{
  "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": 2.5088911056518555,
      "learning_rate": 0.0,
      "loss": 1.602,
      "step": 1
    },
    {
      "epoch": 0.008564164324902984,
      "grad_norm": 1.8938051462173462,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.2932,
      "step": 2
    },
    {
      "epoch": 0.012846246487354477,
      "grad_norm": 2.1013832092285156,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.4833,
      "step": 3
    },
    {
      "epoch": 0.017128328649805968,
      "grad_norm": 2.3183069229125977,
      "learning_rate": 1e-05,
      "loss": 1.4841,
      "step": 4
    },
    {
      "epoch": 0.02141041081225746,
      "grad_norm": 2.284543514251709,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 1.5066,
      "step": 5
    },
    {
      "epoch": 0.025692492974708953,
      "grad_norm": 2.2618205547332764,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.509,
      "step": 6
    },
    {
      "epoch": 0.029974575137160443,
      "grad_norm": 2.1478371620178223,
      "learning_rate": 2e-05,
      "loss": 1.4126,
      "step": 7
    },
    {
      "epoch": 0.034256657299611935,
      "grad_norm": 1.6768583059310913,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 1.127,
      "step": 8
    },
    {
      "epoch": 0.03853873946206343,
      "grad_norm": 1.7240148782730103,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 1.1236,
      "step": 9
    },
    {
      "epoch": 0.04282082162451492,
      "grad_norm": 1.3401944637298584,
      "learning_rate": 3e-05,
      "loss": 0.9918,
      "step": 10
    },
    {
      "epoch": 0.04710290378696641,
      "grad_norm": 1.072590708732605,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.8782,
      "step": 11
    },
    {
      "epoch": 0.051384985949417906,
      "grad_norm": 0.8676239848136902,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.7478,
      "step": 12
    },
    {
      "epoch": 0.055667068111869396,
      "grad_norm": 0.7675440907478333,
      "learning_rate": 4e-05,
      "loss": 0.7276,
      "step": 13
    },
    {
      "epoch": 0.059949150274320885,
      "grad_norm": 0.7409543991088867,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.6637,
      "step": 14
    },
    {
      "epoch": 0.06423123243677238,
      "grad_norm": 0.5381997227668762,
      "learning_rate": 4.666666666666667e-05,
      "loss": 0.633,
      "step": 15
    },
    {
      "epoch": 0.06851331459922387,
      "grad_norm": 0.4906686246395111,
      "learning_rate": 5e-05,
      "loss": 0.5315,
      "step": 16
    },
    {
      "epoch": 0.07279539676167536,
      "grad_norm": 0.5198220610618591,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.5256,
      "step": 17
    },
    {
      "epoch": 0.07707747892412686,
      "grad_norm": 0.4853769540786743,
      "learning_rate": 5.666666666666667e-05,
      "loss": 0.5323,
      "step": 18
    },
    {
      "epoch": 0.08135956108657835,
      "grad_norm": 0.36931294202804565,
      "learning_rate": 6e-05,
      "loss": 0.4003,
      "step": 19
    },
    {
      "epoch": 0.08564164324902984,
      "grad_norm": 0.4454424977302551,
      "learning_rate": 6.333333333333333e-05,
      "loss": 0.4547,
      "step": 20
    },
    {
      "epoch": 0.08992372541148133,
      "grad_norm": 0.3773050904273987,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.4051,
      "step": 21
    },
    {
      "epoch": 0.09420580757393282,
      "grad_norm": 0.3983994722366333,
      "learning_rate": 7e-05,
      "loss": 0.3534,
      "step": 22
    },
    {
      "epoch": 0.09848788973638432,
      "grad_norm": 0.3229769766330719,
      "learning_rate": 7.333333333333333e-05,
      "loss": 0.3644,
      "step": 23
    },
    {
      "epoch": 0.10276997189883581,
      "grad_norm": 0.27993354201316833,
      "learning_rate": 7.666666666666667e-05,
      "loss": 0.3343,
      "step": 24
    },
    {
      "epoch": 0.1070520540612873,
      "grad_norm": 0.2869669497013092,
      "learning_rate": 8e-05,
      "loss": 0.303,
      "step": 25
    },
    {
      "epoch": 0.11133413622373879,
      "grad_norm": 0.26860982179641724,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.3018,
      "step": 26
    },
    {
      "epoch": 0.11561621838619028,
      "grad_norm": 0.29772505164146423,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.3158,
      "step": 27
    },
    {
      "epoch": 0.11989830054864177,
      "grad_norm": 0.31327569484710693,
      "learning_rate": 9e-05,
      "loss": 0.3588,
      "step": 28
    },
    {
      "epoch": 0.12418038271109327,
      "grad_norm": 0.2718559205532074,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.3219,
      "step": 29
    },
    {
      "epoch": 0.12846246487354476,
      "grad_norm": 0.3206578493118286,
      "learning_rate": 9.666666666666667e-05,
      "loss": 0.3552,
      "step": 30
    },
    {
      "epoch": 0.13274454703599625,
      "grad_norm": 0.30618807673454285,
      "learning_rate": 0.0001,
      "loss": 0.3667,
      "step": 31
    },
    {
      "epoch": 0.13702662919844774,
      "grad_norm": 0.2761961817741394,
      "learning_rate": 9.950980392156864e-05,
      "loss": 0.3217,
      "step": 32
    },
    {
      "epoch": 0.14130871136089923,
      "grad_norm": 0.3033706545829773,
      "learning_rate": 9.901960784313727e-05,
      "loss": 0.301,
      "step": 33
    },
    {
      "epoch": 0.14559079352335072,
      "grad_norm": 0.2686208188533783,
      "learning_rate": 9.852941176470589e-05,
      "loss": 0.3279,
      "step": 34
    },
    {
      "epoch": 0.1498728756858022,
      "grad_norm": 0.22647158801555634,
      "learning_rate": 9.80392156862745e-05,
      "loss": 0.3133,
      "step": 35
    },
    {
      "epoch": 0.15415495784825373,
      "grad_norm": 0.2989174425601959,
      "learning_rate": 9.754901960784314e-05,
      "loss": 0.2982,
      "step": 36
    },
    {
      "epoch": 0.15843704001070522,
      "grad_norm": 0.27390947937965393,
      "learning_rate": 9.705882352941177e-05,
      "loss": 0.3133,
      "step": 37
    },
    {
      "epoch": 0.1627191221731567,
      "grad_norm": 0.22332409024238586,
      "learning_rate": 9.65686274509804e-05,
      "loss": 0.2802,
      "step": 38
    },
    {
      "epoch": 0.1670012043356082,
      "grad_norm": 0.21666789054870605,
      "learning_rate": 9.607843137254903e-05,
      "loss": 0.2921,
      "step": 39
    },
    {
      "epoch": 0.17128328649805968,
      "grad_norm": 0.2647346556186676,
      "learning_rate": 9.558823529411765e-05,
      "loss": 0.3016,
      "step": 40
    },
    {
      "epoch": 0.17556536866051117,
      "grad_norm": 0.23542706668376923,
      "learning_rate": 9.509803921568627e-05,
      "loss": 0.269,
      "step": 41
    },
    {
      "epoch": 0.17984745082296266,
      "grad_norm": 0.24950803816318512,
      "learning_rate": 9.46078431372549e-05,
      "loss": 0.2803,
      "step": 42
    },
    {
      "epoch": 0.18412953298541415,
      "grad_norm": 0.22942230105400085,
      "learning_rate": 9.411764705882353e-05,
      "loss": 0.2889,
      "step": 43
    },
    {
      "epoch": 0.18841161514786564,
      "grad_norm": 0.22736386954784393,
      "learning_rate": 9.362745098039217e-05,
      "loss": 0.2712,
      "step": 44
    },
    {
      "epoch": 0.19269369731031713,
      "grad_norm": 0.24545833468437195,
      "learning_rate": 9.313725490196079e-05,
      "loss": 0.2489,
      "step": 45
    },
    {
      "epoch": 0.19697577947276865,
      "grad_norm": 0.2351815104484558,
      "learning_rate": 9.264705882352942e-05,
      "loss": 0.2637,
      "step": 46
    },
    {
      "epoch": 0.20125786163522014,
      "grad_norm": 0.24491284787654877,
      "learning_rate": 9.215686274509804e-05,
      "loss": 0.2789,
      "step": 47
    },
    {
      "epoch": 0.20553994379767163,
      "grad_norm": 0.26231712102890015,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.2765,
      "step": 48
    },
    {
      "epoch": 0.20982202596012312,
      "grad_norm": 0.29606589674949646,
      "learning_rate": 9.11764705882353e-05,
      "loss": 0.3004,
      "step": 49
    },
    {
      "epoch": 0.2141041081225746,
      "grad_norm": 0.27311965823173523,
      "learning_rate": 9.068627450980393e-05,
      "loss": 0.2955,
      "step": 50
    },
    {
      "epoch": 0.2183861902850261,
      "grad_norm": 0.24474990367889404,
      "learning_rate": 9.019607843137255e-05,
      "loss": 0.2772,
      "step": 51
    },
    {
      "epoch": 0.22266827244747758,
      "grad_norm": 0.2573692202568054,
      "learning_rate": 8.970588235294118e-05,
      "loss": 0.2791,
      "step": 52
    },
    {
      "epoch": 0.22695035460992907,
      "grad_norm": 0.24064259231090546,
      "learning_rate": 8.921568627450981e-05,
      "loss": 0.2674,
      "step": 53
    },
    {
      "epoch": 0.23123243677238056,
      "grad_norm": 0.2550707161426544,
      "learning_rate": 8.872549019607843e-05,
      "loss": 0.3038,
      "step": 54
    },
    {
      "epoch": 0.23551451893483205,
      "grad_norm": 0.2474234700202942,
      "learning_rate": 8.823529411764706e-05,
      "loss": 0.2517,
      "step": 55
    },
    {
      "epoch": 0.23979660109728354,
      "grad_norm": 0.2420043647289276,
      "learning_rate": 8.774509803921568e-05,
      "loss": 0.2773,
      "step": 56
    },
    {
      "epoch": 0.24407868325973506,
      "grad_norm": 0.2720290422439575,
      "learning_rate": 8.725490196078432e-05,
      "loss": 0.275,
      "step": 57
    },
    {
      "epoch": 0.24836076542218655,
      "grad_norm": 0.24702514708042145,
      "learning_rate": 8.676470588235295e-05,
      "loss": 0.3,
      "step": 58
    },
    {
      "epoch": 0.252642847584638,
      "grad_norm": 0.2610642910003662,
      "learning_rate": 8.627450980392158e-05,
      "loss": 0.2686,
      "step": 59
    },
    {
      "epoch": 0.2569249297470895,
      "grad_norm": 0.23535649478435516,
      "learning_rate": 8.57843137254902e-05,
      "loss": 0.2807,
      "step": 60
    },
    {
      "epoch": 0.261207011909541,
      "grad_norm": 0.25134730339050293,
      "learning_rate": 8.529411764705883e-05,
      "loss": 0.2742,
      "step": 61
    },
    {
      "epoch": 0.2654890940719925,
      "grad_norm": 0.22608330845832825,
      "learning_rate": 8.480392156862745e-05,
      "loss": 0.2722,
      "step": 62
    },
    {
      "epoch": 0.269771176234444,
      "grad_norm": 0.2923852205276489,
      "learning_rate": 8.431372549019608e-05,
      "loss": 0.2899,
      "step": 63
    },
    {
      "epoch": 0.2740532583968955,
      "grad_norm": 0.24233442544937134,
      "learning_rate": 8.382352941176471e-05,
      "loss": 0.2386,
      "step": 64
    },
    {
      "epoch": 0.278335340559347,
      "grad_norm": 0.24746356904506683,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.2638,
      "step": 65
    },
    {
      "epoch": 0.28261742272179846,
      "grad_norm": 0.22780101001262665,
      "learning_rate": 8.284313725490198e-05,
      "loss": 0.2502,
      "step": 66
    },
    {
      "epoch": 0.28689950488425,
      "grad_norm": 0.26735028624534607,
      "learning_rate": 8.23529411764706e-05,
      "loss": 0.3082,
      "step": 67
    },
    {
      "epoch": 0.29118158704670144,
      "grad_norm": 0.2815140187740326,
      "learning_rate": 8.186274509803921e-05,
      "loss": 0.3034,
      "step": 68
    },
    {
      "epoch": 0.29546366920915296,
      "grad_norm": 0.23998597264289856,
      "learning_rate": 8.137254901960785e-05,
      "loss": 0.2663,
      "step": 69
    },
    {
      "epoch": 0.2997457513716044,
      "grad_norm": 0.25277936458587646,
      "learning_rate": 8.088235294117648e-05,
      "loss": 0.2692,
      "step": 70
    },
    {
      "epoch": 0.30402783353405594,
      "grad_norm": 0.26299694180488586,
      "learning_rate": 8.039215686274511e-05,
      "loss": 0.2957,
      "step": 71
    },
    {
      "epoch": 0.30830991569650745,
      "grad_norm": 0.2495986670255661,
      "learning_rate": 7.990196078431373e-05,
      "loss": 0.258,
      "step": 72
    },
    {
      "epoch": 0.3125919978589589,
      "grad_norm": 0.2295389175415039,
      "learning_rate": 7.941176470588235e-05,
      "loss": 0.2664,
      "step": 73
    },
    {
      "epoch": 0.31687408002141043,
      "grad_norm": 0.254192590713501,
      "learning_rate": 7.892156862745098e-05,
      "loss": 0.2459,
      "step": 74
    },
    {
      "epoch": 0.3211561621838619,
      "grad_norm": 0.2466706931591034,
      "learning_rate": 7.843137254901961e-05,
      "loss": 0.2395,
      "step": 75
    },
    {
      "epoch": 0.3254382443463134,
      "grad_norm": 0.26901692152023315,
      "learning_rate": 7.794117647058824e-05,
      "loss": 0.3027,
      "step": 76
    },
    {
      "epoch": 0.32972032650876487,
      "grad_norm": 0.2709638774394989,
      "learning_rate": 7.745098039215687e-05,
      "loss": 0.2532,
      "step": 77
    },
    {
      "epoch": 0.3340024086712164,
      "grad_norm": 0.22372817993164062,
      "learning_rate": 7.696078431372549e-05,
      "loss": 0.2456,
      "step": 78
    },
    {
      "epoch": 0.33828449083366785,
      "grad_norm": 0.26552456617355347,
      "learning_rate": 7.647058823529411e-05,
      "loss": 0.2892,
      "step": 79
    },
    {
      "epoch": 0.34256657299611937,
      "grad_norm": 0.26157811284065247,
      "learning_rate": 7.598039215686274e-05,
      "loss": 0.2446,
      "step": 80
    },
    {
      "epoch": 0.34684865515857083,
      "grad_norm": 0.2531943917274475,
      "learning_rate": 7.549019607843137e-05,
      "loss": 0.266,
      "step": 81
    },
    {
      "epoch": 0.35113073732102235,
      "grad_norm": 0.260826051235199,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.2621,
      "step": 82
    },
    {
      "epoch": 0.35541281948347386,
      "grad_norm": 0.2922424376010895,
      "learning_rate": 7.450980392156864e-05,
      "loss": 0.2687,
      "step": 83
    },
    {
      "epoch": 0.3596949016459253,
      "grad_norm": 0.22803695499897003,
      "learning_rate": 7.401960784313726e-05,
      "loss": 0.2019,
      "step": 84
    },
    {
      "epoch": 0.36397698380837684,
      "grad_norm": 0.292816698551178,
      "learning_rate": 7.352941176470589e-05,
      "loss": 0.248,
      "step": 85
    },
    {
      "epoch": 0.3682590659708283,
      "grad_norm": 0.26142618060112,
      "learning_rate": 7.303921568627451e-05,
      "loss": 0.2444,
      "step": 86
    },
    {
      "epoch": 0.3725411481332798,
      "grad_norm": 0.2799445390701294,
      "learning_rate": 7.254901960784314e-05,
      "loss": 0.2423,
      "step": 87
    },
    {
      "epoch": 0.3768232302957313,
      "grad_norm": 0.2834770679473877,
      "learning_rate": 7.205882352941177e-05,
      "loss": 0.2921,
      "step": 88
    },
    {
      "epoch": 0.3811053124581828,
      "grad_norm": 0.2628571093082428,
      "learning_rate": 7.156862745098039e-05,
      "loss": 0.2621,
      "step": 89
    },
    {
      "epoch": 0.38538739462063426,
      "grad_norm": 0.284971684217453,
      "learning_rate": 7.107843137254902e-05,
      "loss": 0.2596,
      "step": 90
    },
    {
      "epoch": 0.3896694767830858,
      "grad_norm": 0.2779354751110077,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.2581,
      "step": 91
    },
    {
      "epoch": 0.3939515589455373,
      "grad_norm": 0.2709386348724365,
      "learning_rate": 7.009803921568627e-05,
      "loss": 0.2344,
      "step": 92
    },
    {
      "epoch": 0.39823364110798876,
      "grad_norm": 0.27376848459243774,
      "learning_rate": 6.96078431372549e-05,
      "loss": 0.2531,
      "step": 93
    },
    {
      "epoch": 0.4025157232704403,
      "grad_norm": 0.24732176959514618,
      "learning_rate": 6.911764705882354e-05,
      "loss": 0.2263,
      "step": 94
    },
    {
      "epoch": 0.40679780543289173,
      "grad_norm": 0.27054837346076965,
      "learning_rate": 6.862745098039216e-05,
      "loss": 0.2693,
      "step": 95
    },
    {
      "epoch": 0.41107988759534325,
      "grad_norm": 0.2752648591995239,
      "learning_rate": 6.813725490196079e-05,
      "loss": 0.2812,
      "step": 96
    },
    {
      "epoch": 0.4153619697577947,
      "grad_norm": 0.29322928190231323,
      "learning_rate": 6.764705882352942e-05,
      "loss": 0.2584,
      "step": 97
    },
    {
      "epoch": 0.41964405192024623,
      "grad_norm": 0.2726430892944336,
      "learning_rate": 6.715686274509804e-05,
      "loss": 0.2738,
      "step": 98
    },
    {
      "epoch": 0.4239261340826977,
      "grad_norm": 0.25086814165115356,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.2401,
      "step": 99
    },
    {
      "epoch": 0.4282082162451492,
      "grad_norm": 0.25943422317504883,
      "learning_rate": 6.61764705882353e-05,
      "loss": 0.2747,
      "step": 100
    },
    {
      "epoch": 0.43249029840760067,
      "grad_norm": 0.2715889811515808,
      "learning_rate": 6.568627450980392e-05,
      "loss": 0.2465,
      "step": 101
    },
    {
      "epoch": 0.4367723805700522,
      "grad_norm": 0.28668537735939026,
      "learning_rate": 6.519607843137255e-05,
      "loss": 0.2671,
      "step": 102
    },
    {
      "epoch": 0.4410544627325037,
      "grad_norm": 0.2939082086086273,
      "learning_rate": 6.470588235294118e-05,
      "loss": 0.2606,
      "step": 103
    },
    {
      "epoch": 0.44533654489495517,
      "grad_norm": 0.29065361618995667,
      "learning_rate": 6.421568627450982e-05,
      "loss": 0.252,
      "step": 104
    },
    {
      "epoch": 0.4496186270574067,
      "grad_norm": 0.28489866852760315,
      "learning_rate": 6.372549019607843e-05,
      "loss": 0.2576,
      "step": 105
    },
    {
      "epoch": 0.45390070921985815,
      "grad_norm": 0.3433012068271637,
      "learning_rate": 6.323529411764705e-05,
      "loss": 0.3029,
      "step": 106
    },
    {
      "epoch": 0.45818279138230966,
      "grad_norm": 0.283680260181427,
      "learning_rate": 6.274509803921569e-05,
      "loss": 0.2023,
      "step": 107
    },
    {
      "epoch": 0.4624648735447611,
      "grad_norm": 0.2868785858154297,
      "learning_rate": 6.225490196078432e-05,
      "loss": 0.2439,
      "step": 108
    },
    {
      "epoch": 0.46674695570721264,
      "grad_norm": 0.28326234221458435,
      "learning_rate": 6.176470588235295e-05,
      "loss": 0.2913,
      "step": 109
    },
    {
      "epoch": 0.4710290378696641,
      "grad_norm": 0.26688653230667114,
      "learning_rate": 6.127450980392158e-05,
      "loss": 0.2482,
      "step": 110
    },
    {
      "epoch": 0.4753111200321156,
      "grad_norm": 0.23991777002811432,
      "learning_rate": 6.078431372549019e-05,
      "loss": 0.2305,
      "step": 111
    },
    {
      "epoch": 0.4795932021945671,
      "grad_norm": 0.26562249660491943,
      "learning_rate": 6.0294117647058825e-05,
      "loss": 0.2378,
      "step": 112
    },
    {
      "epoch": 0.4838752843570186,
      "grad_norm": 0.2699893116950989,
      "learning_rate": 5.980392156862745e-05,
      "loss": 0.2396,
      "step": 113
    },
    {
      "epoch": 0.4881573665194701,
      "grad_norm": 0.2860610783100128,
      "learning_rate": 5.931372549019608e-05,
      "loss": 0.2459,
      "step": 114
    },
    {
      "epoch": 0.4924394486819216,
      "grad_norm": 0.27508506178855896,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.2636,
      "step": 115
    },
    {
      "epoch": 0.4967215308443731,
      "grad_norm": 0.29485681653022766,
      "learning_rate": 5.833333333333334e-05,
      "loss": 0.2317,
      "step": 116
    },
    {
      "epoch": 0.5010036130068246,
      "grad_norm": 0.2937762141227722,
      "learning_rate": 5.784313725490197e-05,
      "loss": 0.2473,
      "step": 117
    },
    {
      "epoch": 0.505285695169276,
      "grad_norm": 0.31636714935302734,
      "learning_rate": 5.735294117647059e-05,
      "loss": 0.2752,
      "step": 118
    },
    {
      "epoch": 0.5095677773317275,
      "grad_norm": 0.2664767801761627,
      "learning_rate": 5.6862745098039215e-05,
      "loss": 0.2285,
      "step": 119
    },
    {
      "epoch": 0.513849859494179,
      "grad_norm": 0.251086950302124,
      "learning_rate": 5.637254901960785e-05,
      "loss": 0.2388,
      "step": 120
    },
    {
      "epoch": 0.5181319416566306,
      "grad_norm": 0.29327136278152466,
      "learning_rate": 5.588235294117647e-05,
      "loss": 0.2528,
      "step": 121
    },
    {
      "epoch": 0.522414023819082,
      "grad_norm": 0.2633054256439209,
      "learning_rate": 5.5392156862745104e-05,
      "loss": 0.2405,
      "step": 122
    },
    {
      "epoch": 0.5266961059815335,
      "grad_norm": 0.2686323821544647,
      "learning_rate": 5.490196078431373e-05,
      "loss": 0.2394,
      "step": 123
    },
    {
      "epoch": 0.530978188143985,
      "grad_norm": 0.31154799461364746,
      "learning_rate": 5.441176470588235e-05,
      "loss": 0.2939,
      "step": 124
    },
    {
      "epoch": 0.5352602703064365,
      "grad_norm": 0.29867416620254517,
      "learning_rate": 5.392156862745098e-05,
      "loss": 0.2271,
      "step": 125
    },
    {
      "epoch": 0.539542352468888,
      "grad_norm": 0.2981470227241516,
      "learning_rate": 5.343137254901961e-05,
      "loss": 0.2697,
      "step": 126
    },
    {
      "epoch": 0.5438244346313394,
      "grad_norm": 0.2520562410354614,
      "learning_rate": 5.294117647058824e-05,
      "loss": 0.2272,
      "step": 127
    },
    {
      "epoch": 0.548106516793791,
      "grad_norm": 0.2620241940021515,
      "learning_rate": 5.245098039215687e-05,
      "loss": 0.2369,
      "step": 128
    },
    {
      "epoch": 0.5523885989562425,
      "grad_norm": 0.2934142053127289,
      "learning_rate": 5.1960784313725495e-05,
      "loss": 0.2464,
      "step": 129
    },
    {
      "epoch": 0.556670681118694,
      "grad_norm": 0.283558189868927,
      "learning_rate": 5.147058823529411e-05,
      "loss": 0.2899,
      "step": 130
    },
    {
      "epoch": 0.5609527632811454,
      "grad_norm": 0.3288978934288025,
      "learning_rate": 5.0980392156862745e-05,
      "loss": 0.243,
      "step": 131
    },
    {
      "epoch": 0.5652348454435969,
      "grad_norm": 0.26702696084976196,
      "learning_rate": 5.049019607843137e-05,
      "loss": 0.2373,
      "step": 132
    },
    {
      "epoch": 0.5695169276060484,
      "grad_norm": 0.30316781997680664,
      "learning_rate": 5e-05,
      "loss": 0.2188,
      "step": 133
    },
    {
      "epoch": 0.5737990097685,
      "grad_norm": 0.30304622650146484,
      "learning_rate": 4.9509803921568634e-05,
      "loss": 0.2901,
      "step": 134
    },
    {
      "epoch": 0.5780810919309515,
      "grad_norm": 0.28154560923576355,
      "learning_rate": 4.901960784313725e-05,
      "loss": 0.2262,
      "step": 135
    },
    {
      "epoch": 0.5823631740934029,
      "grad_norm": 0.29088521003723145,
      "learning_rate": 4.8529411764705885e-05,
      "loss": 0.2471,
      "step": 136
    },
    {
      "epoch": 0.5866452562558544,
      "grad_norm": 0.2862875461578369,
      "learning_rate": 4.803921568627452e-05,
      "loss": 0.247,
      "step": 137
    },
    {
      "epoch": 0.5909273384183059,
      "grad_norm": 0.31327253580093384,
      "learning_rate": 4.7549019607843135e-05,
      "loss": 0.2576,
      "step": 138
    },
    {
      "epoch": 0.5952094205807574,
      "grad_norm": 0.29111480712890625,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.2352,
      "step": 139
    },
    {
      "epoch": 0.5994915027432088,
      "grad_norm": 0.2540854215621948,
      "learning_rate": 4.656862745098039e-05,
      "loss": 0.2131,
      "step": 140
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.28845465183258057,
      "learning_rate": 4.607843137254902e-05,
      "loss": 0.2473,
      "step": 141
    },
    {
      "epoch": 0.6080556670681119,
      "grad_norm": 0.2802583873271942,
      "learning_rate": 4.558823529411765e-05,
      "loss": 0.2147,
      "step": 142
    },
    {
      "epoch": 0.6123377492305634,
      "grad_norm": 0.2936670482158661,
      "learning_rate": 4.5098039215686275e-05,
      "loss": 0.2471,
      "step": 143
    },
    {
      "epoch": 0.6166198313930149,
      "grad_norm": 0.33588525652885437,
      "learning_rate": 4.460784313725491e-05,
      "loss": 0.2661,
      "step": 144
    },
    {
      "epoch": 0.6209019135554663,
      "grad_norm": 0.29438409209251404,
      "learning_rate": 4.411764705882353e-05,
      "loss": 0.2441,
      "step": 145
    },
    {
      "epoch": 0.6251839957179178,
      "grad_norm": 0.3029739260673523,
      "learning_rate": 4.362745098039216e-05,
      "loss": 0.2506,
      "step": 146
    },
    {
      "epoch": 0.6294660778803693,
      "grad_norm": 0.28387150168418884,
      "learning_rate": 4.313725490196079e-05,
      "loss": 0.2624,
      "step": 147
    },
    {
      "epoch": 0.6337481600428209,
      "grad_norm": 0.29315584897994995,
      "learning_rate": 4.2647058823529415e-05,
      "loss": 0.2523,
      "step": 148
    },
    {
      "epoch": 0.6380302422052723,
      "grad_norm": 0.28650233149528503,
      "learning_rate": 4.215686274509804e-05,
      "loss": 0.2191,
      "step": 149
    },
    {
      "epoch": 0.6423123243677238,
      "grad_norm": 0.28433752059936523,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.2155,
      "step": 150
    },
    {
      "epoch": 0.6465944065301753,
      "grad_norm": 0.31756916642189026,
      "learning_rate": 4.11764705882353e-05,
      "loss": 0.2465,
      "step": 151
    },
    {
      "epoch": 0.6508764886926268,
      "grad_norm": 0.31745150685310364,
      "learning_rate": 4.068627450980392e-05,
      "loss": 0.2316,
      "step": 152
    },
    {
      "epoch": 0.6551585708550783,
      "grad_norm": 0.30417391657829285,
      "learning_rate": 4.0196078431372555e-05,
      "loss": 0.2314,
      "step": 153
    },
    {
      "epoch": 0.6594406530175297,
      "grad_norm": 0.3365858197212219,
      "learning_rate": 3.970588235294117e-05,
      "loss": 0.2677,
      "step": 154
    },
    {
      "epoch": 0.6637227351799813,
      "grad_norm": 0.3298024833202362,
      "learning_rate": 3.9215686274509805e-05,
      "loss": 0.2352,
      "step": 155
    },
    {
      "epoch": 0.6680048173424328,
      "grad_norm": 0.3242587447166443,
      "learning_rate": 3.872549019607844e-05,
      "loss": 0.249,
      "step": 156
    },
    {
      "epoch": 0.6722868995048843,
      "grad_norm": 0.31644657254219055,
      "learning_rate": 3.8235294117647055e-05,
      "loss": 0.2915,
      "step": 157
    },
    {
      "epoch": 0.6765689816673357,
      "grad_norm": 0.3301697373390198,
      "learning_rate": 3.774509803921569e-05,
      "loss": 0.2582,
      "step": 158
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.300255686044693,
      "learning_rate": 3.725490196078432e-05,
      "loss": 0.2591,
      "step": 159
    },
    {
      "epoch": 0.6851331459922387,
      "grad_norm": 0.29023420810699463,
      "learning_rate": 3.6764705882352945e-05,
      "loss": 0.2306,
      "step": 160
    },
    {
      "epoch": 0.6894152281546903,
      "grad_norm": 0.27386027574539185,
      "learning_rate": 3.627450980392157e-05,
      "loss": 0.1964,
      "step": 161
    },
    {
      "epoch": 0.6936973103171417,
      "grad_norm": 0.29281091690063477,
      "learning_rate": 3.5784313725490195e-05,
      "loss": 0.2865,
      "step": 162
    },
    {
      "epoch": 0.6979793924795932,
      "grad_norm": 0.24590419232845306,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.2086,
      "step": 163
    },
    {
      "epoch": 0.7022614746420447,
      "grad_norm": 0.26177820563316345,
      "learning_rate": 3.480392156862745e-05,
      "loss": 0.2023,
      "step": 164
    },
    {
      "epoch": 0.7065435568044962,
      "grad_norm": 0.29054680466651917,
      "learning_rate": 3.431372549019608e-05,
      "loss": 0.2307,
      "step": 165
    },
    {
      "epoch": 0.7108256389669477,
      "grad_norm": 0.30803316831588745,
      "learning_rate": 3.382352941176471e-05,
      "loss": 0.2301,
      "step": 166
    },
    {
      "epoch": 0.7151077211293991,
      "grad_norm": 0.30045220255851746,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.2558,
      "step": 167
    },
    {
      "epoch": 0.7193898032918506,
      "grad_norm": 0.2771556079387665,
      "learning_rate": 3.284313725490196e-05,
      "loss": 0.2287,
      "step": 168
    },
    {
      "epoch": 0.7236718854543022,
      "grad_norm": 0.31324565410614014,
      "learning_rate": 3.235294117647059e-05,
      "loss": 0.2033,
      "step": 169
    },
    {
      "epoch": 0.7279539676167537,
      "grad_norm": 0.2909330725669861,
      "learning_rate": 3.186274509803922e-05,
      "loss": 0.239,
      "step": 170
    },
    {
      "epoch": 0.7322360497792051,
      "grad_norm": 0.3129582703113556,
      "learning_rate": 3.137254901960784e-05,
      "loss": 0.248,
      "step": 171
    },
    {
      "epoch": 0.7365181319416566,
      "grad_norm": 0.2871990203857422,
      "learning_rate": 3.0882352941176475e-05,
      "loss": 0.2207,
      "step": 172
    },
    {
      "epoch": 0.7408002141041081,
      "grad_norm": 0.348622590303421,
      "learning_rate": 3.0392156862745097e-05,
      "loss": 0.2737,
      "step": 173
    },
    {
      "epoch": 0.7450822962665596,
      "grad_norm": 0.31700730323791504,
      "learning_rate": 2.9901960784313725e-05,
      "loss": 0.2241,
      "step": 174
    },
    {
      "epoch": 0.7493643784290112,
      "grad_norm": 0.283666729927063,
      "learning_rate": 2.9411764705882354e-05,
      "loss": 0.2328,
      "step": 175
    },
    {
      "epoch": 0.7536464605914626,
      "grad_norm": 0.3109316825866699,
      "learning_rate": 2.8921568627450986e-05,
      "loss": 0.2515,
      "step": 176
    },
    {
      "epoch": 0.7579285427539141,
      "grad_norm": 0.2863878309726715,
      "learning_rate": 2.8431372549019608e-05,
      "loss": 0.2307,
      "step": 177
    },
    {
      "epoch": 0.7622106249163656,
      "grad_norm": 0.2934282720088959,
      "learning_rate": 2.7941176470588236e-05,
      "loss": 0.2276,
      "step": 178
    },
    {
      "epoch": 0.7664927070788171,
      "grad_norm": 0.28136464953422546,
      "learning_rate": 2.7450980392156865e-05,
      "loss": 0.2223,
      "step": 179
    },
    {
      "epoch": 0.7707747892412685,
      "grad_norm": 0.29069915413856506,
      "learning_rate": 2.696078431372549e-05,
      "loss": 0.2068,
      "step": 180
    },
    {
      "epoch": 0.77505687140372,
      "grad_norm": 0.3577599823474884,
      "learning_rate": 2.647058823529412e-05,
      "loss": 0.2688,
      "step": 181
    },
    {
      "epoch": 0.7793389535661716,
      "grad_norm": 0.29474157094955444,
      "learning_rate": 2.5980392156862747e-05,
      "loss": 0.2173,
      "step": 182
    },
    {
      "epoch": 0.7836210357286231,
      "grad_norm": 0.3038179278373718,
      "learning_rate": 2.5490196078431373e-05,
      "loss": 0.25,
      "step": 183
    },
    {
      "epoch": 0.7879031178910746,
      "grad_norm": 0.35067787766456604,
      "learning_rate": 2.5e-05,
      "loss": 0.2567,
      "step": 184
    },
    {
      "epoch": 0.792185200053526,
      "grad_norm": 0.30648499727249146,
      "learning_rate": 2.4509803921568626e-05,
      "loss": 0.2676,
      "step": 185
    },
    {
      "epoch": 0.7964672822159775,
      "grad_norm": 0.3093171715736389,
      "learning_rate": 2.401960784313726e-05,
      "loss": 0.2658,
      "step": 186
    },
    {
      "epoch": 0.800749364378429,
      "grad_norm": 0.3015124499797821,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.2332,
      "step": 187
    },
    {
      "epoch": 0.8050314465408805,
      "grad_norm": 0.29681718349456787,
      "learning_rate": 2.303921568627451e-05,
      "loss": 0.2472,
      "step": 188
    },
    {
      "epoch": 0.809313528703332,
      "grad_norm": 0.2751936912536621,
      "learning_rate": 2.2549019607843138e-05,
      "loss": 0.1986,
      "step": 189
    },
    {
      "epoch": 0.8135956108657835,
      "grad_norm": 0.2982531487941742,
      "learning_rate": 2.2058823529411766e-05,
      "loss": 0.224,
      "step": 190
    },
    {
      "epoch": 0.817877693028235,
      "grad_norm": 0.29918336868286133,
      "learning_rate": 2.1568627450980395e-05,
      "loss": 0.193,
      "step": 191
    },
    {
      "epoch": 0.8221597751906865,
      "grad_norm": 0.2874895930290222,
      "learning_rate": 2.107843137254902e-05,
      "loss": 0.2357,
      "step": 192
    },
    {
      "epoch": 0.8264418573531379,
      "grad_norm": 0.2816848158836365,
      "learning_rate": 2.058823529411765e-05,
      "loss": 0.2081,
      "step": 193
    },
    {
      "epoch": 0.8307239395155894,
      "grad_norm": 0.3022446632385254,
      "learning_rate": 2.0098039215686277e-05,
      "loss": 0.2222,
      "step": 194
    },
    {
      "epoch": 0.8350060216780409,
      "grad_norm": 0.3225061297416687,
      "learning_rate": 1.9607843137254903e-05,
      "loss": 0.2637,
      "step": 195
    },
    {
      "epoch": 0.8392881038404925,
      "grad_norm": 0.32750946283340454,
      "learning_rate": 1.9117647058823528e-05,
      "loss": 0.2409,
      "step": 196
    },
    {
      "epoch": 0.843570186002944,
      "grad_norm": 0.3293348550796509,
      "learning_rate": 1.862745098039216e-05,
      "loss": 0.2746,
      "step": 197
    },
    {
      "epoch": 0.8478522681653954,
      "grad_norm": 0.25766468048095703,
      "learning_rate": 1.8137254901960785e-05,
      "loss": 0.2057,
      "step": 198
    },
    {
      "epoch": 0.8521343503278469,
      "grad_norm": 0.32190749049186707,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.2355,
      "step": 199
    },
    {
      "epoch": 0.8564164324902984,
      "grad_norm": 0.29550763964653015,
      "learning_rate": 1.715686274509804e-05,
      "loss": 0.2067,
      "step": 200
    },
    {
      "epoch": 0.8606985146527499,
      "grad_norm": 0.30072906613349915,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.2451,
      "step": 201
    },
    {
      "epoch": 0.8649805968152013,
      "grad_norm": 0.30366024374961853,
      "learning_rate": 1.6176470588235296e-05,
      "loss": 0.2311,
      "step": 202
    },
    {
      "epoch": 0.8692626789776529,
      "grad_norm": 0.28947317600250244,
      "learning_rate": 1.568627450980392e-05,
      "loss": 0.2376,
      "step": 203
    },
    {
      "epoch": 0.8735447611401044,
      "grad_norm": 0.27679264545440674,
      "learning_rate": 1.5196078431372548e-05,
      "loss": 0.2312,
      "step": 204
    },
    {
      "epoch": 0.8778268433025559,
      "grad_norm": 0.2838965654373169,
      "learning_rate": 1.4705882352941177e-05,
      "loss": 0.202,
      "step": 205
    },
    {
      "epoch": 0.8821089254650074,
      "grad_norm": 0.2983388602733612,
      "learning_rate": 1.4215686274509804e-05,
      "loss": 0.2283,
      "step": 206
    },
    {
      "epoch": 0.8863910076274588,
      "grad_norm": 0.27599766850471497,
      "learning_rate": 1.3725490196078432e-05,
      "loss": 0.1915,
      "step": 207
    },
    {
      "epoch": 0.8906730897899103,
      "grad_norm": 0.31190070509910583,
      "learning_rate": 1.323529411764706e-05,
      "loss": 0.259,
      "step": 208
    },
    {
      "epoch": 0.8949551719523618,
      "grad_norm": 0.3071896433830261,
      "learning_rate": 1.2745098039215686e-05,
      "loss": 0.2539,
      "step": 209
    },
    {
      "epoch": 0.8992372541148134,
      "grad_norm": 0.30813899636268616,
      "learning_rate": 1.2254901960784313e-05,
      "loss": 0.2237,
      "step": 210
    },
    {
      "epoch": 0.9035193362772648,
      "grad_norm": 0.3494122624397278,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.2455,
      "step": 211
    },
    {
      "epoch": 0.9078014184397163,
      "grad_norm": 0.28135284781455994,
      "learning_rate": 1.1274509803921569e-05,
      "loss": 0.1908,
      "step": 212
    },
    {
      "epoch": 0.9120835006021678,
      "grad_norm": 0.33502623438835144,
      "learning_rate": 1.0784313725490197e-05,
      "loss": 0.2404,
      "step": 213
    },
    {
      "epoch": 0.9163655827646193,
      "grad_norm": 0.30870556831359863,
      "learning_rate": 1.0294117647058824e-05,
      "loss": 0.2493,
      "step": 214
    },
    {
      "epoch": 0.9206476649270708,
      "grad_norm": 0.33104825019836426,
      "learning_rate": 9.803921568627451e-06,
      "loss": 0.2354,
      "step": 215
    },
    {
      "epoch": 0.9249297470895222,
      "grad_norm": 0.29549095034599304,
      "learning_rate": 9.31372549019608e-06,
      "loss": 0.2467,
      "step": 216
    },
    {
      "epoch": 0.9292118292519738,
      "grad_norm": 0.28787434101104736,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.23,
      "step": 217
    },
    {
      "epoch": 0.9334939114144253,
      "grad_norm": 0.3066050708293915,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.2257,
      "step": 218
    },
    {
      "epoch": 0.9377759935768768,
      "grad_norm": 0.35698211193084717,
      "learning_rate": 7.84313725490196e-06,
      "loss": 0.2715,
      "step": 219
    },
    {
      "epoch": 0.9420580757393282,
      "grad_norm": 0.2793384790420532,
      "learning_rate": 7.3529411764705884e-06,
      "loss": 0.1954,
      "step": 220
    },
    {
      "epoch": 0.9463401579017797,
      "grad_norm": 0.35188594460487366,
      "learning_rate": 6.862745098039216e-06,
      "loss": 0.2827,
      "step": 221
    },
    {
      "epoch": 0.9506222400642312,
      "grad_norm": 0.3033970594406128,
      "learning_rate": 6.372549019607843e-06,
      "loss": 0.24,
      "step": 222
    },
    {
      "epoch": 0.9549043222266828,
      "grad_norm": 0.3248239755630493,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.2356,
      "step": 223
    },
    {
      "epoch": 0.9591864043891342,
      "grad_norm": 0.28153759241104126,
      "learning_rate": 5.392156862745099e-06,
      "loss": 0.223,
      "step": 224
    },
    {
      "epoch": 0.9634684865515857,
      "grad_norm": 0.2797453701496124,
      "learning_rate": 4.901960784313726e-06,
      "loss": 0.2256,
      "step": 225
    },
    {
      "epoch": 0.9677505687140372,
      "grad_norm": 0.3379395008087158,
      "learning_rate": 4.411764705882353e-06,
      "loss": 0.2705,
      "step": 226
    },
    {
      "epoch": 0.9720326508764887,
      "grad_norm": 0.2997606694698334,
      "learning_rate": 3.92156862745098e-06,
      "loss": 0.2062,
      "step": 227
    },
    {
      "epoch": 0.9763147330389402,
      "grad_norm": 0.32395467162132263,
      "learning_rate": 3.431372549019608e-06,
      "loss": 0.2581,
      "step": 228
    },
    {
      "epoch": 0.9805968152013916,
      "grad_norm": 0.31389373540878296,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 0.2435,
      "step": 229
    },
    {
      "epoch": 0.9848788973638432,
      "grad_norm": 0.3246689736843109,
      "learning_rate": 2.450980392156863e-06,
      "loss": 0.2319,
      "step": 230
    },
    {
      "epoch": 0.9891609795262947,
      "grad_norm": 0.32743996381759644,
      "learning_rate": 1.96078431372549e-06,
      "loss": 0.2158,
      "step": 231
    },
    {
      "epoch": 0.9934430616887462,
      "grad_norm": 0.31922298669815063,
      "learning_rate": 1.4705882352941177e-06,
      "loss": 0.2236,
      "step": 232
    },
    {
      "epoch": 0.9977251438511976,
      "grad_norm": 0.30808648467063904,
      "learning_rate": 9.80392156862745e-07,
      "loss": 0.1959,
      "step": 233
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.49088653922080994,
      "learning_rate": 4.901960784313725e-07,
      "loss": 0.212,
      "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": 3.6606806196747264e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}