444,
      "grad_norm": 2.7722433588291966,
      "learning_rate": 1e-05,
      "loss": 0.494,
      "step": 23
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 2.118018603780166,
      "learning_rate": 9.999344418328161e-06,
      "loss": 0.3882,
      "step": 24
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 2.227514369635361,
      "learning_rate": 9.997377845227577e-06,
      "loss": 0.451,
      "step": 25
    },
    {
      "epoch": 0.3611111111111111,
      "grad_norm": 1.8352056819401141,
      "learning_rate": 9.994100796397954e-06,
      "loss": 0.4151,
      "step": 26
    },
    {
      "epoch": 0.375,
      "grad_norm": 2.0303315554358488,
      "learning_rate": 9.98951413118856e-06,
      "loss": 0.383,
      "step": 27
    },
    {
      "epoch": 0.3888888888888889,
      "grad_norm": 1.9876736681198277,
      "learning_rate": 9.983619052372847e-06,
      "loss": 0.3647,
      "step": 28
    },
    {
      "epoch": 0.4027777777777778,
      "grad_norm": 1.7843457351766752,
      "learning_rate": 9.97641710583307e-06,
      "loss": 0.3629,
      "step": 29
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 1.4930651496160667,
      "learning_rate": 9.96791018015489e-06,
      "loss": 0.3443,
      "step": 30
    },
    {
      "epoch": 0.4305555555555556,
      "grad_norm": 1.5759590184712353,
      "learning_rate": 9.958100506132127e-06,
      "loss": 0.324,
      "step": 31
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 1.705745384345177,
      "learning_rate": 9.946990656181782e-06,
      "loss": 0.3095,
      "step": 32
    },
    {
      "epoch": 0.4583333333333333,
      "grad_norm": 2.0524444353062403,
      "learning_rate": 9.934583543669454e-06,
      "loss": 0.4268,
      "step": 33
    },
    {
      "epoch": 0.4722222222222222,
      "grad_norm": 1.6876043891458643,
      "learning_rate": 9.920882422145372e-06,
      "loss": 0.4535,
      "step": 34
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 1.7064855964835324,
      "learning_rate": 9.905890884491196e-06,
      "loss": 0.3926,
      "step": 35
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.823753890056008,
      "learning_rate": 9.889612861977855e-06,
      "loss": 0.4343,
      "step": 36
    },
    {
      "epoch": 0.5138888888888888,
      "grad_norm": 1.8529840016274717,
      "learning_rate": 9.872052623234632e-06,
      "loss": 0.4281,
      "step": 37
    },
    {
      "epoch": 0.5277777777777778,
      "grad_norm": 1.955408547526647,
      "learning_rate": 9.853214773129796e-06,
      "loss": 0.4148,
      "step": 38
    },
    {
      "epoch": 0.5416666666666666,
      "grad_norm": 1.6681385782954192,
      "learning_rate": 9.833104251563058e-06,
      "loss": 0.3554,
      "step": 39
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 1.7455115554623304,
      "learning_rate": 9.811726332170153e-06,
      "loss": 0.3806,
      "step": 40
    },
    {
      "epoch": 0.5694444444444444,
      "grad_norm": 1.808587406097976,
      "learning_rate": 9.789086620939936e-06,
      "loss": 0.318,
      "step": 41
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 2.0415709012615637,
      "learning_rate": 9.765191054744305e-06,
      "loss": 0.403,
      "step": 42
    },
    {
      "epoch": 0.5972222222222222,
      "grad_norm": 1.5062509351045965,
      "learning_rate": 9.740045899781353e-06,
      "loss": 0.3219,
      "step": 43
    },
    {
      "epoch": 0.6111111111111112,
      "grad_norm": 1.8233835485592458,
      "learning_rate": 9.713657749932172e-06,
      "loss": 0.3496,
      "step": 44
    },
    {
      "epoch": 0.625,
      "grad_norm": 1.55248703557286,
      "learning_rate": 9.68603352503172e-06,
      "loss": 0.3158,
      "step": 45
    },
    {
      "epoch": 0.6388888888888888,
      "grad_norm": 1.2775572351579332,
      "learning_rate": 9.657180469054213e-06,
      "loss": 0.3026,
      "step": 46
    },
    {
      "epoch": 0.6527777777777778,
      "grad_norm": 1.6282146735823924,
      "learning_rate": 9.627106148213521e-06,
      "loss": 0.2771,
      "step": 47
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.9376858537829651,
      "learning_rate": 9.595818448979061e-06,
      "loss": 0.2237,
      "step": 48
    },
    {
      "epoch": 0.6805555555555556,
      "grad_norm": 1.6898368689684822,
      "learning_rate": 9.563325576007702e-06,
      "loss": 0.3631,
      "step": 49
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 1.614524493713582,
      "learning_rate": 9.529636049992235e-06,
      "loss": 0.2762,
      "step": 50
    },
    {
      "epoch": 0.7083333333333334,
      "grad_norm": 1.4097198478742305,
      "learning_rate": 9.494758705426978e-06,
      "loss": 0.2929,
      "step": 51
    },
    {
      "epoch": 0.7222222222222222,
      "grad_norm": 1.2338452097484094,
      "learning_rate": 9.458702688291072e-06,
      "loss": 0.2741,
      "step": 52
    },
    {
      "epoch": 0.7361111111111112,
      "grad_norm": 1.5056629548281208,
      "learning_rate": 9.421477453650118e-06,
      "loss": 0.2946,
      "step": 53
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.597626165346704,
      "learning_rate": 9.38309276317674e-06,
      "loss": 0.3472,
      "step": 54
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 1.2539185771603638,
      "learning_rate": 9.343558682590757e-06,
      "loss": 0.3247,
      "step": 55
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 1.5073835531227189,
      "learning_rate": 9.302885579019626e-06,
      "loss": 0.3089,
      "step": 56
    },
    {
      "epoch": 0.7916666666666666,
      "grad_norm": 1.1716436585494405,
      "learning_rate": 9.261084118279846e-06,
      "loss": 0.2243,
      "step": 57
    },
    {
      "epoch": 0.8055555555555556,
      "grad_norm": 1.615562349308342,
      "learning_rate": 9.218165262080024e-06,
      "loss": 0.3813,
      "step": 58
    },
    {
      "epoch": 0.8194444444444444,
      "grad_norm": 1.126432528226203,
      "learning_rate": 9.174140265146356e-06,
      "loss": 0.2131,
      "step": 59
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 4.944982415856606,
      "learning_rate": 9.129020672271283e-06,
      "loss": 0.3533,
      "step": 60
    },
    {
      "epoch": 0.8472222222222222,
      "grad_norm": 1.8627430033355397,
      "learning_rate": 9.082818315286054e-06,
      "loss": 0.2986,
      "step": 61
    },
    {
      "epoch": 0.8611111111111112,
      "grad_norm": 1.247452599538563,
      "learning_rate": 9.035545309958048e-06,
      "loss": 0.309,
      "step": 62
    },
    {
      "epoch": 0.875,
      "grad_norm": 1.4040277782456316,
      "learning_rate": 8.987214052813605e-06,
      "loss": 0.3614,
      "step": 63
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 1.726105636424933,
      "learning_rate": 8.937837217887273e-06,
      "loss": 0.2789,
      "step": 64
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 1.3818975343126196,
      "learning_rate": 8.887427753398249e-06,
      "loss": 0.295,
      "step": 65
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 1.40890129019135,
      "learning_rate": 8.83599887835493e-06,
      "loss": 0.3319,
      "step": 66
    },
    {
      "epoch": 0.9305555555555556,
      "grad_norm": 1.1840920751941846,
      "learning_rate": 8.783564079088478e-06,
      "loss": 0.2327,
      "step": 67
    },
    {
      "epoch": 0.9444444444444444,
      "grad_norm": 1.6408315149496815,
      "learning_rate": 8.730137105716231e-06,
      "loss": 0.3666,
      "step": 68
    },
    {
      "epoch": 0.9583333333333334,
      "grad_norm": 1.3727819885809827,
      "learning_rate": 8.675731968536004e-06,
      "loss": 0.2455,
      "step": 69
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 1.2514474549155656,
      "learning_rate": 8.620362934352109e-06,
      "loss": 0.2722,
      "step": 70
    },
    {
      "epoch": 0.9861111111111112,
      "grad_norm": 1.7164065718751929,
      "learning_rate": 8.564044522734147e-06,
      "loss": 0.382,
      "step": 71
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.471855648534107,
      "learning_rate": 8.506791502209497e-06,
      "loss": 0.2948,
      "step": 72
    },
    {
      "epoch": 1.0138888888888888,
      "grad_norm": 0.9301287212201296,
      "learning_rate": 8.448618886390523e-06,
      "loss": 0.1474,
      "step": 73
    },
    {
      "epoch": 1.0277777777777777,
      "grad_norm": 1.3078419454322334,
      "learning_rate": 8.389541930037516e-06,
      "loss": 0.1895,
      "step": 74
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 1.2172523796628973,
      "learning_rate": 8.329576125058406e-06,
      "loss": 0.2487,
      "step": 75
    },
    {
      "epoch": 1.0555555555555556,
      "grad_norm": 1.1087278861942516,
      "learning_rate": 8.268737196446264e-06,
      "loss": 0.157,
      "step": 76
    },
    {
      "epoch": 1.0694444444444444,
      "grad_norm": 1.1299518077880732,
      "learning_rate": 8.207041098155701e-06,
      "loss": 0.2375,
      "step": 77
    },
    {
      "epoch": 1.0833333333333333,
      "grad_norm": 1.0822564340528005,
      "learning_rate": 8.144504008919224e-06,
      "loss": 0.1659,
      "step": 78
    },
    {
      "epoch": 1.0972222222222223,
      "grad_norm": 1.3112345018169005,
      "learning_rate": 8.081142328004638e-06,
      "loss": 0.224,
      "step": 79
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 1.2763077943589756,
      "learning_rate": 8.016972670914624e-06,
      "loss": 0.2167,
      "step": 80
    },
    {
      "epoch": 1.125,
      "grad_norm": 1.3430559699696374,
      "learning_rate": 7.952011865029614e-06,
      "loss": 0.2392,
      "step": 81
    },
    {
      "epoch": 1.1388888888888888,
      "grad_norm": 1.2018334484401647,
      "learning_rate": 7.886276945195098e-06,
      "loss": 0.2069,
      "step": 82
    },
    {
      "epoch": 1.1527777777777777,
      "grad_norm": 1.036290700090147,
      "learning_rate": 7.819785149254534e-06,
      "loss": 0.1471,
      "step": 83
    },
    {
      "epoch": 1.1666666666666667,
      "grad_norm": 1.1280595707414307,
      "learning_rate": 7.752553913529019e-06,
      "loss": 0.2057,
      "step": 84
    },
    {
      "epoch": 1.1805555555555556,
      "grad_norm": 1.214652271680462,
      "learning_rate": 7.68460086824492e-06,
      "loss": 0.1965,
      "step": 85
    },
    {
      "epoch": 1.1944444444444444,
      "grad_norm": 1.2598327043593027,
      "learning_rate": 7.61594383291065e-06,
      "loss": 0.1703,
      "step": 86
    },
    {
      "epoch": 1.2083333333333333,
      "grad_norm": 1.2134283432839454,
      "learning_rate": 7.546600811643816e-06,
      "loss": 0.196,
      "step": 87
    },
    {
      "epoch": 1.2222222222222223,
      "grad_norm": 1.3533098727352946,
      "learning_rate": 7.476589988449939e-06,
      "loss": 0.2001,
      "step": 88
    },
    {
      "epoch": 1.2361111111111112,
      "grad_norm": 1.4766618558288398,
      "learning_rate": 7.405929722454026e-06,
      "loss": 0.2699,
      "step": 89
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.0266767933675947,
      "learning_rate": 7.334638543086203e-06,
      "loss": 0.1974,
      "step": 90
    },
    {
      "epoch": 1.2638888888888888,
      "grad_norm": 1.1946430741459986,
      "learning_rate": 7.262735145222696e-06,
      "loss": 0.2071,
      "step": 91
    },
    {
      "epoch": 1.2777777777777777,
      "grad_norm": 1.088741034243188,
      "learning_rate": 7.190238384283413e-06,
      "loss": 0.1784,
      "step": 92
    },
    {
      "epoch": 1.2916666666666667,
      "grad_norm": 1.2058899949869668,
      "learning_rate": 7.117167271287453e-06,
      "loss": 0.1919,
      "step": 93
    },
    {
      "epoch": 1.3055555555555556,
      "grad_norm": 1.2812552163649695,
      "learning_rate": 7.043540967867782e-06,
      "loss": 0.2557,
      "step": 94
    },
    {
      "epoch": 1.3194444444444444,
      "grad_norm": 1.153293550437067,
      "learning_rate": 6.969378781246436e-06,
      "loss": 0.196,
      "step": 95
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 1.2650327267209625,
      "learning_rate": 6.894700159171535e-06,
      "loss": 0.2084,
      "step": 96
    },
    {
      "epoch": 1.3472222222222223,
      "grad_norm": 1.419643472824945,
      "learning_rate": 6.819524684817439e-06,
      "loss": 0.2273,
      "step": 97
    },
    {
      "epoch": 1.3611111111111112,
      "grad_norm": 1.311304155370212,
      "learning_rate": 6.743872071649411e-06,
      "loss": 0.2371,
      "step": 98
    },
    {
      "epoch": 1.375,
      "grad_norm": 1.29047515065847,
      "learning_rate": 6.667762158254104e-06,
      "loss": 0.2121,
      "step": 99
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 1.0545245202437472,
      "learning_rate": 6.591214903137221e-06,
      "loss": 0.1825,
      "step": 100
    },
    {
      "epoch": 1.4027777777777777,
      "grad_norm": 1.148615805854231,
      "learning_rate": 6.514250379489754e-06,
      "loss": 0.1796,
      "step": 101
    },
    {
      "epoch": 1.4166666666666667,
      "grad_norm": 1.1846417241397067,
      "learning_rate": 6.436888769924142e-06,
      "loss": 0.2014,
      "step": 102
    },
    {
      "epoch": 1.4305555555555556,
      "grad_norm": 0.955857633717501,
      "learning_rate": 6.3591503611817155e-06,
      "loss": 0.1732,
      "step": 103
    },
    {
      "epoch": 1.4444444444444444,
      "grad_norm": 0.984712498915309,
      "learning_rate": 6.281055538812861e-06,
      "loss": 0.2064,
      "step": 104
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 1.2434367138483469,
      "learning_rate": 6.202624781831269e-06,
      "loss": 0.171,
      "step": 105
    },
    {
      "epoch": 1.4722222222222223,
      "grad_norm": 1.398866584019501,
      "learning_rate": 6.123878657343648e-06,
      "loss": 0.2194,
      "step": 106
    },
    {
      "epoch": 1.4861111111111112,
      "grad_norm": 0.9074792928724427,
      "learning_rate": 6.044837815156377e-06,
      "loss": 0.1361,
      "step": 107
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.095019701696862,
      "learning_rate": 5.965522982360441e-06,
      "loss": 0.15,
      "step": 108
    },
    {
      "epoch": 1.5138888888888888,
      "grad_norm": 0.9623783171866812,
      "learning_rate": 5.885954957896115e-06,
      "loss": 0.1418,
      "step": 109
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 1.1829230695836053,
      "learning_rate": 5.806154607098799e-06,
      "loss": 0.192,
      "step": 110
    },
    {
      "epoch": 1.5416666666666665,
      "grad_norm": 1.1362402732847106,
      "learning_rate": 5.726142856227453e-06,
      "loss": 0.1664,
      "step": 111
    },
    {
      "epoch": 1.5555555555555556,
      "grad_norm": 1.2625152001636257,
      "learning_rate": 5.645940686977033e-06,
      "loss": 0.2085,
      "step": 112
    },
    {
      "epoch": 1.5694444444444444,
      "grad_norm": 1.1510624195725618,
      "learning_rate": 5.5655691309764225e-06,
      "loss": 0.1538,
      "step": 113
    },
    {
      "epoch": 1.5833333333333335,
      "grad_norm": 1.229794711647949,
      "learning_rate": 5.485049264273241e-06,
      "loss": 0.1561,
      "step": 114
    },
    {
      "epoch": 1.5972222222222223,
      "grad_norm": 1.3857203830431617,
      "learning_rate": 5.404402201807022e-06,
      "loss": 0.181,
      "step": 115
    },
    {
      "epoch": 1.6111111111111112,
      "grad_norm": 1.2727585502342074,
      "learning_rate": 5.323649091872179e-06,
      "loss": 0.179,
      "step": 116
    },
    {
      "epoch": 1.625,
      "grad_norm": 1.0778825699382022,
      "learning_rate": 5.242811110572243e-06,
      "loss": 0.2143,
      "step": 117
    },
    {
      "epoch": 1.6388888888888888,
      "grad_norm": 1.195532097387673,
      "learning_rate": 5.161909456266781e-06,
      "loss": 0.1725,
      "step": 118
    },
    {
      "epoch": 1.6527777777777777,
      "grad_norm": 1.3708575186061502,
      "learning_rate": 5.080965344012509e-06,
      "loss": 0.2151,
      "step": 119
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.1094569783026915,
      "learning_rate": 5e-06,
      "loss": 0.1856,
      "step": 120
    },
    {
      "epoch": 1.6805555555555556,
      "grad_norm": 0.9572962419430803,
      "learning_rate": 4.919034655987493e-06,
      "loss": 0.1237,
      "step": 121
    },
    {
      "epoch": 1.6944444444444444,
      "grad_norm": 1.3697526165030556,
      "learning_rate": 4.838090543733222e-06,
      "loss": 0.1917,
      "step": 122
    },
    {
      "epoch": 1.7083333333333335,
      "grad_norm": 0.9797668524681393,
      "learning_rate": 4.757188889427761e-06,
      "loss": 0.1403,
      "step": 123
    },
    {
      "epoch": 1.7222222222222223,
      "grad_norm": 1.1177025083917884,
      "learning_rate": 4.6763509081278215e-06,
      "loss": 0.1607,
      "step": 124
    },
    {
      "epoch": 1.7361111111111112,
      "grad_norm": 0.8736820887389185,
      "learning_rate": 4.59559779819298e-06,
      "loss": 0.1283,
      "step": 125
    },
    {
      "epoch": 1.75,
      "grad_norm": 1.2109157586571146,
      "learning_rate": 4.51495073572676e-06,
      "loss": 0.1741,
      "step": 126
    },
    {
      "epoch": 1.7638888888888888,
      "grad_norm": 1.2535877794776555,
      "learning_rate": 4.434430869023579e-06,
      "loss": 0.1698,
      "step": 127
    },
    {
      "epoch": 1.7777777777777777,
      "grad_norm": 1.1351713435610022,
      "learning_rate": 4.3540593130229695e-06,
      "loss": 0.2004,
      "step": 128
    },
    {
      "epoch": 1.7916666666666665,
      "grad_norm": 1.2258838345245995,
      "learning_rate": 4.27385714377255e-06,
      "loss": 0.1492,
      "step": 129
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 0.9541560169459083,
      "learning_rate": 4.1938453929012014e-06,
      "loss": 0.1332,
      "step": 130
    },
    {
      "epoch": 1.8194444444444444,
      "grad_norm": 0.9127960838926955,
      "learning_rate": 4.1140450421038865e-06,
      "loss": 0.1248,
      "step": 131
    },
    {
      "epoch": 1.8333333333333335,
      "grad_norm": 1.1807911880234168,
      "learning_rate": 4.034477017639561e-06,
      "loss": 0.2108,
      "step": 132
    },
    {
      "epoch": 1.8472222222222223,
      "grad_norm": 1.2833323008527484,
      "learning_rate": 3.955162184843625e-06,
      "loss": 0.1841,
      "step": 133
    },
    {
      "epoch": 1.8611111111111112,
      "grad_norm": 0.9828910910855105,
      "learning_rate": 3.8761213426563546e-06,
      "loss": 0.1318,
      "step": 134
    },
    {
      "epoch": 1.875,
      "grad_norm": 1.186035054075873,
      "learning_rate": 3.7973752181687336e-06,
      "loss": 0.163,
      "step": 135
    },
    {
      "epoch": 1.8888888888888888,
      "grad_norm": 1.1779706537877848,
      "learning_rate": 3.7189444611871383e-06,
      "loss": 0.1882,
      "step": 136
    },
    {
      "epoch": 1.9027777777777777,
      "grad_norm": 1.1354177782906174,
      "learning_rate": 3.6408496388182857e-06,
      "loss": 0.2008,
      "step": 137
    },
    {
      "epoch": 1.9166666666666665,
      "grad_norm": 1.4059571469655092,
      "learning_rate": 3.5631112300758595e-06,
      "loss": 0.1944,
      "step": 138
    },
    {
      "epoch": 1.9305555555555556,
      "grad_norm": 1.2175488933819723,
      "learning_rate": 3.4857496205102475e-06,
      "loss": 0.1748,
      "step": 139
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 1.5683237293883645,
      "learning_rate": 3.4087850968627823e-06,
      "loss": 0.2047,
      "step": 140
    },
    {
      "epoch": 1.9583333333333335,
      "grad_norm": 1.0174411846084366,
      "learning_rate": 3.3322378417458985e-06,
      "loss": 0.1486,
      "step": 141
    },
    {
      "epoch": 1.9722222222222223,
      "grad_norm": 1.103126125153067,
      "learning_rate": 3.2561279283505888e-06,
      "loss": 0.1531,
      "step": 142
    },
    {
      "epoch": 1.9861111111111112,
      "grad_norm": 1.215924865841863,
      "learning_rate": 3.180475315182563e-06,
      "loss": 0.1923,
      "step": 143
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3959859902744824,
      "learning_rate": 3.1052998408284664e-06,
      "loss": 0.1844,
      "step": 144
    },
    {
      "epoch": 2.013888888888889,
      "grad_norm": 0.8850545165982937,
      "learning_rate": 3.0306212187535653e-06,
      "loss": 0.1055,
      "step": 145
    },
    {
      "epoch": 2.0277777777777777,
      "grad_norm": 0.7392472686706152,
      "learning_rate": 2.9564590321322206e-06,
      "loss": 0.1103,
      "step": 146
    },
    {
      "epoch": 2.0416666666666665,
      "grad_norm": 1.0474626499239874,
      "learning_rate": 2.882832728712551e-06,
      "loss": 0.122,
      "step": 147
    },
    {
      "epoch": 2.0555555555555554,
      "grad_norm": 1.054666922857715,
      "learning_rate": 2.8097616157165886e-06,
      "loss": 0.1122,
      "step": 148
    },
    {
      "epoch": 2.0694444444444446,
      "grad_norm": 0.8634988046465579,
      "learning_rate": 2.7372648547773063e-06,
      "loss": 0.1,
      "step": 149
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.8345756241699817,
      "learning_rate": 2.665361456913797e-06,
      "loss": 0.078,
      "step": 150
    },
    {
      "epoch": 2.0972222222222223,
      "grad_norm": 0.9176687953898187,
      "learning_rate": 2.594070277545975e-06,
      "loss": 0.0844,
      "step": 151
    },
    {
      "epoch": 2.111111111111111,
      "grad_norm": 0.854021939355199,
      "learning_rate": 2.5234100115500643e-06,
      "loss": 0.1098,
      "step": 152
    },
    {
      "epoch": 2.125,
      "grad_norm": 1.0989534848214524,
      "learning_rate": 2.4533991883561868e-06,
      "loss": 0.0988,
      "step": 153
    },
    {
      "epoch": 2.138888888888889,
      "grad_norm": 0.8709103509567065,
      "learning_rate": 2.38405616708935e-06,
      "loss": 0.0934,
      "step": 154
    },
    {
      "epoch": 2.1527777777777777,
      "grad_norm": 1.0463943582674686,
      "learning_rate": 2.315399131755081e-06,
      "loss": 0.1018,
      "step": 155
    },
    {
      "epoch": 2.1666666666666665,
      "grad_norm": 0.8604664734980788,
      "learning_rate": 2.2474460864709825e-06,
      "loss": 0.0678,
      "step": 156
    },
    {
      "epoch": 2.1805555555555554,
      "grad_norm": 1.1367319188924512,
      "learning_rate": 2.1802148507454675e-06,
      "loss": 0.0875,
      "step": 157
    },
    {
      "epoch": 2.1944444444444446,
      "grad_norm": 1.0542976126951007,
      "learning_rate": 2.1137230548049042e-06,
      "loss": 0.0941,
      "step": 158
    },
    {
      "epoch": 2.2083333333333335,
      "grad_norm": 0.9687196847921362,
      "learning_rate": 2.0479881349703885e-06,
      "loss": 0.076,
      "step": 159
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 1.0348517428949806,
      "learning_rate": 1.983027329085377e-06,
      "loss": 0.1031,
      "step": 160
    },
    {
      "epoch": 2.236111111111111,
      "grad_norm": 1.0051006664347915,
      "learning_rate": 1.9188576719953635e-06,
      "loss": 0.0841,
      "step": 161
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.8687076592134144,
      "learning_rate": 1.8554959910807773e-06,
      "loss": 0.0829,
      "step": 162
    },
    {
      "epoch": 2.263888888888889,
      "grad_norm": 0.8929133761206014,
      "learning_rate": 1.7929589018443016e-06,
      "loss": 0.0874,
      "step": 163
    },
    {
      "epoch": 2.2777777777777777,
      "grad_norm": 1.0709519883057992,
      "learning_rate": 1.7312628035537388e-06,
      "loss": 0.091,
      "step": 164
    },
    {
      "epoch": 2.2916666666666665,
      "grad_norm": 1.2071519721282518,
      "learning_rate": 1.6704238749415958e-06,
      "loss": 0.1085,
      "step": 165
    },
    {
      "epoch": 2.3055555555555554,
      "grad_norm": 0.9678231989795624,
      "learning_rate": 1.6104580699624839e-06,
      "loss": 0.0835,
      "step": 166
    },
    {
      "epoch": 2.3194444444444446,
      "grad_norm": 0.9103991479696918,
      "learning_rate": 1.5513811136094786e-06,
      "loss": 0.0794,
      "step": 167
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 1.1173344715303095,
      "learning_rate": 1.4932084977905043e-06,
      "loss": 0.094,
      "step": 168
    },
    {
      "epoch": 2.3472222222222223,
      "grad_norm": 1.0350900131357783,
      "learning_rate": 1.4359554772658551e-06,
      "loss": 0.0839,
      "step": 169
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 1.0879004203081726,
      "learning_rate": 1.3796370656478936e-06,
      "loss": 0.0782,
      "step": 170
    },
    {
      "epoch": 2.375,
      "grad_norm": 1.0433434569228495,
      "learning_rate": 1.3242680314639995e-06,
      "loss": 0.092,
      "step": 171
    },
    {
      "epoch": 2.388888888888889,
      "grad_norm": 1.0472031021923118,
      "learning_rate": 1.2698628942837698e-06,
      "loss": 0.108,
      "step": 172
    },
    {
      "epoch": 2.4027777777777777,
      "grad_norm": 1.0870664996240922,
      "learning_rate": 1.2164359209115235e-06,
      "loss": 0.1135,
      "step": 173
    },
    {
      "epoch": 2.4166666666666665,
      "grad_norm": 1.031076769540047,
      "learning_rate": 1.164001121645069e-06,
      "loss": 0.081,
      "step": 174
    },
    {
      "epoch": 2.4305555555555554,
      "grad_norm": 0.8977382545078525,
      "learning_rate": 1.1125722466017547e-06,
      "loss": 0.1045,
      "step": 175
    },
    {
      "epoch": 2.4444444444444446,
      "grad_norm": 0.9045454769961017,
      "learning_rate": 1.062162782112729e-06,
      "loss": 0.0676,
      "step": 176
    },
    {
      "epoch": 2.4583333333333335,
      "grad_norm": 1.1670937520014997,
      "learning_rate": 1.012785947186397e-06,
      "loss": 0.0974,
      "step": 177
    },
    {
      "epoch": 2.4722222222222223,
      "grad_norm": 0.8952684127297614,
      "learning_rate": 9.644546900419533e-07,
      "loss": 0.0779,
      "step": 178
    },
    {
      "epoch": 2.486111111111111,
      "grad_norm": 1.015060814284756,
      "learning_rate": 9.171816847139447e-07,
      "loss": 0.0568,
      "step": 179
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.180957273322683,
      "learning_rate": 8.709793277287182e-07,
      "loss": 0.0878,
      "step": 180
    },
    {
      "epoch": 2.513888888888889,
      "grad_norm": 1.0199431884781058,
      "learning_rate": 8.258597348536452e-07,
      "loss": 0.0901,
      "step": 181
    },
    {
      "epoch": 2.5277777777777777,
      "grad_norm": 1.2119814715127268,
      "learning_rate": 7.818347379199781e-07,
      "loss": 0.0812,
      "step": 182
    },
    {
      "epoch": 2.5416666666666665,
      "grad_norm": 0.8216318676650253,
      "learning_rate": 7.389158817201541e-07,
      "loss": 0.0648,
      "step": 183
    },
    {
      "epoch": 2.5555555555555554,
      "grad_norm": 1.100604251520046,
      "learning_rate": 6.971144209803738e-07,
      "loss": 0.0696,
      "step": 184
    },
    {
      "epoch": 2.5694444444444446,
      "grad_norm": 0.814335995041825,
      "learning_rate": 6.564413174092443e-07,
      "loss": 0.0581,
      "step": 185
    },
    {
      "epoch": 2.5833333333333335,
      "grad_norm": 0.9464152422858562,
      "learning_rate": 6.16907236823262e-07,
      "loss": 0.1157,
      "step": 186
    },
    {
      "epoch": 2.5972222222222223,
      "grad_norm": 0.9089497799859633,
      "learning_rate": 5.785225463498828e-07,
      "loss": 0.0685,
      "step": 187
    },
    {
      "epoch": 2.611111111111111,
      "grad_norm": 1.202682179578493,
      "learning_rate": 5.412973117089288e-07,
      "loss": 0.1026,
      "step": 188
    },
    {
      "epoch": 2.625,
      "grad_norm": 0.9670028636668581,
      "learning_rate": 5.05241294573024e-07,
      "loss": 0.0821,
      "step": 189
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 1.344972972936964,
      "learning_rate": 4.7036395000776556e-07,
      "loss": 0.0792,
      "step": 190
    },
    {
      "epoch": 2.6527777777777777,
      "grad_norm": 1.0445586830246816,
      "learning_rate": 4.3667442399229985e-07,
      "loss": 0.084,
      "step": 191
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.9833141979119899,
      "learning_rate": 4.041815510209396e-07,
      "loss": 0.0823,
      "step": 192
    },
    {
      "epoch": 2.6805555555555554,
      "grad_norm": 0.88332376145521,
      "learning_rate": 3.728938517864794e-07,
      "loss": 0.0873,
      "step": 193
    },
    {
      "epoch": 2.6944444444444446,
      "grad_norm": 1.184763496189507,
      "learning_rate": 3.4281953094578877e-07,
      "loss": 0.0784,
      "step": 194
    },
    {
      "epoch": 2.7083333333333335,
      "grad_norm": 1.0275513222738037,
      "learning_rate": 3.1396647496828245e-07,
      "loss": 0.077,
      "step": 195
    },
    {
      "epoch": 2.7222222222222223,
      "grad_norm": 1.0498450260383423,
      "learning_rate": 2.8634225006782867e-07,
      "loss": 0.0756,
      "step": 196
    },
    {
      "epoch": 2.736111111111111,
      "grad_norm": 1.0486480992705092,
      "learning_rate": 2.599541002186479e-07,
      "loss": 0.0834,
      "step": 197
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.8375469665537855,
      "learning_rate": 2.3480894525569564e-07,
      "loss": 0.0628,
      "step": 198
    },
    {
      "epoch": 2.763888888888889,
      "grad_norm": 1.0594132145555728,
      "learning_rate": 2.109133790600648e-07,
      "loss": 0.0918,
      "step": 199
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 1.2282514718503594,
      "learning_rate": 1.8827366782984913e-07,
      "loss": 0.0802,
      "step": 200
    },
    {
      "epoch": 2.7916666666666665,
      "grad_norm": 0.8646193977379852,
      "learning_rate": 1.6689574843694433e-07,
      "loss": 0.0768,
      "step": 201
    },
    {
      "epoch": 2.8055555555555554,
      "grad_norm": 1.0784981205095032,
      "learning_rate": 1.4678522687020414e-07,
      "loss": 0.1026,
      "step": 202
    },
    {
      "epoch": 2.8194444444444446,
      "grad_norm": 1.061569942050643,
      "learning_rate": 1.2794737676536993e-07,
      "loss": 0.0877,
      "step": 203
    },
    {
      "epoch": 2.8333333333333335,
      "grad_norm": 1.2667890122796104,
      "learning_rate": 1.1038713802214718e-07,
      "loss": 0.1095,
      "step": 204
    },
    {
      "epoch": 2.8472222222222223,
      "grad_norm": 1.088419434707559,
      "learning_rate": 9.410911550880474e-08,
      "loss": 0.0845,
      "step": 205
    },
    {
      "epoch": 2.861111111111111,
      "grad_norm": 0.9918023275346972,
      "learning_rate": 7.911757785462882e-08,
      "loss": 0.0764,
      "step": 206
    },
    {
      "epoch": 2.875,
      "grad_norm": 1.2368113950486648,
      "learning_rate": 6.54164563305465e-08,
      "loss": 0.091,
      "step": 207
    },
    {
      "epoch": 2.888888888888889,
      "grad_norm": 1.1107893783681875,
      "learning_rate": 5.3009343818219985e-08,
      "loss": 0.0825,
      "step": 208
    },
    {
      "epoch": 2.9027777777777777,
      "grad_norm": 0.9976272341662271,
      "learning_rate": 4.189949386787462e-08,
      "loss": 0.0823,
      "step": 209
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 0.9916696957761072,
      "learning_rate": 3.2089819845111946e-08,
      "loss": 0.0985,
      "step": 210
    },
    {
      "epoch": 2.9305555555555554,
      "grad_norm": 0.9316899060331281,
      "learning_rate": 2.358289416693027e-08,
      "loss": 0.07,
      "step": 211
    },
    {
      "epoch": 2.9444444444444446,
      "grad_norm": 1.1156159922068627,
      "learning_rate": 1.6380947627153143e-08,
      "loss": 0.1053,
      "step": 212
    },
    {
      "epoch": 2.9583333333333335,
      "grad_norm": 0.8676304490269722,
      "learning_rate": 1.0485868811441757e-08,
      "loss": 0.053,
      "step": 213
    },
    {
      "epoch": 2.9722222222222223,
      "grad_norm": 1.039180592375512,
      "learning_rate": 5.899203602046655e-09,
      "loss": 0.0705,
      "step": 214
    },
    {
      "epoch": 2.986111111111111,
      "grad_norm": 1.0376964281779484,
      "learning_rate": 2.6221547724253337e-09,
      "loss": 0.0935,
      "step": 215
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.042899173934309,
      "learning_rate": 6.555816718389896e-10,
      "loss": 0.0738,
      "step": 216
    }
  ],
  "logging_steps": 1,
  "max_steps": 216,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 82435107323904.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}