{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 234,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004282082162451492,
      "grad_norm": 2.3133654594421387,
      "learning_rate": 0.0,
      "loss": 0.9749,
      "step": 1
    },
    {
      "epoch": 0.008564164324902984,
      "grad_norm": 1.8335061073303223,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.7648,
      "step": 2
    },
    {
      "epoch": 0.012846246487354477,
      "grad_norm": 2.2353498935699463,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.9076,
      "step": 3
    },
    {
      "epoch": 0.017128328649805968,
      "grad_norm": 2.1200685501098633,
      "learning_rate": 1e-05,
      "loss": 0.8221,
      "step": 4
    },
    {
      "epoch": 0.02141041081225746,
      "grad_norm": 2.185300350189209,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.8972,
      "step": 5
    },
    {
      "epoch": 0.025692492974708953,
      "grad_norm": 2.383875608444214,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.8918,
      "step": 6
    },
    {
      "epoch": 0.029974575137160443,
      "grad_norm": 2.272465467453003,
      "learning_rate": 2e-05,
      "loss": 0.803,
      "step": 7
    },
    {
      "epoch": 0.034256657299611935,
      "grad_norm": 1.9353933334350586,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.5844,
      "step": 8
    },
    {
      "epoch": 0.03853873946206343,
      "grad_norm": 1.7865619659423828,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.5373,
      "step": 9
    },
    {
      "epoch": 0.04282082162451492,
      "grad_norm": 1.5210086107254028,
      "learning_rate": 3e-05,
      "loss": 0.4767,
      "step": 10
    },
    {
      "epoch": 0.04710290378696641,
      "grad_norm": 1.1693248748779297,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.3285,
      "step": 11
    },
    {
      "epoch": 0.051384985949417906,
      "grad_norm": 0.8487532138824463,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.2321,
      "step": 12
    },
    {
      "epoch": 0.055667068111869396,
      "grad_norm": 0.8226314187049866,
      "learning_rate": 4e-05,
      "loss": 0.2117,
      "step": 13
    },
    {
      "epoch": 0.059949150274320885,
      "grad_norm": 0.7278441190719604,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.1951,
      "step": 14
    },
    {
      "epoch": 0.06423123243677238,
      "grad_norm": 0.6208946704864502,
      "learning_rate": 4.666666666666667e-05,
      "loss": 0.1739,
      "step": 15
    },
    {
      "epoch": 0.06851331459922387,
      "grad_norm": 0.44936591386795044,
      "learning_rate": 5e-05,
      "loss": 0.1144,
      "step": 16
    },
    {
      "epoch": 0.07279539676167536,
      "grad_norm": 0.42960935831069946,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.1089,
      "step": 17
    },
    {
      "epoch": 0.07707747892412686,
      "grad_norm": 0.4222032427787781,
      "learning_rate": 5.666666666666667e-05,
      "loss": 0.1103,
      "step": 18
    },
    {
      "epoch": 0.08135956108657835,
      "grad_norm": 0.3956957459449768,
      "learning_rate": 6e-05,
      "loss": 0.1025,
      "step": 19
    },
    {
      "epoch": 0.08564164324902984,
      "grad_norm": 0.40594249963760376,
      "learning_rate": 6.333333333333333e-05,
      "loss": 0.0923,
      "step": 20
    },
    {
      "epoch": 0.08992372541148133,
      "grad_norm": 0.4704301059246063,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.119,
      "step": 21
    },
    {
      "epoch": 0.09420580757393282,
      "grad_norm": 0.44591984152793884,
      "learning_rate": 7e-05,
      "loss": 0.1016,
      "step": 22
    },
    {
      "epoch": 0.09848788973638432,
      "grad_norm": 0.4180421233177185,
      "learning_rate": 7.333333333333333e-05,
      "loss": 0.1043,
      "step": 23
    },
    {
      "epoch": 0.10276997189883581,
      "grad_norm": 0.36205992102622986,
      "learning_rate": 7.666666666666667e-05,
      "loss": 0.0952,
      "step": 24
    },
    {
      "epoch": 0.1070520540612873,
      "grad_norm": 0.328995019197464,
      "learning_rate": 8e-05,
      "loss": 0.0846,
      "step": 25
    },
    {
      "epoch": 0.11133413622373879,
      "grad_norm": 0.3208652138710022,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.0886,
      "step": 26
    },
    {
      "epoch": 0.11561621838619028,
      "grad_norm": 0.40951845049858093,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.086,
      "step": 27
    },
    {
      "epoch": 0.11989830054864177,
      "grad_norm": 0.42961806058883667,
      "learning_rate": 9e-05,
      "loss": 0.0922,
      "step": 28
    },
    {
      "epoch": 0.12418038271109327,
      "grad_norm": 0.36131271719932556,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.0837,
      "step": 29
    },
    {
      "epoch": 0.12846246487354476,
      "grad_norm": 0.3856489360332489,
      "learning_rate": 9.666666666666667e-05,
      "loss": 0.0874,
      "step": 30
    },
    {
      "epoch": 0.13274454703599625,
      "grad_norm": 0.3986302614212036,
      "learning_rate": 0.0001,
      "loss": 0.0846,
      "step": 31
    },
    {
      "epoch": 0.13702662919844774,
      "grad_norm": 0.40041351318359375,
      "learning_rate": 9.950980392156864e-05,
      "loss": 0.0878,
      "step": 32
    },
    {
      "epoch": 0.14130871136089923,
      "grad_norm": 0.45992231369018555,
      "learning_rate": 9.901960784313727e-05,
      "loss": 0.0851,
      "step": 33
    },
    {
      "epoch": 0.14559079352335072,
      "grad_norm": 0.41897663474082947,
      "learning_rate": 9.852941176470589e-05,
      "loss": 0.0835,
      "step": 34
    },
    {
      "epoch": 0.1498728756858022,
      "grad_norm": 0.3854196071624756,
      "learning_rate": 9.80392156862745e-05,
      "loss": 0.0649,
      "step": 35
    },
    {
      "epoch": 0.15415495784825373,
      "grad_norm": 0.3468564748764038,
      "learning_rate": 9.754901960784314e-05,
      "loss": 0.0619,
      "step": 36
    },
    {
      "epoch": 0.15843704001070522,
      "grad_norm": 0.3781249225139618,
      "learning_rate": 9.705882352941177e-05,
      "loss": 0.0723,
      "step": 37
    },
    {
      "epoch": 0.1627191221731567,
      "grad_norm": 0.33684301376342773,
      "learning_rate": 9.65686274509804e-05,
      "loss": 0.0619,
      "step": 38
    },
    {
      "epoch": 0.1670012043356082,
      "grad_norm": 0.319573312997818,
      "learning_rate": 9.607843137254903e-05,
      "loss": 0.0711,
      "step": 39
    },
    {
      "epoch": 0.17128328649805968,
      "grad_norm": 0.4080142378807068,
      "learning_rate": 9.558823529411765e-05,
      "loss": 0.0637,
      "step": 40
    },
    {
      "epoch": 0.17556536866051117,
      "grad_norm": 0.34714841842651367,
      "learning_rate": 9.509803921568627e-05,
      "loss": 0.0666,
      "step": 41
    },
    {
      "epoch": 0.17984745082296266,
      "grad_norm": 0.3296521306037903,
      "learning_rate": 9.46078431372549e-05,
      "loss": 0.0639,
      "step": 42
    },
    {
      "epoch": 0.18412953298541415,
      "grad_norm": 0.3505910038948059,
      "learning_rate": 9.411764705882353e-05,
      "loss": 0.0756,
      "step": 43
    },
    {
      "epoch": 0.18841161514786564,
      "grad_norm": 0.37628045678138733,
      "learning_rate": 9.362745098039217e-05,
      "loss": 0.0605,
      "step": 44
    },
    {
      "epoch": 0.19269369731031713,
      "grad_norm": 0.3756776452064514,
      "learning_rate": 9.313725490196079e-05,
      "loss": 0.0695,
      "step": 45
    },
    {
      "epoch": 0.19697577947276865,
      "grad_norm": 0.3502165675163269,
      "learning_rate": 9.264705882352942e-05,
      "loss": 0.0715,
      "step": 46
    },
    {
      "epoch": 0.20125786163522014,
      "grad_norm": 0.38754549622535706,
      "learning_rate": 9.215686274509804e-05,
      "loss": 0.0619,
      "step": 47
    },
    {
      "epoch": 0.20553994379767163,
      "grad_norm": 0.342479407787323,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.0697,
      "step": 48
    },
    {
      "epoch": 0.20982202596012312,
      "grad_norm": 0.34569641947746277,
      "learning_rate": 9.11764705882353e-05,
      "loss": 0.0608,
      "step": 49
    },
    {
      "epoch": 0.2141041081225746,
      "grad_norm": 0.3148820996284485,
      "learning_rate": 9.068627450980393e-05,
      "loss": 0.0553,
      "step": 50
    },
    {
      "epoch": 0.2183861902850261,
      "grad_norm": 0.2726571559906006,
      "learning_rate": 9.019607843137255e-05,
      "loss": 0.0525,
      "step": 51
    },
    {
      "epoch": 0.22266827244747758,
      "grad_norm": 0.2933000326156616,
      "learning_rate": 8.970588235294118e-05,
      "loss": 0.0529,
      "step": 52
    },
    {
      "epoch": 0.22695035460992907,
      "grad_norm": 0.32735496759414673,
      "learning_rate": 8.921568627450981e-05,
      "loss": 0.0654,
      "step": 53
    },
    {
      "epoch": 0.23123243677238056,
      "grad_norm": 0.3951069712638855,
      "learning_rate": 8.872549019607843e-05,
      "loss": 0.0672,
      "step": 54
    },
    {
      "epoch": 0.23551451893483205,
      "grad_norm": 0.3004348874092102,
      "learning_rate": 8.823529411764706e-05,
      "loss": 0.0473,
      "step": 55
    },
    {
      "epoch": 0.23979660109728354,
      "grad_norm": 0.3402177691459656,
      "learning_rate": 8.774509803921568e-05,
      "loss": 0.066,
      "step": 56
    },
    {
      "epoch": 0.24407868325973506,
      "grad_norm": 0.33595505356788635,
      "learning_rate": 8.725490196078432e-05,
      "loss": 0.0514,
      "step": 57
    },
    {
      "epoch": 0.24836076542218655,
      "grad_norm": 0.337112158536911,
      "learning_rate": 8.676470588235295e-05,
      "loss": 0.0572,
      "step": 58
    },
    {
      "epoch": 0.252642847584638,
      "grad_norm": 0.33211183547973633,
      "learning_rate": 8.627450980392158e-05,
      "loss": 0.053,
      "step": 59
    },
    {
      "epoch": 0.2569249297470895,
      "grad_norm": 0.42650142312049866,
      "learning_rate": 8.57843137254902e-05,
      "loss": 0.063,
      "step": 60
    },
    {
      "epoch": 0.261207011909541,
      "grad_norm": 0.34502434730529785,
      "learning_rate": 8.529411764705883e-05,
      "loss": 0.0656,
      "step": 61
    },
    {
      "epoch": 0.2654890940719925,
      "grad_norm": 0.3582489490509033,
      "learning_rate": 8.480392156862745e-05,
      "loss": 0.0476,
      "step": 62
    },
    {
      "epoch": 0.269771176234444,
      "grad_norm": 0.27663132548332214,
      "learning_rate": 8.431372549019608e-05,
      "loss": 0.0545,
      "step": 63
    },
    {
      "epoch": 0.2740532583968955,
      "grad_norm": 0.3107115924358368,
      "learning_rate": 8.382352941176471e-05,
      "loss": 0.0412,
      "step": 64
    },
    {
      "epoch": 0.278335340559347,
      "grad_norm": 0.36757537722587585,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.0468,
      "step": 65
    },
    {
      "epoch": 0.28261742272179846,
      "grad_norm": 0.33179935812950134,
      "learning_rate": 8.284313725490198e-05,
      "loss": 0.0601,
      "step": 66
    },
    {
      "epoch": 0.28689950488425,
      "grad_norm": 0.40680983662605286,
      "learning_rate": 8.23529411764706e-05,
      "loss": 0.0565,
      "step": 67
    },
    {
      "epoch": 0.29118158704670144,
      "grad_norm": 0.3238675594329834,
      "learning_rate": 8.186274509803921e-05,
      "loss": 0.0488,
      "step": 68
    },
    {
      "epoch": 0.29546366920915296,
      "grad_norm": 0.39186710119247437,
      "learning_rate": 8.137254901960785e-05,
      "loss": 0.0505,
      "step": 69
    },
    {
      "epoch": 0.2997457513716044,
      "grad_norm": 0.2939775288105011,
      "learning_rate": 8.088235294117648e-05,
      "loss": 0.0484,
      "step": 70
    },
    {
      "epoch": 0.30402783353405594,
      "grad_norm": 0.37278881669044495,
      "learning_rate": 8.039215686274511e-05,
      "loss": 0.0504,
      "step": 71
    },
    {
      "epoch": 0.30830991569650745,
      "grad_norm": 0.3936941921710968,
      "learning_rate": 7.990196078431373e-05,
      "loss": 0.055,
      "step": 72
    },
    {
      "epoch": 0.3125919978589589,
      "grad_norm": 0.38660097122192383,
      "learning_rate": 7.941176470588235e-05,
      "loss": 0.0821,
      "step": 73
    },
    {
      "epoch": 0.31687408002141043,
      "grad_norm": 0.30945825576782227,
      "learning_rate": 7.892156862745098e-05,
      "loss": 0.0442,
      "step": 74
    },
    {
      "epoch": 0.3211561621838619,
      "grad_norm": 0.31412258744239807,
      "learning_rate": 7.843137254901961e-05,
      "loss": 0.0419,
      "step": 75
    },
    {
      "epoch": 0.3254382443463134,
      "grad_norm": 0.34703952074050903,
      "learning_rate": 7.794117647058824e-05,
      "loss": 0.0575,
      "step": 76
    },
    {
      "epoch": 0.32972032650876487,
      "grad_norm": 0.346296489238739,
      "learning_rate": 7.745098039215687e-05,
      "loss": 0.0603,
      "step": 77
    },
    {
      "epoch": 0.3340024086712164,
      "grad_norm": 0.2502746284008026,
      "learning_rate": 7.696078431372549e-05,
      "loss": 0.0417,
      "step": 78
    },
    {
      "epoch": 0.33828449083366785,
      "grad_norm": 0.3630567193031311,
      "learning_rate": 7.647058823529411e-05,
      "loss": 0.0591,
      "step": 79
    },
    {
      "epoch": 0.34256657299611937,
      "grad_norm": 0.33694136142730713,
      "learning_rate": 7.598039215686274e-05,
      "loss": 0.047,
      "step": 80
    },
    {
      "epoch": 0.34684865515857083,
      "grad_norm": 0.22548595070838928,
      "learning_rate": 7.549019607843137e-05,
      "loss": 0.0448,
      "step": 81
    },
    {
      "epoch": 0.35113073732102235,
      "grad_norm": 0.3071158528327942,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0503,
      "step": 82
    },
    {
      "epoch": 0.35541281948347386,
      "grad_norm": 0.3624606132507324,
      "learning_rate": 7.450980392156864e-05,
      "loss": 0.0475,
      "step": 83
    },
    {
      "epoch": 0.3596949016459253,
      "grad_norm": 0.23342928290367126,
      "learning_rate": 7.401960784313726e-05,
      "loss": 0.0337,
      "step": 84
    },
    {
      "epoch": 0.36397698380837684,
      "grad_norm": 0.2770336866378784,
      "learning_rate": 7.352941176470589e-05,
      "loss": 0.0453,
      "step": 85
    },
    {
      "epoch": 0.3682590659708283,
      "grad_norm": 0.27268415689468384,
      "learning_rate": 7.303921568627451e-05,
      "loss": 0.0508,
      "step": 86
    },
    {
      "epoch": 0.3725411481332798,
      "grad_norm": 0.22464269399642944,
      "learning_rate": 7.254901960784314e-05,
      "loss": 0.0348,
      "step": 87
    },
    {
      "epoch": 0.3768232302957313,
      "grad_norm": 0.27807021141052246,
      "learning_rate": 7.205882352941177e-05,
      "loss": 0.0597,
      "step": 88
    },
    {
      "epoch": 0.3811053124581828,
      "grad_norm": 0.30591973662376404,
      "learning_rate": 7.156862745098039e-05,
      "loss": 0.0532,
      "step": 89
    },
    {
      "epoch": 0.38538739462063426,
      "grad_norm": 0.29489806294441223,
      "learning_rate": 7.107843137254902e-05,
      "loss": 0.057,
      "step": 90
    },
    {
      "epoch": 0.3896694767830858,
      "grad_norm": 0.27410319447517395,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.0529,
      "step": 91
    },
    {
      "epoch": 0.3939515589455373,
      "grad_norm": 0.24769103527069092,
      "learning_rate": 7.009803921568627e-05,
      "loss": 0.0371,
      "step": 92
    },
    {
      "epoch": 0.39823364110798876,
      "grad_norm": 0.30047643184661865,
      "learning_rate": 6.96078431372549e-05,
      "loss": 0.06,
      "step": 93
    },
    {
      "epoch": 0.4025157232704403,
      "grad_norm": 0.2860408425331116,
      "learning_rate": 6.911764705882354e-05,
      "loss": 0.046,
      "step": 94
    },
    {
      "epoch": 0.40679780543289173,
      "grad_norm": 0.2957577407360077,
      "learning_rate": 6.862745098039216e-05,
      "loss": 0.0585,
      "step": 95
    },
    {
      "epoch": 0.41107988759534325,
      "grad_norm": 0.3263351023197174,
      "learning_rate": 6.813725490196079e-05,
      "loss": 0.0538,
      "step": 96
    },
    {
      "epoch": 0.4153619697577947,
      "grad_norm": 0.35550421476364136,
      "learning_rate": 6.764705882352942e-05,
      "loss": 0.0533,
      "step": 97
    },
    {
      "epoch": 0.41964405192024623,
      "grad_norm": 0.2950674295425415,
      "learning_rate": 6.715686274509804e-05,
      "loss": 0.05,
      "step": 98
    },
    {
      "epoch": 0.4239261340826977,
      "grad_norm": 0.2513452172279358,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0361,
      "step": 99
    },
    {
      "epoch": 0.4282082162451492,
      "grad_norm": 0.3161787688732147,
      "learning_rate": 6.61764705882353e-05,
      "loss": 0.0483,
      "step": 100
    },
    {
      "epoch": 0.43249029840760067,
      "grad_norm": 0.3478478789329529,
      "learning_rate": 6.568627450980392e-05,
      "loss": 0.0467,
      "step": 101
    },
    {
      "epoch": 0.4367723805700522,
      "grad_norm": 0.33523473143577576,
      "learning_rate": 6.519607843137255e-05,
      "loss": 0.0593,
      "step": 102
    },
    {
      "epoch": 0.4410544627325037,
      "grad_norm": 0.36270254850387573,
      "learning_rate": 6.470588235294118e-05,
      "loss": 0.061,
      "step": 103
    },
    {
      "epoch": 0.44533654489495517,
      "grad_norm": 0.31555601954460144,
      "learning_rate": 6.421568627450982e-05,
      "loss": 0.0404,
      "step": 104
    },
    {
      "epoch": 0.4496186270574067,
      "grad_norm": 0.2953006625175476,
      "learning_rate": 6.372549019607843e-05,
      "loss": 0.0502,
      "step": 105
    },
    {
      "epoch": 0.45390070921985815,
      "grad_norm": 0.30984699726104736,
      "learning_rate": 6.323529411764705e-05,
      "loss": 0.0594,
      "step": 106
    },
    {
      "epoch": 0.45818279138230966,
      "grad_norm": 0.30836424231529236,
      "learning_rate": 6.274509803921569e-05,
      "loss": 0.04,
      "step": 107
    },
    {
      "epoch": 0.4624648735447611,
      "grad_norm": 0.27564123272895813,
      "learning_rate": 6.225490196078432e-05,
      "loss": 0.0372,
      "step": 108
    },
    {
      "epoch": 0.46674695570721264,
      "grad_norm": 0.31601759791374207,
      "learning_rate": 6.176470588235295e-05,
      "loss": 0.0483,
      "step": 109
    },
    {
      "epoch": 0.4710290378696641,
      "grad_norm": 0.40431514382362366,
      "learning_rate": 6.127450980392158e-05,
      "loss": 0.0531,
      "step": 110
    },
    {
      "epoch": 0.4753111200321156,
      "grad_norm": 0.2956823408603668,
      "learning_rate": 6.078431372549019e-05,
      "loss": 0.0393,
      "step": 111
    },
    {
      "epoch": 0.4795932021945671,
      "grad_norm": 0.26746317744255066,
      "learning_rate": 6.0294117647058825e-05,
      "loss": 0.0435,
      "step": 112
    },
    {
      "epoch": 0.4838752843570186,
      "grad_norm": 0.29554516077041626,
      "learning_rate": 5.980392156862745e-05,
      "loss": 0.0419,
      "step": 113
    },
    {
      "epoch": 0.4881573665194701,
      "grad_norm": 0.37722569704055786,
      "learning_rate": 5.931372549019608e-05,
      "loss": 0.0512,
      "step": 114
    },
    {
      "epoch": 0.4924394486819216,
      "grad_norm": 0.2477983981370926,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.038,
      "step": 115
    },
    {
      "epoch": 0.4967215308443731,
      "grad_norm": 0.22754943370819092,
      "learning_rate": 5.833333333333334e-05,
      "loss": 0.0377,
      "step": 116
    },
    {
      "epoch": 0.5010036130068246,
      "grad_norm": 0.25772324204444885,
      "learning_rate": 5.784313725490197e-05,
      "loss": 0.036,
      "step": 117
    },
    {
      "epoch": 0.505285695169276,
      "grad_norm": 0.2738616168498993,
      "learning_rate": 5.735294117647059e-05,
      "loss": 0.0581,
      "step": 118
    },
    {
      "epoch": 0.5095677773317275,
      "grad_norm": 0.31121647357940674,
      "learning_rate": 5.6862745098039215e-05,
      "loss": 0.0479,
      "step": 119
    },
    {
      "epoch": 0.513849859494179,
      "grad_norm": 0.24819445610046387,
      "learning_rate": 5.637254901960785e-05,
      "loss": 0.0375,
      "step": 120
    },
    {
      "epoch": 0.5181319416566306,
      "grad_norm": 0.3565673232078552,
      "learning_rate": 5.588235294117647e-05,
      "loss": 0.0673,
      "step": 121
    },
    {
      "epoch": 0.522414023819082,
      "grad_norm": 0.2924639880657196,
      "learning_rate": 5.5392156862745104e-05,
      "loss": 0.0439,
      "step": 122
    },
    {
      "epoch": 0.5266961059815335,
      "grad_norm": 0.2511173486709595,
      "learning_rate": 5.490196078431373e-05,
      "loss": 0.043,
      "step": 123
    },
    {
      "epoch": 0.530978188143985,
      "grad_norm": 0.3577595353126526,
      "learning_rate": 5.441176470588235e-05,
      "loss": 0.0573,
      "step": 124
    },
    {
      "epoch": 0.5352602703064365,
      "grad_norm": 0.2690516412258148,
      "learning_rate": 5.392156862745098e-05,
      "loss": 0.0462,
      "step": 125
    },
    {
      "epoch": 0.539542352468888,
      "grad_norm": 0.2612890303134918,
      "learning_rate": 5.343137254901961e-05,
      "loss": 0.0452,
      "step": 126
    },
    {
      "epoch": 0.5438244346313394,
      "grad_norm": 0.27475517988204956,
      "learning_rate": 5.294117647058824e-05,
      "loss": 0.0436,
      "step": 127
    },
    {
      "epoch": 0.548106516793791,
      "grad_norm": 0.33208557963371277,
      "learning_rate": 5.245098039215687e-05,
      "loss": 0.0508,
      "step": 128
    },
    {
      "epoch": 0.5523885989562425,
      "grad_norm": 0.28704285621643066,
      "learning_rate": 5.1960784313725495e-05,
      "loss": 0.0436,
      "step": 129
    },
    {
      "epoch": 0.556670681118694,
      "grad_norm": 0.35237687826156616,
      "learning_rate": 5.147058823529411e-05,
      "loss": 0.0549,
      "step": 130
    },
    {
      "epoch": 0.5609527632811454,
      "grad_norm": 0.290622740983963,
      "learning_rate": 5.0980392156862745e-05,
      "loss": 0.048,
      "step": 131
    },
    {
      "epoch": 0.5652348454435969,
      "grad_norm": 0.250241219997406,
      "learning_rate": 5.049019607843137e-05,
      "loss": 0.0441,
      "step": 132
    },
    {
      "epoch": 0.5695169276060484,
      "grad_norm": 0.3002930283546448,
      "learning_rate": 5e-05,
      "loss": 0.0411,
      "step": 133
    },
    {
      "epoch": 0.5737990097685,
      "grad_norm": 0.30466729402542114,
      "learning_rate": 4.9509803921568634e-05,
      "loss": 0.0452,
      "step": 134
    },
    {
      "epoch": 0.5780810919309515,
      "grad_norm": 0.2513613998889923,
      "learning_rate": 4.901960784313725e-05,
      "loss": 0.0408,
      "step": 135
    },
    {
      "epoch": 0.5823631740934029,
      "grad_norm": 0.30394110083580017,
      "learning_rate": 4.8529411764705885e-05,
      "loss": 0.0451,
      "step": 136
    },
    {
      "epoch": 0.5866452562558544,
      "grad_norm": 0.2374386042356491,
      "learning_rate": 4.803921568627452e-05,
      "loss": 0.0391,
      "step": 137
    },
    {
      "epoch": 0.5909273384183059,
      "grad_norm": 0.3214227259159088,
      "learning_rate": 4.7549019607843135e-05,
      "loss": 0.0428,
      "step": 138
    },
    {
      "epoch": 0.5952094205807574,
      "grad_norm": 0.25820499658584595,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.0458,
      "step": 139
    },
    {
      "epoch": 0.5994915027432088,
      "grad_norm": 0.21727867424488068,
      "learning_rate": 4.656862745098039e-05,
      "loss": 0.0313,
      "step": 140
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.3863449692726135,
      "learning_rate": 4.607843137254902e-05,
      "loss": 0.0578,
      "step": 141
    },
    {
      "epoch": 0.6080556670681119,
      "grad_norm": 0.2634093463420868,
      "learning_rate": 4.558823529411765e-05,
      "loss": 0.0373,
      "step": 142
    },
    {
      "epoch": 0.6123377492305634,
      "grad_norm": 0.2833387553691864,
      "learning_rate": 4.5098039215686275e-05,
      "loss": 0.051,
      "step": 143
    },
    {
      "epoch": 0.6166198313930149,
      "grad_norm": 0.3331435024738312,
      "learning_rate": 4.460784313725491e-05,
      "loss": 0.0639,
      "step": 144
    },
    {
      "epoch": 0.6209019135554663,
      "grad_norm": 0.2673349380493164,
      "learning_rate": 4.411764705882353e-05,
      "loss": 0.0509,
      "step": 145
    },
    {
      "epoch": 0.6251839957179178,
      "grad_norm": 0.3010377585887909,
      "learning_rate": 4.362745098039216e-05,
      "loss": 0.0521,
      "step": 146
    },
    {
      "epoch": 0.6294660778803693,
      "grad_norm": 0.28252658247947693,
      "learning_rate": 4.313725490196079e-05,
      "loss": 0.0455,
      "step": 147
    },
    {
      "epoch": 0.6337481600428209,
      "grad_norm": 0.28520438075065613,
      "learning_rate": 4.2647058823529415e-05,
      "loss": 0.0551,
      "step": 148
    },
    {
      "epoch": 0.6380302422052723,
      "grad_norm": 0.24448159337043762,
      "learning_rate": 4.215686274509804e-05,
      "loss": 0.0362,
      "step": 149
    },
    {
      "epoch": 0.6423123243677238,
      "grad_norm": 0.2240142524242401,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.0377,
      "step": 150
    },
    {
      "epoch": 0.6465944065301753,
      "grad_norm": 0.2896118760108948,
      "learning_rate": 4.11764705882353e-05,
      "loss": 0.0378,
      "step": 151
    },
    {
      "epoch": 0.6508764886926268,
      "grad_norm": 0.2663087844848633,
      "learning_rate": 4.068627450980392e-05,
      "loss": 0.0407,
      "step": 152
    },
    {
      "epoch": 0.6551585708550783,
      "grad_norm": 0.3086792230606079,
      "learning_rate": 4.0196078431372555e-05,
      "loss": 0.0323,
      "step": 153
    },
    {
      "epoch": 0.6594406530175297,
      "grad_norm": 0.2776995897293091,
      "learning_rate": 3.970588235294117e-05,
      "loss": 0.0437,
      "step": 154
    },
    {
      "epoch": 0.6637227351799813,
      "grad_norm": 0.3368375897407532,
      "learning_rate": 3.9215686274509805e-05,
      "loss": 0.0525,
      "step": 155
    },
    {
      "epoch": 0.6680048173424328,
      "grad_norm": 0.25399115681648254,
      "learning_rate": 3.872549019607844e-05,
      "loss": 0.0352,
      "step": 156
    },
    {
      "epoch": 0.6722868995048843,
      "grad_norm": 0.3201920986175537,
      "learning_rate": 3.8235294117647055e-05,
      "loss": 0.046,
      "step": 157
    },
    {
      "epoch": 0.6765689816673357,
      "grad_norm": 0.36324596405029297,
      "learning_rate": 3.774509803921569e-05,
      "loss": 0.0429,
      "step": 158
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.27535951137542725,
      "learning_rate": 3.725490196078432e-05,
      "loss": 0.036,
      "step": 159
    },
    {
      "epoch": 0.6851331459922387,
      "grad_norm": 0.2501991391181946,
      "learning_rate": 3.6764705882352945e-05,
      "loss": 0.0333,
      "step": 160
    },
    {
      "epoch": 0.6894152281546903,
      "grad_norm": 0.26270344853401184,
      "learning_rate": 3.627450980392157e-05,
      "loss": 0.0359,
      "step": 161
    },
    {
      "epoch": 0.6936973103171417,
      "grad_norm": 0.2966891825199127,
      "learning_rate": 3.5784313725490195e-05,
      "loss": 0.0448,
      "step": 162
    },
    {
      "epoch": 0.6979793924795932,
      "grad_norm": 0.24465243518352509,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.0488,
      "step": 163
    },
    {
      "epoch": 0.7022614746420447,
      "grad_norm": 0.2697572410106659,
      "learning_rate": 3.480392156862745e-05,
      "loss": 0.0295,
      "step": 164
    },
    {
      "epoch": 0.7065435568044962,
      "grad_norm": 0.27013951539993286,
      "learning_rate": 3.431372549019608e-05,
      "loss": 0.0395,
      "step": 165
    },
    {
      "epoch": 0.7108256389669477,
      "grad_norm": 0.3045112192630768,
      "learning_rate": 3.382352941176471e-05,
      "loss": 0.0411,
      "step": 166
    },
    {
      "epoch": 0.7151077211293991,
      "grad_norm": 0.24719510972499847,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0354,
      "step": 167
    },
    {
      "epoch": 0.7193898032918506,
      "grad_norm": 0.1892707198858261,
      "learning_rate": 3.284313725490196e-05,
      "loss": 0.0234,
      "step": 168
    },
    {
      "epoch": 0.7236718854543022,
      "grad_norm": 0.30045631527900696,
      "learning_rate": 3.235294117647059e-05,
      "loss": 0.0447,
      "step": 169
    },
    {
      "epoch": 0.7279539676167537,
      "grad_norm": 0.309243768453598,
      "learning_rate": 3.186274509803922e-05,
      "loss": 0.0404,
      "step": 170
    },
    {
      "epoch": 0.7322360497792051,
      "grad_norm": 0.24024784564971924,
      "learning_rate": 3.137254901960784e-05,
      "loss": 0.0395,
      "step": 171
    },
    {
      "epoch": 0.7365181319416566,
      "grad_norm": 0.2552880048751831,
      "learning_rate": 3.0882352941176475e-05,
      "loss": 0.0359,
      "step": 172
    },
    {
      "epoch": 0.7408002141041081,
      "grad_norm": 0.33713486790657043,
      "learning_rate": 3.0392156862745097e-05,
      "loss": 0.0614,
      "step": 173
    },
    {
      "epoch": 0.7450822962665596,
      "grad_norm": 0.2314293533563614,
      "learning_rate": 2.9901960784313725e-05,
      "loss": 0.0348,
      "step": 174
    },
    {
      "epoch": 0.7493643784290112,
      "grad_norm": 0.23023654520511627,
      "learning_rate": 2.9411764705882354e-05,
      "loss": 0.0419,
      "step": 175
    },
    {
      "epoch": 0.7536464605914626,
      "grad_norm": 0.2514519691467285,
      "learning_rate": 2.8921568627450986e-05,
      "loss": 0.0388,
      "step": 176
    },
    {
      "epoch": 0.7579285427539141,
      "grad_norm": 0.2647082209587097,
      "learning_rate": 2.8431372549019608e-05,
      "loss": 0.0463,
      "step": 177
    },
    {
      "epoch": 0.7622106249163656,
      "grad_norm": 0.29537469148635864,
      "learning_rate": 2.7941176470588236e-05,
      "loss": 0.0352,
      "step": 178
    },
    {
      "epoch": 0.7664927070788171,
      "grad_norm": 0.2043161392211914,
      "learning_rate": 2.7450980392156865e-05,
      "loss": 0.0387,
      "step": 179
    },
    {
      "epoch": 0.7707747892412685,
      "grad_norm": 0.305243581533432,
      "learning_rate": 2.696078431372549e-05,
      "loss": 0.0346,
      "step": 180
    },
    {
      "epoch": 0.77505687140372,
      "grad_norm": 0.29974740743637085,
      "learning_rate": 2.647058823529412e-05,
      "loss": 0.0513,
      "step": 181
    },
    {
      "epoch": 0.7793389535661716,
      "grad_norm": 0.28913021087646484,
      "learning_rate": 2.5980392156862747e-05,
      "loss": 0.0341,
      "step": 182
    },
    {
      "epoch": 0.7836210357286231,
      "grad_norm": 0.32380354404449463,
      "learning_rate": 2.5490196078431373e-05,
      "loss": 0.0614,
      "step": 183
    },
    {
      "epoch": 0.7879031178910746,
      "grad_norm": 0.2799640893936157,
      "learning_rate": 2.5e-05,
      "loss": 0.0515,
      "step": 184
    },
    {
      "epoch": 0.792185200053526,
      "grad_norm": 0.28914907574653625,
      "learning_rate": 2.4509803921568626e-05,
      "loss": 0.0393,
      "step": 185
    },
    {
      "epoch": 0.7964672822159775,
      "grad_norm": 0.26201391220092773,
      "learning_rate": 2.401960784313726e-05,
      "loss": 0.0431,
      "step": 186
    },
    {
      "epoch": 0.800749364378429,
      "grad_norm": 0.29834750294685364,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.039,
      "step": 187
    },
    {
      "epoch": 0.8050314465408805,
      "grad_norm": 0.28359827399253845,
      "learning_rate": 2.303921568627451e-05,
      "loss": 0.0474,
      "step": 188
    },
    {
      "epoch": 0.809313528703332,
      "grad_norm": 0.21148712933063507,
      "learning_rate": 2.2549019607843138e-05,
      "loss": 0.0355,
      "step": 189
    },
    {
      "epoch": 0.8135956108657835,
      "grad_norm": 0.30203911662101746,
      "learning_rate": 2.2058823529411766e-05,
      "loss": 0.0423,
      "step": 190
    },
    {
      "epoch": 0.817877693028235,
      "grad_norm": 0.32118844985961914,
      "learning_rate": 2.1568627450980395e-05,
      "loss": 0.0396,
      "step": 191
    },
    {
      "epoch": 0.8221597751906865,
      "grad_norm": 0.2358664572238922,
      "learning_rate": 2.107843137254902e-05,
      "loss": 0.0344,
      "step": 192
    },
    {
      "epoch": 0.8264418573531379,
      "grad_norm": 0.3052966892719269,
      "learning_rate": 2.058823529411765e-05,
      "loss": 0.0459,
      "step": 193
    },
    {
      "epoch": 0.8307239395155894,
      "grad_norm": 0.2716437578201294,
      "learning_rate": 2.0098039215686277e-05,
      "loss": 0.0511,
      "step": 194
    },
    {
      "epoch": 0.8350060216780409,
      "grad_norm": 0.3086344301700592,
      "learning_rate": 1.9607843137254903e-05,
      "loss": 0.0477,
      "step": 195
    },
    {
      "epoch": 0.8392881038404925,
      "grad_norm": 0.26460692286491394,
      "learning_rate": 1.9117647058823528e-05,
      "loss": 0.0355,
      "step": 196
    },
    {
      "epoch": 0.843570186002944,
      "grad_norm": 0.2656954526901245,
      "learning_rate": 1.862745098039216e-05,
      "loss": 0.0504,
      "step": 197
    },
    {
      "epoch": 0.8478522681653954,
      "grad_norm": 0.2059815526008606,
      "learning_rate": 1.8137254901960785e-05,
      "loss": 0.0404,
      "step": 198
    },
    {
      "epoch": 0.8521343503278469,
      "grad_norm": 0.28695958852767944,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.0502,
      "step": 199
    },
    {
      "epoch": 0.8564164324902984,
      "grad_norm": 0.278156042098999,
      "learning_rate": 1.715686274509804e-05,
      "loss": 0.0325,
      "step": 200
    },
    {
      "epoch": 0.8606985146527499,
      "grad_norm": 0.33728834986686707,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.0468,
      "step": 201
    },
    {
      "epoch": 0.8649805968152013,
      "grad_norm": 0.2665461003780365,
      "learning_rate": 1.6176470588235296e-05,
      "loss": 0.041,
      "step": 202
    },
    {
      "epoch": 0.8692626789776529,
      "grad_norm": 0.2768334150314331,
      "learning_rate": 1.568627450980392e-05,
      "loss": 0.0477,
      "step": 203
    },
    {
      "epoch": 0.8735447611401044,
      "grad_norm": 0.2847307324409485,
      "learning_rate": 1.5196078431372548e-05,
      "loss": 0.0368,
      "step": 204
    },
    {
      "epoch": 0.8778268433025559,
      "grad_norm": 0.26865965127944946,
      "learning_rate": 1.4705882352941177e-05,
      "loss": 0.0321,
      "step": 205
    },
    {
      "epoch": 0.8821089254650074,
      "grad_norm": 0.23639728128910065,
      "learning_rate": 1.4215686274509804e-05,
      "loss": 0.0267,
      "step": 206
    },
    {
      "epoch": 0.8863910076274588,
      "grad_norm": 0.22090226411819458,
      "learning_rate": 1.3725490196078432e-05,
      "loss": 0.0283,
      "step": 207
    },
    {
      "epoch": 0.8906730897899103,
      "grad_norm": 0.24385924637317657,
      "learning_rate": 1.323529411764706e-05,
      "loss": 0.0486,
      "step": 208
    },
    {
      "epoch": 0.8949551719523618,
      "grad_norm": 0.267991304397583,
      "learning_rate": 1.2745098039215686e-05,
      "loss": 0.049,
      "step": 209
    },
    {
      "epoch": 0.8992372541148134,
      "grad_norm": 0.2858865261077881,
      "learning_rate": 1.2254901960784313e-05,
      "loss": 0.0416,
      "step": 210
    },
    {
      "epoch": 0.9035193362772648,
      "grad_norm": 0.30461645126342773,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.0459,
      "step": 211
    },
    {
      "epoch": 0.9078014184397163,
      "grad_norm": 0.3148707449436188,
      "learning_rate": 1.1274509803921569e-05,
      "loss": 0.0448,
      "step": 212
    },
    {
      "epoch": 0.9120835006021678,
      "grad_norm": 0.29912617802619934,
      "learning_rate": 1.0784313725490197e-05,
      "loss": 0.0391,
      "step": 213
    },
    {
      "epoch": 0.9163655827646193,
      "grad_norm": 0.30762436985969543,
      "learning_rate": 1.0294117647058824e-05,
      "loss": 0.0475,
      "step": 214
    },
    {
      "epoch": 0.9206476649270708,
      "grad_norm": 0.27570030093193054,
      "learning_rate": 9.803921568627451e-06,
      "loss": 0.0371,
      "step": 215
    },
    {
      "epoch": 0.9249297470895222,
      "grad_norm": 0.26668745279312134,
      "learning_rate": 9.31372549019608e-06,
      "loss": 0.0532,
      "step": 216
    },
    {
      "epoch": 0.9292118292519738,
      "grad_norm": 0.20003274083137512,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.0328,
      "step": 217
    },
    {
      "epoch": 0.9334939114144253,
      "grad_norm": 0.23162154853343964,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.0348,
      "step": 218
    },
    {
      "epoch": 0.9377759935768768,
      "grad_norm": 0.28092819452285767,
      "learning_rate": 7.84313725490196e-06,
      "loss": 0.041,
      "step": 219
    },
    {
      "epoch": 0.9420580757393282,
      "grad_norm": 0.20697717368602753,
      "learning_rate": 7.3529411764705884e-06,
      "loss": 0.0344,
      "step": 220
    },
    {
      "epoch": 0.9463401579017797,
      "grad_norm": 0.2537749111652374,
      "learning_rate": 6.862745098039216e-06,
      "loss": 0.0364,
      "step": 221
    },
    {
      "epoch": 0.9506222400642312,
      "grad_norm": 0.23425333201885223,
      "learning_rate": 6.372549019607843e-06,
      "loss": 0.0391,
      "step": 222
    },
    {
      "epoch": 0.9549043222266828,
      "grad_norm": 0.25059816241264343,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.0506,
      "step": 223
    },
    {
      "epoch": 0.9591864043891342,
      "grad_norm": 0.2436205893754959,
      "learning_rate": 5.392156862745099e-06,
      "loss": 0.0418,
      "step": 224
    },
    {
      "epoch": 0.9634684865515857,
      "grad_norm": 0.255195289850235,
      "learning_rate": 4.901960784313726e-06,
      "loss": 0.0412,
      "step": 225
    },
    {
      "epoch": 0.9677505687140372,
      "grad_norm": 0.27884790301322937,
      "learning_rate": 4.411764705882353e-06,
      "loss": 0.0497,
      "step": 226
    },
    {
      "epoch": 0.9720326508764887,
      "grad_norm": 0.25363633036613464,
      "learning_rate": 3.92156862745098e-06,
      "loss": 0.0408,
      "step": 227
    },
    {
      "epoch": 0.9763147330389402,
      "grad_norm": 0.30385518074035645,
      "learning_rate": 3.431372549019608e-06,
      "loss": 0.0552,
      "step": 228
    },
    {
      "epoch": 0.9805968152013916,
      "grad_norm": 0.27926793694496155,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 0.036,
      "step": 229
    },
    {
      "epoch": 0.9848788973638432,
      "grad_norm": 0.236245259642601,
      "learning_rate": 2.450980392156863e-06,
      "loss": 0.029,
      "step": 230
    },
    {
      "epoch": 0.9891609795262947,
      "grad_norm": 0.25143948197364807,
      "learning_rate": 1.96078431372549e-06,
      "loss": 0.0303,
      "step": 231
    },
    {
      "epoch": 0.9934430616887462,
      "grad_norm": 0.29792022705078125,
      "learning_rate": 1.4705882352941177e-06,
      "loss": 0.0498,
      "step": 232
    },
    {
      "epoch": 0.9977251438511976,
      "grad_norm": 0.32305002212524414,
      "learning_rate": 9.80392156862745e-07,
      "loss": 0.0375,
      "step": 233
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.4602526128292084,
      "learning_rate": 4.901960784313725e-07,
      "loss": 0.0426,
      "step": 234
    }
  ],
  "logging_steps": 1,
  "max_steps": 234,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.927681385546916e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}