{
  "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": 2.4878523349761963,
      "learning_rate": 0.0,
      "loss": 1.3598,
      "step": 1
    },
    {
      "epoch": 0.008564164324902984,
      "grad_norm": 1.9838305711746216,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0978,
      "step": 2
    },
    {
      "epoch": 0.012846246487354477,
      "grad_norm": 2.3863673210144043,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.3175,
      "step": 3
    },
    {
      "epoch": 0.017128328649805968,
      "grad_norm": 2.275620222091675,
      "learning_rate": 1e-05,
      "loss": 1.2824,
      "step": 4
    },
    {
      "epoch": 0.02141041081225746,
      "grad_norm": 2.2979214191436768,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 1.2941,
      "step": 5
    },
    {
      "epoch": 0.025692492974708953,
      "grad_norm": 2.187796115875244,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.2794,
      "step": 6
    },
    {
      "epoch": 0.029974575137160443,
      "grad_norm": 1.9881962537765503,
      "learning_rate": 2e-05,
      "loss": 1.1338,
      "step": 7
    },
    {
      "epoch": 0.034256657299611935,
      "grad_norm": 1.3727648258209229,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.9047,
      "step": 8
    },
    {
      "epoch": 0.03853873946206343,
      "grad_norm": 1.1047203540802002,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.8618,
      "step": 9
    },
    {
      "epoch": 0.04282082162451492,
      "grad_norm": 0.843003511428833,
      "learning_rate": 3e-05,
      "loss": 0.7817,
      "step": 10
    },
    {
      "epoch": 0.04710290378696641,
      "grad_norm": 0.7555856108665466,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.7282,
      "step": 11
    },
    {
      "epoch": 0.051384985949417906,
      "grad_norm": 0.8716446757316589,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.6094,
      "step": 12
    },
    {
      "epoch": 0.055667068111869396,
      "grad_norm": 0.6328527927398682,
      "learning_rate": 4e-05,
      "loss": 0.5976,
      "step": 13
    },
    {
      "epoch": 0.059949150274320885,
      "grad_norm": 0.621186375617981,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.5339,
      "step": 14
    },
    {
      "epoch": 0.06423123243677238,
      "grad_norm": 0.5601040124893188,
      "learning_rate": 4.666666666666667e-05,
      "loss": 0.5452,
      "step": 15
    },
    {
      "epoch": 0.06851331459922387,
      "grad_norm": 0.4794214963912964,
      "learning_rate": 5e-05,
      "loss": 0.4495,
      "step": 16
    },
    {
      "epoch": 0.07279539676167536,
      "grad_norm": 0.4371148943901062,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.426,
      "step": 17
    },
    {
      "epoch": 0.07707747892412686,
      "grad_norm": 0.3789004385471344,
      "learning_rate": 5.666666666666667e-05,
      "loss": 0.4357,
      "step": 18
    },
    {
      "epoch": 0.08135956108657835,
      "grad_norm": 0.32028523087501526,
      "learning_rate": 6e-05,
      "loss": 0.3448,
      "step": 19
    },
    {
      "epoch": 0.08564164324902984,
      "grad_norm": 0.45856499671936035,
      "learning_rate": 6.333333333333333e-05,
      "loss": 0.3894,
      "step": 20
    },
    {
      "epoch": 0.08992372541148133,
      "grad_norm": 0.37712252140045166,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.373,
      "step": 21
    },
    {
      "epoch": 0.09420580757393282,
      "grad_norm": 0.32988882064819336,
      "learning_rate": 7e-05,
      "loss": 0.3239,
      "step": 22
    },
    {
      "epoch": 0.09848788973638432,
      "grad_norm": 0.31284961104393005,
      "learning_rate": 7.333333333333333e-05,
      "loss": 0.3487,
      "step": 23
    },
    {
      "epoch": 0.10276997189883581,
      "grad_norm": 0.2780960202217102,
      "learning_rate": 7.666666666666667e-05,
      "loss": 0.3289,
      "step": 24
    },
    {
      "epoch": 0.1070520540612873,
      "grad_norm": 0.27915439009666443,
      "learning_rate": 8e-05,
      "loss": 0.2919,
      "step": 25
    },
    {
      "epoch": 0.11133413622373879,
      "grad_norm": 0.2739494740962982,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.2972,
      "step": 26
    },
    {
      "epoch": 0.11561621838619028,
      "grad_norm": 0.30776894092559814,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.3112,
      "step": 27
    },
    {
      "epoch": 0.11989830054864177,
      "grad_norm": 0.30830278992652893,
      "learning_rate": 9e-05,
      "loss": 0.3515,
      "step": 28
    },
    {
      "epoch": 0.12418038271109327,
      "grad_norm": 0.2607559263706207,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.3136,
      "step": 29
    },
    {
      "epoch": 0.12846246487354476,
      "grad_norm": 0.3013528883457184,
      "learning_rate": 9.666666666666667e-05,
      "loss": 0.3447,
      "step": 30
    },
    {
      "epoch": 0.13274454703599625,
      "grad_norm": 0.2951507568359375,
      "learning_rate": 0.0001,
      "loss": 0.3561,
      "step": 31
    },
    {
      "epoch": 0.13702662919844774,
      "grad_norm": 0.27145230770111084,
      "learning_rate": 9.950980392156864e-05,
      "loss": 0.3174,
      "step": 32
    },
    {
      "epoch": 0.14130871136089923,
      "grad_norm": 0.2763392925262451,
      "learning_rate": 9.901960784313727e-05,
      "loss": 0.2994,
      "step": 33
    },
    {
      "epoch": 0.14559079352335072,
      "grad_norm": 0.24706591665744781,
      "learning_rate": 9.852941176470589e-05,
      "loss": 0.3189,
      "step": 34
    },
    {
      "epoch": 0.1498728756858022,
      "grad_norm": 0.2073615938425064,
      "learning_rate": 9.80392156862745e-05,
      "loss": 0.3057,
      "step": 35
    },
    {
      "epoch": 0.15415495784825373,
      "grad_norm": 0.2470332384109497,
      "learning_rate": 9.754901960784314e-05,
      "loss": 0.2915,
      "step": 36
    },
    {
      "epoch": 0.15843704001070522,
      "grad_norm": 0.24675661325454712,
      "learning_rate": 9.705882352941177e-05,
      "loss": 0.3069,
      "step": 37
    },
    {
      "epoch": 0.1627191221731567,
      "grad_norm": 0.21644169092178345,
      "learning_rate": 9.65686274509804e-05,
      "loss": 0.2811,
      "step": 38
    },
    {
      "epoch": 0.1670012043356082,
      "grad_norm": 0.20190708339214325,
      "learning_rate": 9.607843137254903e-05,
      "loss": 0.2883,
      "step": 39
    },
    {
      "epoch": 0.17128328649805968,
      "grad_norm": 0.25285035371780396,
      "learning_rate": 9.558823529411765e-05,
      "loss": 0.2939,
      "step": 40
    },
    {
      "epoch": 0.17556536866051117,
      "grad_norm": 0.24124890565872192,
      "learning_rate": 9.509803921568627e-05,
      "loss": 0.2679,
      "step": 41
    },
    {
      "epoch": 0.17984745082296266,
      "grad_norm": 0.23497405648231506,
      "learning_rate": 9.46078431372549e-05,
      "loss": 0.2784,
      "step": 42
    },
    {
      "epoch": 0.18412953298541415,
      "grad_norm": 0.24517400562763214,
      "learning_rate": 9.411764705882353e-05,
      "loss": 0.2858,
      "step": 43
    },
    {
      "epoch": 0.18841161514786564,
      "grad_norm": 0.2252051681280136,
      "learning_rate": 9.362745098039217e-05,
      "loss": 0.2668,
      "step": 44
    },
    {
      "epoch": 0.19269369731031713,
      "grad_norm": 0.24736210703849792,
      "learning_rate": 9.313725490196079e-05,
      "loss": 0.245,
      "step": 45
    },
    {
      "epoch": 0.19697577947276865,
      "grad_norm": 0.2217230349779129,
      "learning_rate": 9.264705882352942e-05,
      "loss": 0.2639,
      "step": 46
    },
    {
      "epoch": 0.20125786163522014,
      "grad_norm": 0.23976930975914001,
      "learning_rate": 9.215686274509804e-05,
      "loss": 0.2799,
      "step": 47
    },
    {
      "epoch": 0.20553994379767163,
      "grad_norm": 0.2588768005371094,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.275,
      "step": 48
    },
    {
      "epoch": 0.20982202596012312,
      "grad_norm": 0.2617010772228241,
      "learning_rate": 9.11764705882353e-05,
      "loss": 0.2939,
      "step": 49
    },
    {
      "epoch": 0.2141041081225746,
      "grad_norm": 0.2588669955730438,
      "learning_rate": 9.068627450980393e-05,
      "loss": 0.2987,
      "step": 50
    },
    {
      "epoch": 0.2183861902850261,
      "grad_norm": 0.22989527881145477,
      "learning_rate": 9.019607843137255e-05,
      "loss": 0.2742,
      "step": 51
    },
    {
      "epoch": 0.22266827244747758,
      "grad_norm": 0.24354051053524017,
      "learning_rate": 8.970588235294118e-05,
      "loss": 0.2771,
      "step": 52
    },
    {
      "epoch": 0.22695035460992907,
      "grad_norm": 0.22785232961177826,
      "learning_rate": 8.921568627450981e-05,
      "loss": 0.2641,
      "step": 53
    },
    {
      "epoch": 0.23123243677238056,
      "grad_norm": 0.2526540458202362,
      "learning_rate": 8.872549019607843e-05,
      "loss": 0.3021,
      "step": 54
    },
    {
      "epoch": 0.23551451893483205,
      "grad_norm": 0.22112172842025757,
      "learning_rate": 8.823529411764706e-05,
      "loss": 0.2512,
      "step": 55
    },
    {
      "epoch": 0.23979660109728354,
      "grad_norm": 0.2365880012512207,
      "learning_rate": 8.774509803921568e-05,
      "loss": 0.2737,
      "step": 56
    },
    {
      "epoch": 0.24407868325973506,
      "grad_norm": 0.25947895646095276,
      "learning_rate": 8.725490196078432e-05,
      "loss": 0.2726,
      "step": 57
    },
    {
      "epoch": 0.24836076542218655,
      "grad_norm": 0.23627030849456787,
      "learning_rate": 8.676470588235295e-05,
      "loss": 0.2965,
      "step": 58
    },
    {
      "epoch": 0.252642847584638,
      "grad_norm": 0.27099597454071045,
      "learning_rate": 8.627450980392158e-05,
      "loss": 0.2655,
      "step": 59
    },
    {
      "epoch": 0.2569249297470895,
      "grad_norm": 0.22191449999809265,
      "learning_rate": 8.57843137254902e-05,
      "loss": 0.278,
      "step": 60
    },
    {
      "epoch": 0.261207011909541,
      "grad_norm": 0.24128779768943787,
      "learning_rate": 8.529411764705883e-05,
      "loss": 0.273,
      "step": 61
    },
    {
      "epoch": 0.2654890940719925,
      "grad_norm": 0.22755272686481476,
      "learning_rate": 8.480392156862745e-05,
      "loss": 0.2723,
      "step": 62
    },
    {
      "epoch": 0.269771176234444,
      "grad_norm": 0.26044610142707825,
      "learning_rate": 8.431372549019608e-05,
      "loss": 0.2911,
      "step": 63
    },
    {
      "epoch": 0.2740532583968955,
      "grad_norm": 0.23918823897838593,
      "learning_rate": 8.382352941176471e-05,
      "loss": 0.2378,
      "step": 64
    },
    {
      "epoch": 0.278335340559347,
      "grad_norm": 0.22860293090343475,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.2628,
      "step": 65
    },
    {
      "epoch": 0.28261742272179846,
      "grad_norm": 0.22429755330085754,
      "learning_rate": 8.284313725490198e-05,
      "loss": 0.2484,
      "step": 66
    },
    {
      "epoch": 0.28689950488425,
      "grad_norm": 0.26635828614234924,
      "learning_rate": 8.23529411764706e-05,
      "loss": 0.3084,
      "step": 67
    },
    {
      "epoch": 0.29118158704670144,
      "grad_norm": 0.3146519362926483,
      "learning_rate": 8.186274509803921e-05,
      "loss": 0.304,
      "step": 68
    },
    {
      "epoch": 0.29546366920915296,
      "grad_norm": 0.23846609890460968,
      "learning_rate": 8.137254901960785e-05,
      "loss": 0.264,
      "step": 69
    },
    {
      "epoch": 0.2997457513716044,
      "grad_norm": 0.2358589768409729,
      "learning_rate": 8.088235294117648e-05,
      "loss": 0.2685,
      "step": 70
    },
    {
      "epoch": 0.30402783353405594,
      "grad_norm": 0.2648678421974182,
      "learning_rate": 8.039215686274511e-05,
      "loss": 0.2931,
      "step": 71
    },
    {
      "epoch": 0.30830991569650745,
      "grad_norm": 0.25414445996284485,
      "learning_rate": 7.990196078431373e-05,
      "loss": 0.2574,
      "step": 72
    },
    {
      "epoch": 0.3125919978589589,
      "grad_norm": 0.2337549328804016,
      "learning_rate": 7.941176470588235e-05,
      "loss": 0.2677,
      "step": 73
    },
    {
      "epoch": 0.31687408002141043,
      "grad_norm": 0.2523665726184845,
      "learning_rate": 7.892156862745098e-05,
      "loss": 0.2451,
      "step": 74
    },
    {
      "epoch": 0.3211561621838619,
      "grad_norm": 0.2410557121038437,
      "learning_rate": 7.843137254901961e-05,
      "loss": 0.2381,
      "step": 75
    },
    {
      "epoch": 0.3254382443463134,
      "grad_norm": 0.2714288830757141,
      "learning_rate": 7.794117647058824e-05,
      "loss": 0.3045,
      "step": 76
    },
    {
      "epoch": 0.32972032650876487,
      "grad_norm": 0.27958807349205017,
      "learning_rate": 7.745098039215687e-05,
      "loss": 0.2509,
      "step": 77
    },
    {
      "epoch": 0.3340024086712164,
      "grad_norm": 0.2251473218202591,
      "learning_rate": 7.696078431372549e-05,
      "loss": 0.2444,
      "step": 78
    },
    {
      "epoch": 0.33828449083366785,
      "grad_norm": 0.27283602952957153,
      "learning_rate": 7.647058823529411e-05,
      "loss": 0.288,
      "step": 79
    },
    {
      "epoch": 0.34256657299611937,
      "grad_norm": 0.2540000081062317,
      "learning_rate": 7.598039215686274e-05,
      "loss": 0.2435,
      "step": 80
    },
    {
      "epoch": 0.34684865515857083,
      "grad_norm": 0.2543659210205078,
      "learning_rate": 7.549019607843137e-05,
      "loss": 0.2655,
      "step": 81
    },
    {
      "epoch": 0.35113073732102235,
      "grad_norm": 0.26397407054901123,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.2618,
      "step": 82
    },
    {
      "epoch": 0.35541281948347386,
      "grad_norm": 0.293526828289032,
      "learning_rate": 7.450980392156864e-05,
      "loss": 0.2688,
      "step": 83
    },
    {
      "epoch": 0.3596949016459253,
      "grad_norm": 0.22264349460601807,
      "learning_rate": 7.401960784313726e-05,
      "loss": 0.2009,
      "step": 84
    },
    {
      "epoch": 0.36397698380837684,
      "grad_norm": 0.2644459307193756,
      "learning_rate": 7.352941176470589e-05,
      "loss": 0.2466,
      "step": 85
    },
    {
      "epoch": 0.3682590659708283,
      "grad_norm": 0.25630417466163635,
      "learning_rate": 7.303921568627451e-05,
      "loss": 0.2425,
      "step": 86
    },
    {
      "epoch": 0.3725411481332798,
      "grad_norm": 0.2758299708366394,
      "learning_rate": 7.254901960784314e-05,
      "loss": 0.2405,
      "step": 87
    },
    {
      "epoch": 0.3768232302957313,
      "grad_norm": 0.2609292268753052,
      "learning_rate": 7.205882352941177e-05,
      "loss": 0.2867,
      "step": 88
    },
    {
      "epoch": 0.3811053124581828,
      "grad_norm": 0.23372113704681396,
      "learning_rate": 7.156862745098039e-05,
      "loss": 0.2593,
      "step": 89
    },
    {
      "epoch": 0.38538739462063426,
      "grad_norm": 0.2724294066429138,
      "learning_rate": 7.107843137254902e-05,
      "loss": 0.259,
      "step": 90
    },
    {
      "epoch": 0.3896694767830858,
      "grad_norm": 0.2579306960105896,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.2578,
      "step": 91
    },
    {
      "epoch": 0.3939515589455373,
      "grad_norm": 0.24242401123046875,
      "learning_rate": 7.009803921568627e-05,
      "loss": 0.2354,
      "step": 92
    },
    {
      "epoch": 0.39823364110798876,
      "grad_norm": 0.2605730891227722,
      "learning_rate": 6.96078431372549e-05,
      "loss": 0.2515,
      "step": 93
    },
    {
      "epoch": 0.4025157232704403,
      "grad_norm": 0.23951297998428345,
      "learning_rate": 6.911764705882354e-05,
      "loss": 0.2265,
      "step": 94
    },
    {
      "epoch": 0.40679780543289173,
      "grad_norm": 0.2644834816455841,
      "learning_rate": 6.862745098039216e-05,
      "loss": 0.2682,
      "step": 95
    },
    {
      "epoch": 0.41107988759534325,
      "grad_norm": 0.2700228691101074,
      "learning_rate": 6.813725490196079e-05,
      "loss": 0.2795,
      "step": 96
    },
    {
      "epoch": 0.4153619697577947,
      "grad_norm": 0.29942718148231506,
      "learning_rate": 6.764705882352942e-05,
      "loss": 0.2562,
      "step": 97
    },
    {
      "epoch": 0.41964405192024623,
      "grad_norm": 0.28121066093444824,
      "learning_rate": 6.715686274509804e-05,
      "loss": 0.274,
      "step": 98
    },
    {
      "epoch": 0.4239261340826977,
      "grad_norm": 0.24854238331317902,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.2352,
      "step": 99
    },
    {
      "epoch": 0.4282082162451492,
      "grad_norm": 0.28277692198753357,
      "learning_rate": 6.61764705882353e-05,
      "loss": 0.2747,
      "step": 100
    },
    {
      "epoch": 0.43249029840760067,
      "grad_norm": 0.27323755621910095,
      "learning_rate": 6.568627450980392e-05,
      "loss": 0.2455,
      "step": 101
    },
    {
      "epoch": 0.4367723805700522,
      "grad_norm": 0.282482385635376,
      "learning_rate": 6.519607843137255e-05,
      "loss": 0.2651,
      "step": 102
    },
    {
      "epoch": 0.4410544627325037,
      "grad_norm": 0.29933255910873413,
      "learning_rate": 6.470588235294118e-05,
      "loss": 0.2596,
      "step": 103
    },
    {
      "epoch": 0.44533654489495517,
      "grad_norm": 0.284763902425766,
      "learning_rate": 6.421568627450982e-05,
      "loss": 0.2478,
      "step": 104
    },
    {
      "epoch": 0.4496186270574067,
      "grad_norm": 0.27358904480934143,
      "learning_rate": 6.372549019607843e-05,
      "loss": 0.256,
      "step": 105
    },
    {
      "epoch": 0.45390070921985815,
      "grad_norm": 0.41629382967948914,
      "learning_rate": 6.323529411764705e-05,
      "loss": 0.3009,
      "step": 106
    },
    {
      "epoch": 0.45818279138230966,
      "grad_norm": 0.30577167868614197,
      "learning_rate": 6.274509803921569e-05,
      "loss": 0.2021,
      "step": 107
    },
    {
      "epoch": 0.4624648735447611,
      "grad_norm": 0.27379924058914185,
      "learning_rate": 6.225490196078432e-05,
      "loss": 0.2402,
      "step": 108
    },
    {
      "epoch": 0.46674695570721264,
      "grad_norm": 0.2772539556026459,
      "learning_rate": 6.176470588235295e-05,
      "loss": 0.2898,
      "step": 109
    },
    {
      "epoch": 0.4710290378696641,
      "grad_norm": 0.2680152356624603,
      "learning_rate": 6.127450980392158e-05,
      "loss": 0.2498,
      "step": 110
    },
    {
      "epoch": 0.4753111200321156,
      "grad_norm": 0.2348792850971222,
      "learning_rate": 6.078431372549019e-05,
      "loss": 0.2321,
      "step": 111
    },
    {
      "epoch": 0.4795932021945671,
      "grad_norm": 0.25828760862350464,
      "learning_rate": 6.0294117647058825e-05,
      "loss": 0.2378,
      "step": 112
    },
    {
      "epoch": 0.4838752843570186,
      "grad_norm": 0.26433736085891724,
      "learning_rate": 5.980392156862745e-05,
      "loss": 0.2409,
      "step": 113
    },
    {
      "epoch": 0.4881573665194701,
      "grad_norm": 0.30242738127708435,
      "learning_rate": 5.931372549019608e-05,
      "loss": 0.2462,
      "step": 114
    },
    {
      "epoch": 0.4924394486819216,
      "grad_norm": 0.266008585691452,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.2617,
      "step": 115
    },
    {
      "epoch": 0.4967215308443731,
      "grad_norm": 0.28391969203948975,
      "learning_rate": 5.833333333333334e-05,
      "loss": 0.2316,
      "step": 116
    },
    {
      "epoch": 0.5010036130068246,
      "grad_norm": 0.2897227108478546,
      "learning_rate": 5.784313725490197e-05,
      "loss": 0.2456,
      "step": 117
    },
    {
      "epoch": 0.505285695169276,
      "grad_norm": 0.28793928027153015,
      "learning_rate": 5.735294117647059e-05,
      "loss": 0.2712,
      "step": 118
    },
    {
      "epoch": 0.5095677773317275,
      "grad_norm": 0.2534613609313965,
      "learning_rate": 5.6862745098039215e-05,
      "loss": 0.2242,
      "step": 119
    },
    {
      "epoch": 0.513849859494179,
      "grad_norm": 0.24495765566825867,
      "learning_rate": 5.637254901960785e-05,
      "loss": 0.2411,
      "step": 120
    },
    {
      "epoch": 0.5181319416566306,
      "grad_norm": 0.2918499708175659,
      "learning_rate": 5.588235294117647e-05,
      "loss": 0.2528,
      "step": 121
    },
    {
      "epoch": 0.522414023819082,
      "grad_norm": 0.25384247303009033,
      "learning_rate": 5.5392156862745104e-05,
      "loss": 0.2394,
      "step": 122
    },
    {
      "epoch": 0.5266961059815335,
      "grad_norm": 0.25959160923957825,
      "learning_rate": 5.490196078431373e-05,
      "loss": 0.2368,
      "step": 123
    },
    {
      "epoch": 0.530978188143985,
      "grad_norm": 0.30009910464286804,
      "learning_rate": 5.441176470588235e-05,
      "loss": 0.2929,
      "step": 124
    },
    {
      "epoch": 0.5352602703064365,
      "grad_norm": 0.28385186195373535,
      "learning_rate": 5.392156862745098e-05,
      "loss": 0.2246,
      "step": 125
    },
    {
      "epoch": 0.539542352468888,
      "grad_norm": 0.28721216320991516,
      "learning_rate": 5.343137254901961e-05,
      "loss": 0.2672,
      "step": 126
    },
    {
      "epoch": 0.5438244346313394,
      "grad_norm": 0.24484668672084808,
      "learning_rate": 5.294117647058824e-05,
      "loss": 0.2273,
      "step": 127
    },
    {
      "epoch": 0.548106516793791,
      "grad_norm": 0.2539582848548889,
      "learning_rate": 5.245098039215687e-05,
      "loss": 0.2365,
      "step": 128
    },
    {
      "epoch": 0.5523885989562425,
      "grad_norm": 0.2860145568847656,
      "learning_rate": 5.1960784313725495e-05,
      "loss": 0.2444,
      "step": 129
    },
    {
      "epoch": 0.556670681118694,
      "grad_norm": 0.27066272497177124,
      "learning_rate": 5.147058823529411e-05,
      "loss": 0.2888,
      "step": 130
    },
    {
      "epoch": 0.5609527632811454,
      "grad_norm": 0.29372817277908325,
      "learning_rate": 5.0980392156862745e-05,
      "loss": 0.2379,
      "step": 131
    },
    {
      "epoch": 0.5652348454435969,
      "grad_norm": 0.2668856084346771,
      "learning_rate": 5.049019607843137e-05,
      "loss": 0.2374,
      "step": 132
    },
    {
      "epoch": 0.5695169276060484,
      "grad_norm": 0.30546411871910095,
      "learning_rate": 5e-05,
      "loss": 0.2202,
      "step": 133
    },
    {
      "epoch": 0.5737990097685,
      "grad_norm": 0.30588456988334656,
      "learning_rate": 4.9509803921568634e-05,
      "loss": 0.2897,
      "step": 134
    },
    {
      "epoch": 0.5780810919309515,
      "grad_norm": 0.3344208002090454,
      "learning_rate": 4.901960784313725e-05,
      "loss": 0.2261,
      "step": 135
    },
    {
      "epoch": 0.5823631740934029,
      "grad_norm": 0.29046499729156494,
      "learning_rate": 4.8529411764705885e-05,
      "loss": 0.2483,
      "step": 136
    },
    {
      "epoch": 0.5866452562558544,
      "grad_norm": 0.27813631296157837,
      "learning_rate": 4.803921568627452e-05,
      "loss": 0.2484,
      "step": 137
    },
    {
      "epoch": 0.5909273384183059,
      "grad_norm": 0.3182491362094879,
      "learning_rate": 4.7549019607843135e-05,
      "loss": 0.255,
      "step": 138
    },
    {
      "epoch": 0.5952094205807574,
      "grad_norm": 0.26866352558135986,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.2362,
      "step": 139
    },
    {
      "epoch": 0.5994915027432088,
      "grad_norm": 0.24540072679519653,
      "learning_rate": 4.656862745098039e-05,
      "loss": 0.2125,
      "step": 140
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.274228036403656,
      "learning_rate": 4.607843137254902e-05,
      "loss": 0.2477,
      "step": 141
    },
    {
      "epoch": 0.6080556670681119,
      "grad_norm": 0.2659485340118408,
      "learning_rate": 4.558823529411765e-05,
      "loss": 0.2144,
      "step": 142
    },
    {
      "epoch": 0.6123377492305634,
      "grad_norm": 0.2808791697025299,
      "learning_rate": 4.5098039215686275e-05,
      "loss": 0.2484,
      "step": 143
    },
    {
      "epoch": 0.6166198313930149,
      "grad_norm": 0.3063105344772339,
      "learning_rate": 4.460784313725491e-05,
      "loss": 0.2648,
      "step": 144
    },
    {
      "epoch": 0.6209019135554663,
      "grad_norm": 0.28035491704940796,
      "learning_rate": 4.411764705882353e-05,
      "loss": 0.2438,
      "step": 145
    },
    {
      "epoch": 0.6251839957179178,
      "grad_norm": 0.2870001196861267,
      "learning_rate": 4.362745098039216e-05,
      "loss": 0.2488,
      "step": 146
    },
    {
      "epoch": 0.6294660778803693,
      "grad_norm": 0.27585291862487793,
      "learning_rate": 4.313725490196079e-05,
      "loss": 0.2626,
      "step": 147
    },
    {
      "epoch": 0.6337481600428209,
      "grad_norm": 0.28468048572540283,
      "learning_rate": 4.2647058823529415e-05,
      "loss": 0.2515,
      "step": 148
    },
    {
      "epoch": 0.6380302422052723,
      "grad_norm": 0.27381083369255066,
      "learning_rate": 4.215686274509804e-05,
      "loss": 0.2161,
      "step": 149
    },
    {
      "epoch": 0.6423123243677238,
      "grad_norm": 0.2693536579608917,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.2149,
      "step": 150
    },
    {
      "epoch": 0.6465944065301753,
      "grad_norm": 0.3121880292892456,
      "learning_rate": 4.11764705882353e-05,
      "loss": 0.2478,
      "step": 151
    },
    {
      "epoch": 0.6508764886926268,
      "grad_norm": 0.31520259380340576,
      "learning_rate": 4.068627450980392e-05,
      "loss": 0.2294,
      "step": 152
    },
    {
      "epoch": 0.6551585708550783,
      "grad_norm": 0.3019797205924988,
      "learning_rate": 4.0196078431372555e-05,
      "loss": 0.2308,
      "step": 153
    },
    {
      "epoch": 0.6594406530175297,
      "grad_norm": 0.3353118598461151,
      "learning_rate": 3.970588235294117e-05,
      "loss": 0.2701,
      "step": 154
    },
    {
      "epoch": 0.6637227351799813,
      "grad_norm": 0.3227437138557434,
      "learning_rate": 3.9215686274509805e-05,
      "loss": 0.2342,
      "step": 155
    },
    {
      "epoch": 0.6680048173424328,
      "grad_norm": 0.33070555329322815,
      "learning_rate": 3.872549019607844e-05,
      "loss": 0.2508,
      "step": 156
    },
    {
      "epoch": 0.6722868995048843,
      "grad_norm": 0.3071269094944,
      "learning_rate": 3.8235294117647055e-05,
      "loss": 0.2911,
      "step": 157
    },
    {
      "epoch": 0.6765689816673357,
      "grad_norm": 0.326850563287735,
      "learning_rate": 3.774509803921569e-05,
      "loss": 0.2593,
      "step": 158
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.2925606369972229,
      "learning_rate": 3.725490196078432e-05,
      "loss": 0.2601,
      "step": 159
    },
    {
      "epoch": 0.6851331459922387,
      "grad_norm": 0.28228139877319336,
      "learning_rate": 3.6764705882352945e-05,
      "loss": 0.2305,
      "step": 160
    },
    {
      "epoch": 0.6894152281546903,
      "grad_norm": 0.29046449065208435,
      "learning_rate": 3.627450980392157e-05,
      "loss": 0.1943,
      "step": 161
    },
    {
      "epoch": 0.6936973103171417,
      "grad_norm": 0.27570340037345886,
      "learning_rate": 3.5784313725490195e-05,
      "loss": 0.2848,
      "step": 162
    },
    {
      "epoch": 0.6979793924795932,
      "grad_norm": 0.23822766542434692,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.2077,
      "step": 163
    },
    {
      "epoch": 0.7022614746420447,
      "grad_norm": 0.2474467009305954,
      "learning_rate": 3.480392156862745e-05,
      "loss": 0.203,
      "step": 164
    },
    {
      "epoch": 0.7065435568044962,
      "grad_norm": 0.27342796325683594,
      "learning_rate": 3.431372549019608e-05,
      "loss": 0.2287,
      "step": 165
    },
    {
      "epoch": 0.7108256389669477,
      "grad_norm": 0.28830328583717346,
      "learning_rate": 3.382352941176471e-05,
      "loss": 0.2265,
      "step": 166
    },
    {
      "epoch": 0.7151077211293991,
      "grad_norm": 0.2837151288986206,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.2556,
      "step": 167
    },
    {
      "epoch": 0.7193898032918506,
      "grad_norm": 0.2651543915271759,
      "learning_rate": 3.284313725490196e-05,
      "loss": 0.2267,
      "step": 168
    },
    {
      "epoch": 0.7236718854543022,
      "grad_norm": 0.3021429479122162,
      "learning_rate": 3.235294117647059e-05,
      "loss": 0.2023,
      "step": 169
    },
    {
      "epoch": 0.7279539676167537,
      "grad_norm": 0.2809842526912689,
      "learning_rate": 3.186274509803922e-05,
      "loss": 0.2365,
      "step": 170
    },
    {
      "epoch": 0.7322360497792051,
      "grad_norm": 0.29352620244026184,
      "learning_rate": 3.137254901960784e-05,
      "loss": 0.2504,
      "step": 171
    },
    {
      "epoch": 0.7365181319416566,
      "grad_norm": 0.2841671407222748,
      "learning_rate": 3.0882352941176475e-05,
      "loss": 0.2211,
      "step": 172
    },
    {
      "epoch": 0.7408002141041081,
      "grad_norm": 0.33483561873435974,
      "learning_rate": 3.0392156862745097e-05,
      "loss": 0.2745,
      "step": 173
    },
    {
      "epoch": 0.7450822962665596,
      "grad_norm": 0.30535849928855896,
      "learning_rate": 2.9901960784313725e-05,
      "loss": 0.2242,
      "step": 174
    },
    {
      "epoch": 0.7493643784290112,
      "grad_norm": 0.28258150815963745,
      "learning_rate": 2.9411764705882354e-05,
      "loss": 0.2339,
      "step": 175
    },
    {
      "epoch": 0.7536464605914626,
      "grad_norm": 0.3096565902233124,
      "learning_rate": 2.8921568627450986e-05,
      "loss": 0.2527,
      "step": 176
    },
    {
      "epoch": 0.7579285427539141,
      "grad_norm": 0.28737732768058777,
      "learning_rate": 2.8431372549019608e-05,
      "loss": 0.2317,
      "step": 177
    },
    {
      "epoch": 0.7622106249163656,
      "grad_norm": 0.28823569416999817,
      "learning_rate": 2.7941176470588236e-05,
      "loss": 0.226,
      "step": 178
    },
    {
      "epoch": 0.7664927070788171,
      "grad_norm": 0.27827608585357666,
      "learning_rate": 2.7450980392156865e-05,
      "loss": 0.2213,
      "step": 179
    },
    {
      "epoch": 0.7707747892412685,
      "grad_norm": 0.29937097430229187,
      "learning_rate": 2.696078431372549e-05,
      "loss": 0.208,
      "step": 180
    },
    {
      "epoch": 0.77505687140372,
      "grad_norm": 0.3469245433807373,
      "learning_rate": 2.647058823529412e-05,
      "loss": 0.2647,
      "step": 181
    },
    {
      "epoch": 0.7793389535661716,
      "grad_norm": 0.29952043294906616,
      "learning_rate": 2.5980392156862747e-05,
      "loss": 0.2171,
      "step": 182
    },
    {
      "epoch": 0.7836210357286231,
      "grad_norm": 0.3126024007797241,
      "learning_rate": 2.5490196078431373e-05,
      "loss": 0.249,
      "step": 183
    },
    {
      "epoch": 0.7879031178910746,
      "grad_norm": 0.3616069257259369,
      "learning_rate": 2.5e-05,
      "loss": 0.2499,
      "step": 184
    },
    {
      "epoch": 0.792185200053526,
      "grad_norm": 0.31493785977363586,
      "learning_rate": 2.4509803921568626e-05,
      "loss": 0.2707,
      "step": 185
    },
    {
      "epoch": 0.7964672822159775,
      "grad_norm": 0.2993806004524231,
      "learning_rate": 2.401960784313726e-05,
      "loss": 0.2657,
      "step": 186
    },
    {
      "epoch": 0.800749364378429,
      "grad_norm": 0.2947325110435486,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.2339,
      "step": 187
    },
    {
      "epoch": 0.8050314465408805,
      "grad_norm": 0.28110820055007935,
      "learning_rate": 2.303921568627451e-05,
      "loss": 0.2478,
      "step": 188
    },
    {
      "epoch": 0.809313528703332,
      "grad_norm": 0.2690654397010803,
      "learning_rate": 2.2549019607843138e-05,
      "loss": 0.1976,
      "step": 189
    },
    {
      "epoch": 0.8135956108657835,
      "grad_norm": 0.2916092872619629,
      "learning_rate": 2.2058823529411766e-05,
      "loss": 0.2254,
      "step": 190
    },
    {
      "epoch": 0.817877693028235,
      "grad_norm": 0.29713618755340576,
      "learning_rate": 2.1568627450980395e-05,
      "loss": 0.1932,
      "step": 191
    },
    {
      "epoch": 0.8221597751906865,
      "grad_norm": 0.2790399491786957,
      "learning_rate": 2.107843137254902e-05,
      "loss": 0.2343,
      "step": 192
    },
    {
      "epoch": 0.8264418573531379,
      "grad_norm": 0.2714719772338867,
      "learning_rate": 2.058823529411765e-05,
      "loss": 0.2083,
      "step": 193
    },
    {
      "epoch": 0.8307239395155894,
      "grad_norm": 0.289072185754776,
      "learning_rate": 2.0098039215686277e-05,
      "loss": 0.2243,
      "step": 194
    },
    {
      "epoch": 0.8350060216780409,
      "grad_norm": 0.3053552210330963,
      "learning_rate": 1.9607843137254903e-05,
      "loss": 0.2626,
      "step": 195
    },
    {
      "epoch": 0.8392881038404925,
      "grad_norm": 0.3143557906150818,
      "learning_rate": 1.9117647058823528e-05,
      "loss": 0.2381,
      "step": 196
    },
    {
      "epoch": 0.843570186002944,
      "grad_norm": 0.3242422342300415,
      "learning_rate": 1.862745098039216e-05,
      "loss": 0.2725,
      "step": 197
    },
    {
      "epoch": 0.8478522681653954,
      "grad_norm": 0.2495286464691162,
      "learning_rate": 1.8137254901960785e-05,
      "loss": 0.207,
      "step": 198
    },
    {
      "epoch": 0.8521343503278469,
      "grad_norm": 0.32010331749916077,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.2357,
      "step": 199
    },
    {
      "epoch": 0.8564164324902984,
      "grad_norm": 0.2888106107711792,
      "learning_rate": 1.715686274509804e-05,
      "loss": 0.2045,
      "step": 200
    },
    {
      "epoch": 0.8606985146527499,
      "grad_norm": 0.2900044620037079,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.2444,
      "step": 201
    },
    {
      "epoch": 0.8649805968152013,
      "grad_norm": 0.2952995002269745,
      "learning_rate": 1.6176470588235296e-05,
      "loss": 0.2298,
      "step": 202
    },
    {
      "epoch": 0.8692626789776529,
      "grad_norm": 0.2806972563266754,
      "learning_rate": 1.568627450980392e-05,
      "loss": 0.2379,
      "step": 203
    },
    {
      "epoch": 0.8735447611401044,
      "grad_norm": 0.27058354020118713,
      "learning_rate": 1.5196078431372548e-05,
      "loss": 0.2306,
      "step": 204
    },
    {
      "epoch": 0.8778268433025559,
      "grad_norm": 0.2771023213863373,
      "learning_rate": 1.4705882352941177e-05,
      "loss": 0.2002,
      "step": 205
    },
    {
      "epoch": 0.8821089254650074,
      "grad_norm": 0.30181995034217834,
      "learning_rate": 1.4215686274509804e-05,
      "loss": 0.2283,
      "step": 206
    },
    {
      "epoch": 0.8863910076274588,
      "grad_norm": 0.27368202805519104,
      "learning_rate": 1.3725490196078432e-05,
      "loss": 0.1936,
      "step": 207
    },
    {
      "epoch": 0.8906730897899103,
      "grad_norm": 0.30678945779800415,
      "learning_rate": 1.323529411764706e-05,
      "loss": 0.257,
      "step": 208
    },
    {
      "epoch": 0.8949551719523618,
      "grad_norm": 0.30001550912857056,
      "learning_rate": 1.2745098039215686e-05,
      "loss": 0.2509,
      "step": 209
    },
    {
      "epoch": 0.8992372541148134,
      "grad_norm": 0.30607905983924866,
      "learning_rate": 1.2254901960784313e-05,
      "loss": 0.2236,
      "step": 210
    },
    {
      "epoch": 0.9035193362772648,
      "grad_norm": 0.3329964876174927,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.2425,
      "step": 211
    },
    {
      "epoch": 0.9078014184397163,
      "grad_norm": 0.2751823365688324,
      "learning_rate": 1.1274509803921569e-05,
      "loss": 0.1894,
      "step": 212
    },
    {
      "epoch": 0.9120835006021678,
      "grad_norm": 0.3222876489162445,
      "learning_rate": 1.0784313725490197e-05,
      "loss": 0.2358,
      "step": 213
    },
    {
      "epoch": 0.9163655827646193,
      "grad_norm": 0.3155864179134369,
      "learning_rate": 1.0294117647058824e-05,
      "loss": 0.2525,
      "step": 214
    },
    {
      "epoch": 0.9206476649270708,
      "grad_norm": 0.3156728744506836,
      "learning_rate": 9.803921568627451e-06,
      "loss": 0.2353,
      "step": 215
    },
    {
      "epoch": 0.9249297470895222,
      "grad_norm": 0.2930629551410675,
      "learning_rate": 9.31372549019608e-06,
      "loss": 0.2465,
      "step": 216
    },
    {
      "epoch": 0.9292118292519738,
      "grad_norm": 0.27664482593536377,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.2282,
      "step": 217
    },
    {
      "epoch": 0.9334939114144253,
      "grad_norm": 0.29967495799064636,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.2238,
      "step": 218
    },
    {
      "epoch": 0.9377759935768768,
      "grad_norm": 0.36104142665863037,
      "learning_rate": 7.84313725490196e-06,
      "loss": 0.2694,
      "step": 219
    },
    {
      "epoch": 0.9420580757393282,
      "grad_norm": 0.2655661106109619,
      "learning_rate": 7.3529411764705884e-06,
      "loss": 0.1941,
      "step": 220
    },
    {
      "epoch": 0.9463401579017797,
      "grad_norm": 0.3580377399921417,
      "learning_rate": 6.862745098039216e-06,
      "loss": 0.2869,
      "step": 221
    },
    {
      "epoch": 0.9506222400642312,
      "grad_norm": 0.2969386875629425,
      "learning_rate": 6.372549019607843e-06,
      "loss": 0.2383,
      "step": 222
    },
    {
      "epoch": 0.9549043222266828,
      "grad_norm": 0.3826366066932678,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.2347,
      "step": 223
    },
    {
      "epoch": 0.9591864043891342,
      "grad_norm": 0.2728925049304962,
      "learning_rate": 5.392156862745099e-06,
      "loss": 0.222,
      "step": 224
    },
    {
      "epoch": 0.9634684865515857,
      "grad_norm": 0.274934858083725,
      "learning_rate": 4.901960784313726e-06,
      "loss": 0.2248,
      "step": 225
    },
    {
      "epoch": 0.9677505687140372,
      "grad_norm": 0.3285166025161743,
      "learning_rate": 4.411764705882353e-06,
      "loss": 0.2686,
      "step": 226
    },
    {
      "epoch": 0.9720326508764887,
      "grad_norm": 0.28881922364234924,
      "learning_rate": 3.92156862745098e-06,
      "loss": 0.2061,
      "step": 227
    },
    {
      "epoch": 0.9763147330389402,
      "grad_norm": 0.32320910692214966,
      "learning_rate": 3.431372549019608e-06,
      "loss": 0.2577,
      "step": 228
    },
    {
      "epoch": 0.9805968152013916,
      "grad_norm": 0.30604419112205505,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 0.2411,
      "step": 229
    },
    {
      "epoch": 0.9848788973638432,
      "grad_norm": 0.3247486352920532,
      "learning_rate": 2.450980392156863e-06,
      "loss": 0.2319,
      "step": 230
    },
    {
      "epoch": 0.9891609795262947,
      "grad_norm": 0.3221021890640259,
      "learning_rate": 1.96078431372549e-06,
      "loss": 0.2147,
      "step": 231
    },
    {
      "epoch": 0.9934430616887462,
      "grad_norm": 0.31727492809295654,
      "learning_rate": 1.4705882352941177e-06,
      "loss": 0.2233,
      "step": 232
    },
    {
      "epoch": 0.9977251438511976,
      "grad_norm": 0.3005445897579193,
      "learning_rate": 9.80392156862745e-07,
      "loss": 0.1958,
      "step": 233
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.49040892720222473,
      "learning_rate": 4.901960784313725e-07,
      "loss": 0.2109,
      "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
}