{
  "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": 2.1201584339141846,
      "learning_rate": 0.0,
      "loss": 1.3954,
      "step": 1
    },
    {
      "epoch": 0.008564164324902984,
      "grad_norm": 1.642530918121338,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.1107,
      "step": 2
    },
    {
      "epoch": 0.012846246487354477,
      "grad_norm": 1.9669904708862305,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.3024,
      "step": 3
    },
    {
      "epoch": 0.017128328649805968,
      "grad_norm": 1.974590539932251,
      "learning_rate": 1e-05,
      "loss": 1.2922,
      "step": 4
    },
    {
      "epoch": 0.02141041081225746,
      "grad_norm": 1.991250991821289,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 1.3049,
      "step": 5
    },
    {
      "epoch": 0.025692492974708953,
      "grad_norm": 1.8808647394180298,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.2983,
      "step": 6
    },
    {
      "epoch": 0.029974575137160443,
      "grad_norm": 1.7501676082611084,
      "learning_rate": 2e-05,
      "loss": 1.1921,
      "step": 7
    },
    {
      "epoch": 0.034256657299611935,
      "grad_norm": 1.0341113805770874,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.9469,
      "step": 8
    },
    {
      "epoch": 0.03853873946206343,
      "grad_norm": 0.9029558300971985,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.9475,
      "step": 9
    },
    {
      "epoch": 0.04282082162451492,
      "grad_norm": 0.7903841733932495,
      "learning_rate": 3e-05,
      "loss": 0.9012,
      "step": 10
    },
    {
      "epoch": 0.04710290378696641,
      "grad_norm": 0.7263166904449463,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.8324,
      "step": 11
    },
    {
      "epoch": 0.051384985949417906,
      "grad_norm": 0.7035966515541077,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.7309,
      "step": 12
    },
    {
      "epoch": 0.055667068111869396,
      "grad_norm": 0.597355306148529,
      "learning_rate": 4e-05,
      "loss": 0.6999,
      "step": 13
    },
    {
      "epoch": 0.059949150274320885,
      "grad_norm": 0.610143780708313,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.6403,
      "step": 14
    },
    {
      "epoch": 0.06423123243677238,
      "grad_norm": 0.5835250616073608,
      "learning_rate": 4.666666666666667e-05,
      "loss": 0.6454,
      "step": 15
    },
    {
      "epoch": 0.06851331459922387,
      "grad_norm": 0.6275021433830261,
      "learning_rate": 5e-05,
      "loss": 0.5325,
      "step": 16
    },
    {
      "epoch": 0.07279539676167536,
      "grad_norm": 0.5470361709594727,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.4997,
      "step": 17
    },
    {
      "epoch": 0.07707747892412686,
      "grad_norm": 0.6093178987503052,
      "learning_rate": 5.666666666666667e-05,
      "loss": 0.5068,
      "step": 18
    },
    {
      "epoch": 0.08135956108657835,
      "grad_norm": 0.46858343482017517,
      "learning_rate": 6e-05,
      "loss": 0.3882,
      "step": 19
    },
    {
      "epoch": 0.08564164324902984,
      "grad_norm": 0.5791119337081909,
      "learning_rate": 6.333333333333333e-05,
      "loss": 0.4179,
      "step": 20
    },
    {
      "epoch": 0.08992372541148133,
      "grad_norm": 0.4088153839111328,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.3728,
      "step": 21
    },
    {
      "epoch": 0.09420580757393282,
      "grad_norm": 0.460022509098053,
      "learning_rate": 7e-05,
      "loss": 0.3339,
      "step": 22
    },
    {
      "epoch": 0.09848788973638432,
      "grad_norm": 0.38594093918800354,
      "learning_rate": 7.333333333333333e-05,
      "loss": 0.3499,
      "step": 23
    },
    {
      "epoch": 0.10276997189883581,
      "grad_norm": 0.3029627501964569,
      "learning_rate": 7.666666666666667e-05,
      "loss": 0.3318,
      "step": 24
    },
    {
      "epoch": 0.1070520540612873,
      "grad_norm": 0.3492797613143921,
      "learning_rate": 8e-05,
      "loss": 0.2966,
      "step": 25
    },
    {
      "epoch": 0.11133413622373879,
      "grad_norm": 0.3025495409965515,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.2981,
      "step": 26
    },
    {
      "epoch": 0.11561621838619028,
      "grad_norm": 0.31970080733299255,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.3114,
      "step": 27
    },
    {
      "epoch": 0.11989830054864177,
      "grad_norm": 0.40258461236953735,
      "learning_rate": 9e-05,
      "loss": 0.3558,
      "step": 28
    },
    {
      "epoch": 0.12418038271109327,
      "grad_norm": 0.2629145383834839,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.3108,
      "step": 29
    },
    {
      "epoch": 0.12846246487354476,
      "grad_norm": 0.3107101321220398,
      "learning_rate": 9.666666666666667e-05,
      "loss": 0.3469,
      "step": 30
    },
    {
      "epoch": 0.13274454703599625,
      "grad_norm": 0.2993117868900299,
      "learning_rate": 0.0001,
      "loss": 0.3549,
      "step": 31
    },
    {
      "epoch": 0.13702662919844774,
      "grad_norm": 0.252559095621109,
      "learning_rate": 9.950980392156864e-05,
      "loss": 0.3138,
      "step": 32
    },
    {
      "epoch": 0.14130871136089923,
      "grad_norm": 0.28241539001464844,
      "learning_rate": 9.901960784313727e-05,
      "loss": 0.2933,
      "step": 33
    },
    {
      "epoch": 0.14559079352335072,
      "grad_norm": 0.2623656392097473,
      "learning_rate": 9.852941176470589e-05,
      "loss": 0.3187,
      "step": 34
    },
    {
      "epoch": 0.1498728756858022,
      "grad_norm": 0.2127358466386795,
      "learning_rate": 9.80392156862745e-05,
      "loss": 0.305,
      "step": 35
    },
    {
      "epoch": 0.15415495784825373,
      "grad_norm": 0.2847974896430969,
      "learning_rate": 9.754901960784314e-05,
      "loss": 0.2979,
      "step": 36
    },
    {
      "epoch": 0.15843704001070522,
      "grad_norm": 0.2522662580013275,
      "learning_rate": 9.705882352941177e-05,
      "loss": 0.3092,
      "step": 37
    },
    {
      "epoch": 0.1627191221731567,
      "grad_norm": 0.23280668258666992,
      "learning_rate": 9.65686274509804e-05,
      "loss": 0.2787,
      "step": 38
    },
    {
      "epoch": 0.1670012043356082,
      "grad_norm": 0.21782170236110687,
      "learning_rate": 9.607843137254903e-05,
      "loss": 0.2908,
      "step": 39
    },
    {
      "epoch": 0.17128328649805968,
      "grad_norm": 0.26517942547798157,
      "learning_rate": 9.558823529411765e-05,
      "loss": 0.296,
      "step": 40
    },
    {
      "epoch": 0.17556536866051117,
      "grad_norm": 0.24453642964363098,
      "learning_rate": 9.509803921568627e-05,
      "loss": 0.2679,
      "step": 41
    },
    {
      "epoch": 0.17984745082296266,
      "grad_norm": 0.2437601089477539,
      "learning_rate": 9.46078431372549e-05,
      "loss": 0.2774,
      "step": 42
    },
    {
      "epoch": 0.18412953298541415,
      "grad_norm": 0.23632746934890747,
      "learning_rate": 9.411764705882353e-05,
      "loss": 0.286,
      "step": 43
    },
    {
      "epoch": 0.18841161514786564,
      "grad_norm": 0.23719914257526398,
      "learning_rate": 9.362745098039217e-05,
      "loss": 0.2685,
      "step": 44
    },
    {
      "epoch": 0.19269369731031713,
      "grad_norm": 0.24241948127746582,
      "learning_rate": 9.313725490196079e-05,
      "loss": 0.2437,
      "step": 45
    },
    {
      "epoch": 0.19697577947276865,
      "grad_norm": 0.23390215635299683,
      "learning_rate": 9.264705882352942e-05,
      "loss": 0.2655,
      "step": 46
    },
    {
      "epoch": 0.20125786163522014,
      "grad_norm": 0.2400883138179779,
      "learning_rate": 9.215686274509804e-05,
      "loss": 0.2803,
      "step": 47
    },
    {
      "epoch": 0.20553994379767163,
      "grad_norm": 0.2557249665260315,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.272,
      "step": 48
    },
    {
      "epoch": 0.20982202596012312,
      "grad_norm": 0.2652246654033661,
      "learning_rate": 9.11764705882353e-05,
      "loss": 0.2948,
      "step": 49
    },
    {
      "epoch": 0.2141041081225746,
      "grad_norm": 0.26117566227912903,
      "learning_rate": 9.068627450980393e-05,
      "loss": 0.295,
      "step": 50
    },
    {
      "epoch": 0.2183861902850261,
      "grad_norm": 0.24146981537342072,
      "learning_rate": 9.019607843137255e-05,
      "loss": 0.2744,
      "step": 51
    },
    {
      "epoch": 0.22266827244747758,
      "grad_norm": 0.2520713210105896,
      "learning_rate": 8.970588235294118e-05,
      "loss": 0.2754,
      "step": 52
    },
    {
      "epoch": 0.22695035460992907,
      "grad_norm": 0.21968355774879456,
      "learning_rate": 8.921568627450981e-05,
      "loss": 0.2605,
      "step": 53
    },
    {
      "epoch": 0.23123243677238056,
      "grad_norm": 0.25200408697128296,
      "learning_rate": 8.872549019607843e-05,
      "loss": 0.302,
      "step": 54
    },
    {
      "epoch": 0.23551451893483205,
      "grad_norm": 0.2303711473941803,
      "learning_rate": 8.823529411764706e-05,
      "loss": 0.2522,
      "step": 55
    },
    {
      "epoch": 0.23979660109728354,
      "grad_norm": 0.24254676699638367,
      "learning_rate": 8.774509803921568e-05,
      "loss": 0.2768,
      "step": 56
    },
    {
      "epoch": 0.24407868325973506,
      "grad_norm": 0.2681991457939148,
      "learning_rate": 8.725490196078432e-05,
      "loss": 0.2719,
      "step": 57
    },
    {
      "epoch": 0.24836076542218655,
      "grad_norm": 0.23361504077911377,
      "learning_rate": 8.676470588235295e-05,
      "loss": 0.2956,
      "step": 58
    },
    {
      "epoch": 0.252642847584638,
      "grad_norm": 0.2468583881855011,
      "learning_rate": 8.627450980392158e-05,
      "loss": 0.2669,
      "step": 59
    },
    {
      "epoch": 0.2569249297470895,
      "grad_norm": 0.22490642964839935,
      "learning_rate": 8.57843137254902e-05,
      "loss": 0.2798,
      "step": 60
    },
    {
      "epoch": 0.261207011909541,
      "grad_norm": 0.23348553478717804,
      "learning_rate": 8.529411764705883e-05,
      "loss": 0.2715,
      "step": 61
    },
    {
      "epoch": 0.2654890940719925,
      "grad_norm": 0.22062627971172333,
      "learning_rate": 8.480392156862745e-05,
      "loss": 0.2718,
      "step": 62
    },
    {
      "epoch": 0.269771176234444,
      "grad_norm": 0.264909029006958,
      "learning_rate": 8.431372549019608e-05,
      "loss": 0.2892,
      "step": 63
    },
    {
      "epoch": 0.2740532583968955,
      "grad_norm": 0.23317475616931915,
      "learning_rate": 8.382352941176471e-05,
      "loss": 0.2371,
      "step": 64
    },
    {
      "epoch": 0.278335340559347,
      "grad_norm": 0.24755944311618805,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.2621,
      "step": 65
    },
    {
      "epoch": 0.28261742272179846,
      "grad_norm": 0.23592090606689453,
      "learning_rate": 8.284313725490198e-05,
      "loss": 0.248,
      "step": 66
    },
    {
      "epoch": 0.28689950488425,
      "grad_norm": 0.26305922865867615,
      "learning_rate": 8.23529411764706e-05,
      "loss": 0.307,
      "step": 67
    },
    {
      "epoch": 0.29118158704670144,
      "grad_norm": 0.26928380131721497,
      "learning_rate": 8.186274509803921e-05,
      "loss": 0.3037,
      "step": 68
    },
    {
      "epoch": 0.29546366920915296,
      "grad_norm": 0.24364902079105377,
      "learning_rate": 8.137254901960785e-05,
      "loss": 0.2645,
      "step": 69
    },
    {
      "epoch": 0.2997457513716044,
      "grad_norm": 0.24957282841205597,
      "learning_rate": 8.088235294117648e-05,
      "loss": 0.269,
      "step": 70
    },
    {
      "epoch": 0.30402783353405594,
      "grad_norm": 0.27414730191230774,
      "learning_rate": 8.039215686274511e-05,
      "loss": 0.2946,
      "step": 71
    },
    {
      "epoch": 0.30830991569650745,
      "grad_norm": 0.2614569664001465,
      "learning_rate": 7.990196078431373e-05,
      "loss": 0.2589,
      "step": 72
    },
    {
      "epoch": 0.3125919978589589,
      "grad_norm": 0.23539189994335175,
      "learning_rate": 7.941176470588235e-05,
      "loss": 0.2668,
      "step": 73
    },
    {
      "epoch": 0.31687408002141043,
      "grad_norm": 0.2485891580581665,
      "learning_rate": 7.892156862745098e-05,
      "loss": 0.2457,
      "step": 74
    },
    {
      "epoch": 0.3211561621838619,
      "grad_norm": 0.267423152923584,
      "learning_rate": 7.843137254901961e-05,
      "loss": 0.2356,
      "step": 75
    },
    {
      "epoch": 0.3254382443463134,
      "grad_norm": 0.29359403252601624,
      "learning_rate": 7.794117647058824e-05,
      "loss": 0.3035,
      "step": 76
    },
    {
      "epoch": 0.32972032650876487,
      "grad_norm": 0.28489920496940613,
      "learning_rate": 7.745098039215687e-05,
      "loss": 0.2533,
      "step": 77
    },
    {
      "epoch": 0.3340024086712164,
      "grad_norm": 0.2578640282154083,
      "learning_rate": 7.696078431372549e-05,
      "loss": 0.2465,
      "step": 78
    },
    {
      "epoch": 0.33828449083366785,
      "grad_norm": 0.2692528963088989,
      "learning_rate": 7.647058823529411e-05,
      "loss": 0.2907,
      "step": 79
    },
    {
      "epoch": 0.34256657299611937,
      "grad_norm": 0.2575438916683197,
      "learning_rate": 7.598039215686274e-05,
      "loss": 0.246,
      "step": 80
    },
    {
      "epoch": 0.34684865515857083,
      "grad_norm": 0.309518426656723,
      "learning_rate": 7.549019607843137e-05,
      "loss": 0.2634,
      "step": 81
    },
    {
      "epoch": 0.35113073732102235,
      "grad_norm": 0.2875613570213318,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.2614,
      "step": 82
    },
    {
      "epoch": 0.35541281948347386,
      "grad_norm": 0.2768095135688782,
      "learning_rate": 7.450980392156864e-05,
      "loss": 0.2686,
      "step": 83
    },
    {
      "epoch": 0.3596949016459253,
      "grad_norm": 0.22313964366912842,
      "learning_rate": 7.401960784313726e-05,
      "loss": 0.1998,
      "step": 84
    },
    {
      "epoch": 0.36397698380837684,
      "grad_norm": 0.2706241011619568,
      "learning_rate": 7.352941176470589e-05,
      "loss": 0.2476,
      "step": 85
    },
    {
      "epoch": 0.3682590659708283,
      "grad_norm": 0.24844884872436523,
      "learning_rate": 7.303921568627451e-05,
      "loss": 0.2431,
      "step": 86
    },
    {
      "epoch": 0.3725411481332798,
      "grad_norm": 0.2623468041419983,
      "learning_rate": 7.254901960784314e-05,
      "loss": 0.2398,
      "step": 87
    },
    {
      "epoch": 0.3768232302957313,
      "grad_norm": 0.29353752732276917,
      "learning_rate": 7.205882352941177e-05,
      "loss": 0.289,
      "step": 88
    },
    {
      "epoch": 0.3811053124581828,
      "grad_norm": 0.23162683844566345,
      "learning_rate": 7.156862745098039e-05,
      "loss": 0.26,
      "step": 89
    },
    {
      "epoch": 0.38538739462063426,
      "grad_norm": 0.28125250339508057,
      "learning_rate": 7.107843137254902e-05,
      "loss": 0.2618,
      "step": 90
    },
    {
      "epoch": 0.3896694767830858,
      "grad_norm": 0.25823014974594116,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.2563,
      "step": 91
    },
    {
      "epoch": 0.3939515589455373,
      "grad_norm": 0.24386447668075562,
      "learning_rate": 7.009803921568627e-05,
      "loss": 0.2354,
      "step": 92
    },
    {
      "epoch": 0.39823364110798876,
      "grad_norm": 0.26175400614738464,
      "learning_rate": 6.96078431372549e-05,
      "loss": 0.2507,
      "step": 93
    },
    {
      "epoch": 0.4025157232704403,
      "grad_norm": 0.24113620817661285,
      "learning_rate": 6.911764705882354e-05,
      "loss": 0.2269,
      "step": 94
    },
    {
      "epoch": 0.40679780543289173,
      "grad_norm": 0.2782343327999115,
      "learning_rate": 6.862745098039216e-05,
      "loss": 0.2708,
      "step": 95
    },
    {
      "epoch": 0.41107988759534325,
      "grad_norm": 0.2693990468978882,
      "learning_rate": 6.813725490196079e-05,
      "loss": 0.2796,
      "step": 96
    },
    {
      "epoch": 0.4153619697577947,
      "grad_norm": 0.2997571527957916,
      "learning_rate": 6.764705882352942e-05,
      "loss": 0.2574,
      "step": 97
    },
    {
      "epoch": 0.41964405192024623,
      "grad_norm": 0.2795896530151367,
      "learning_rate": 6.715686274509804e-05,
      "loss": 0.2724,
      "step": 98
    },
    {
      "epoch": 0.4239261340826977,
      "grad_norm": 0.24898579716682434,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.2391,
      "step": 99
    },
    {
      "epoch": 0.4282082162451492,
      "grad_norm": 0.2840297818183899,
      "learning_rate": 6.61764705882353e-05,
      "loss": 0.2745,
      "step": 100
    },
    {
      "epoch": 0.43249029840760067,
      "grad_norm": 0.2686343491077423,
      "learning_rate": 6.568627450980392e-05,
      "loss": 0.2444,
      "step": 101
    },
    {
      "epoch": 0.4367723805700522,
      "grad_norm": 0.2764686942100525,
      "learning_rate": 6.519607843137255e-05,
      "loss": 0.2655,
      "step": 102
    },
    {
      "epoch": 0.4410544627325037,
      "grad_norm": 0.2911969721317291,
      "learning_rate": 6.470588235294118e-05,
      "loss": 0.2541,
      "step": 103
    },
    {
      "epoch": 0.44533654489495517,
      "grad_norm": 0.28445762395858765,
      "learning_rate": 6.421568627450982e-05,
      "loss": 0.2494,
      "step": 104
    },
    {
      "epoch": 0.4496186270574067,
      "grad_norm": 0.27755123376846313,
      "learning_rate": 6.372549019607843e-05,
      "loss": 0.2571,
      "step": 105
    },
    {
      "epoch": 0.45390070921985815,
      "grad_norm": 0.33207792043685913,
      "learning_rate": 6.323529411764705e-05,
      "loss": 0.3049,
      "step": 106
    },
    {
      "epoch": 0.45818279138230966,
      "grad_norm": 0.2805379629135132,
      "learning_rate": 6.274509803921569e-05,
      "loss": 0.2016,
      "step": 107
    },
    {
      "epoch": 0.4624648735447611,
      "grad_norm": 0.27627864480018616,
      "learning_rate": 6.225490196078432e-05,
      "loss": 0.2409,
      "step": 108
    },
    {
      "epoch": 0.46674695570721264,
      "grad_norm": 0.276563823223114,
      "learning_rate": 6.176470588235295e-05,
      "loss": 0.2889,
      "step": 109
    },
    {
      "epoch": 0.4710290378696641,
      "grad_norm": 0.273502379655838,
      "learning_rate": 6.127450980392158e-05,
      "loss": 0.2493,
      "step": 110
    },
    {
      "epoch": 0.4753111200321156,
      "grad_norm": 0.24963822960853577,
      "learning_rate": 6.078431372549019e-05,
      "loss": 0.232,
      "step": 111
    },
    {
      "epoch": 0.4795932021945671,
      "grad_norm": 0.2647972106933594,
      "learning_rate": 6.0294117647058825e-05,
      "loss": 0.238,
      "step": 112
    },
    {
      "epoch": 0.4838752843570186,
      "grad_norm": 0.2622900903224945,
      "learning_rate": 5.980392156862745e-05,
      "loss": 0.239,
      "step": 113
    },
    {
      "epoch": 0.4881573665194701,
      "grad_norm": 0.3029134273529053,
      "learning_rate": 5.931372549019608e-05,
      "loss": 0.2478,
      "step": 114
    },
    {
      "epoch": 0.4924394486819216,
      "grad_norm": 0.2673580050468445,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.2628,
      "step": 115
    },
    {
      "epoch": 0.4967215308443731,
      "grad_norm": 0.28343433141708374,
      "learning_rate": 5.833333333333334e-05,
      "loss": 0.229,
      "step": 116
    },
    {
      "epoch": 0.5010036130068246,
      "grad_norm": 0.2858482599258423,
      "learning_rate": 5.784313725490197e-05,
      "loss": 0.2442,
      "step": 117
    },
    {
      "epoch": 0.505285695169276,
      "grad_norm": 0.28911155462265015,
      "learning_rate": 5.735294117647059e-05,
      "loss": 0.2715,
      "step": 118
    },
    {
      "epoch": 0.5095677773317275,
      "grad_norm": 0.2598050534725189,
      "learning_rate": 5.6862745098039215e-05,
      "loss": 0.226,
      "step": 119
    },
    {
      "epoch": 0.513849859494179,
      "grad_norm": 0.24481424689292908,
      "learning_rate": 5.637254901960785e-05,
      "loss": 0.2387,
      "step": 120
    },
    {
      "epoch": 0.5181319416566306,
      "grad_norm": 0.28961485624313354,
      "learning_rate": 5.588235294117647e-05,
      "loss": 0.2497,
      "step": 121
    },
    {
      "epoch": 0.522414023819082,
      "grad_norm": 0.25400567054748535,
      "learning_rate": 5.5392156862745104e-05,
      "loss": 0.2387,
      "step": 122
    },
    {
      "epoch": 0.5266961059815335,
      "grad_norm": 0.277719110250473,
      "learning_rate": 5.490196078431373e-05,
      "loss": 0.2354,
      "step": 123
    },
    {
      "epoch": 0.530978188143985,
      "grad_norm": 0.3029388189315796,
      "learning_rate": 5.441176470588235e-05,
      "loss": 0.2931,
      "step": 124
    },
    {
      "epoch": 0.5352602703064365,
      "grad_norm": 0.30095577239990234,
      "learning_rate": 5.392156862745098e-05,
      "loss": 0.2252,
      "step": 125
    },
    {
      "epoch": 0.539542352468888,
      "grad_norm": 0.2955123484134674,
      "learning_rate": 5.343137254901961e-05,
      "loss": 0.2677,
      "step": 126
    },
    {
      "epoch": 0.5438244346313394,
      "grad_norm": 0.24744829535484314,
      "learning_rate": 5.294117647058824e-05,
      "loss": 0.226,
      "step": 127
    },
    {
      "epoch": 0.548106516793791,
      "grad_norm": 0.27184242010116577,
      "learning_rate": 5.245098039215687e-05,
      "loss": 0.2354,
      "step": 128
    },
    {
      "epoch": 0.5523885989562425,
      "grad_norm": 0.30591636896133423,
      "learning_rate": 5.1960784313725495e-05,
      "loss": 0.244,
      "step": 129
    },
    {
      "epoch": 0.556670681118694,
      "grad_norm": 0.2827044129371643,
      "learning_rate": 5.147058823529411e-05,
      "loss": 0.29,
      "step": 130
    },
    {
      "epoch": 0.5609527632811454,
      "grad_norm": 0.30164632201194763,
      "learning_rate": 5.0980392156862745e-05,
      "loss": 0.2403,
      "step": 131
    },
    {
      "epoch": 0.5652348454435969,
      "grad_norm": 0.2726845145225525,
      "learning_rate": 5.049019607843137e-05,
      "loss": 0.235,
      "step": 132
    },
    {
      "epoch": 0.5695169276060484,
      "grad_norm": 0.3138998746871948,
      "learning_rate": 5e-05,
      "loss": 0.2215,
      "step": 133
    },
    {
      "epoch": 0.5737990097685,
      "grad_norm": 0.3027787506580353,
      "learning_rate": 4.9509803921568634e-05,
      "loss": 0.2885,
      "step": 134
    },
    {
      "epoch": 0.5780810919309515,
      "grad_norm": 0.27734512090682983,
      "learning_rate": 4.901960784313725e-05,
      "loss": 0.223,
      "step": 135
    },
    {
      "epoch": 0.5823631740934029,
      "grad_norm": 0.2831007242202759,
      "learning_rate": 4.8529411764705885e-05,
      "loss": 0.2486,
      "step": 136
    },
    {
      "epoch": 0.5866452562558544,
      "grad_norm": 0.27249404788017273,
      "learning_rate": 4.803921568627452e-05,
      "loss": 0.2481,
      "step": 137
    },
    {
      "epoch": 0.5909273384183059,
      "grad_norm": 0.300654798746109,
      "learning_rate": 4.7549019607843135e-05,
      "loss": 0.2549,
      "step": 138
    },
    {
      "epoch": 0.5952094205807574,
      "grad_norm": 0.26843512058258057,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.2332,
      "step": 139
    },
    {
      "epoch": 0.5994915027432088,
      "grad_norm": 0.25265592336654663,
      "learning_rate": 4.656862745098039e-05,
      "loss": 0.2131,
      "step": 140
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.27025896310806274,
      "learning_rate": 4.607843137254902e-05,
      "loss": 0.245,
      "step": 141
    },
    {
      "epoch": 0.6080556670681119,
      "grad_norm": 0.262179970741272,
      "learning_rate": 4.558823529411765e-05,
      "loss": 0.2138,
      "step": 142
    },
    {
      "epoch": 0.6123377492305634,
      "grad_norm": 0.28520095348358154,
      "learning_rate": 4.5098039215686275e-05,
      "loss": 0.2499,
      "step": 143
    },
    {
      "epoch": 0.6166198313930149,
      "grad_norm": 0.32692384719848633,
      "learning_rate": 4.460784313725491e-05,
      "loss": 0.2636,
      "step": 144
    },
    {
      "epoch": 0.6209019135554663,
      "grad_norm": 0.28875231742858887,
      "learning_rate": 4.411764705882353e-05,
      "loss": 0.244,
      "step": 145
    },
    {
      "epoch": 0.6251839957179178,
      "grad_norm": 0.2917136549949646,
      "learning_rate": 4.362745098039216e-05,
      "loss": 0.2484,
      "step": 146
    },
    {
      "epoch": 0.6294660778803693,
      "grad_norm": 0.28060460090637207,
      "learning_rate": 4.313725490196079e-05,
      "loss": 0.2606,
      "step": 147
    },
    {
      "epoch": 0.6337481600428209,
      "grad_norm": 0.29203474521636963,
      "learning_rate": 4.2647058823529415e-05,
      "loss": 0.2508,
      "step": 148
    },
    {
      "epoch": 0.6380302422052723,
      "grad_norm": 0.323335736989975,
      "learning_rate": 4.215686274509804e-05,
      "loss": 0.2175,
      "step": 149
    },
    {
      "epoch": 0.6423123243677238,
      "grad_norm": 0.27991968393325806,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.2159,
      "step": 150
    },
    {
      "epoch": 0.6465944065301753,
      "grad_norm": 0.3103591501712799,
      "learning_rate": 4.11764705882353e-05,
      "loss": 0.245,
      "step": 151
    },
    {
      "epoch": 0.6508764886926268,
      "grad_norm": 0.31740787625312805,
      "learning_rate": 4.068627450980392e-05,
      "loss": 0.2298,
      "step": 152
    },
    {
      "epoch": 0.6551585708550783,
      "grad_norm": 0.30606138706207275,
      "learning_rate": 4.0196078431372555e-05,
      "loss": 0.2318,
      "step": 153
    },
    {
      "epoch": 0.6594406530175297,
      "grad_norm": 0.3375915288925171,
      "learning_rate": 3.970588235294117e-05,
      "loss": 0.2671,
      "step": 154
    },
    {
      "epoch": 0.6637227351799813,
      "grad_norm": 0.32568517327308655,
      "learning_rate": 3.9215686274509805e-05,
      "loss": 0.2363,
      "step": 155
    },
    {
      "epoch": 0.6680048173424328,
      "grad_norm": 0.31305477023124695,
      "learning_rate": 3.872549019607844e-05,
      "loss": 0.2492,
      "step": 156
    },
    {
      "epoch": 0.6722868995048843,
      "grad_norm": 0.3048565983772278,
      "learning_rate": 3.8235294117647055e-05,
      "loss": 0.2903,
      "step": 157
    },
    {
      "epoch": 0.6765689816673357,
      "grad_norm": 0.3204030394554138,
      "learning_rate": 3.774509803921569e-05,
      "loss": 0.2583,
      "step": 158
    },
    {
      "epoch": 0.6808510638297872,
      "grad_norm": 0.2854422926902771,
      "learning_rate": 3.725490196078432e-05,
      "loss": 0.2568,
      "step": 159
    },
    {
      "epoch": 0.6851331459922387,
      "grad_norm": 0.2798788845539093,
      "learning_rate": 3.6764705882352945e-05,
      "loss": 0.2318,
      "step": 160
    },
    {
      "epoch": 0.6894152281546903,
      "grad_norm": 0.2955876588821411,
      "learning_rate": 3.627450980392157e-05,
      "loss": 0.1954,
      "step": 161
    },
    {
      "epoch": 0.6936973103171417,
      "grad_norm": 0.27800989151000977,
      "learning_rate": 3.5784313725490195e-05,
      "loss": 0.2857,
      "step": 162
    },
    {
      "epoch": 0.6979793924795932,
      "grad_norm": 0.23555797338485718,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.2078,
      "step": 163
    },
    {
      "epoch": 0.7022614746420447,
      "grad_norm": 0.2481900304555893,
      "learning_rate": 3.480392156862745e-05,
      "loss": 0.2028,
      "step": 164
    },
    {
      "epoch": 0.7065435568044962,
      "grad_norm": 0.2711412310600281,
      "learning_rate": 3.431372549019608e-05,
      "loss": 0.2289,
      "step": 165
    },
    {
      "epoch": 0.7108256389669477,
      "grad_norm": 0.2945784330368042,
      "learning_rate": 3.382352941176471e-05,
      "loss": 0.2286,
      "step": 166
    },
    {
      "epoch": 0.7151077211293991,
      "grad_norm": 0.2895250618457794,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.2541,
      "step": 167
    },
    {
      "epoch": 0.7193898032918506,
      "grad_norm": 0.27034303545951843,
      "learning_rate": 3.284313725490196e-05,
      "loss": 0.2265,
      "step": 168
    },
    {
      "epoch": 0.7236718854543022,
      "grad_norm": 0.3107519745826721,
      "learning_rate": 3.235294117647059e-05,
      "loss": 0.2009,
      "step": 169
    },
    {
      "epoch": 0.7279539676167537,
      "grad_norm": 0.2937147915363312,
      "learning_rate": 3.186274509803922e-05,
      "loss": 0.2359,
      "step": 170
    },
    {
      "epoch": 0.7322360497792051,
      "grad_norm": 0.29187825322151184,
      "learning_rate": 3.137254901960784e-05,
      "loss": 0.2482,
      "step": 171
    },
    {
      "epoch": 0.7365181319416566,
      "grad_norm": 0.3703550398349762,
      "learning_rate": 3.0882352941176475e-05,
      "loss": 0.2206,
      "step": 172
    },
    {
      "epoch": 0.7408002141041081,
      "grad_norm": 0.3378922939300537,
      "learning_rate": 3.0392156862745097e-05,
      "loss": 0.2726,
      "step": 173
    },
    {
      "epoch": 0.7450822962665596,
      "grad_norm": 0.3115370571613312,
      "learning_rate": 2.9901960784313725e-05,
      "loss": 0.2262,
      "step": 174
    },
    {
      "epoch": 0.7493643784290112,
      "grad_norm": 0.28524336218833923,
      "learning_rate": 2.9411764705882354e-05,
      "loss": 0.2324,
      "step": 175
    },
    {
      "epoch": 0.7536464605914626,
      "grad_norm": 0.31144171953201294,
      "learning_rate": 2.8921568627450986e-05,
      "loss": 0.2524,
      "step": 176
    },
    {
      "epoch": 0.7579285427539141,
      "grad_norm": 0.28729379177093506,
      "learning_rate": 2.8431372549019608e-05,
      "loss": 0.2298,
      "step": 177
    },
    {
      "epoch": 0.7622106249163656,
      "grad_norm": 0.29088258743286133,
      "learning_rate": 2.7941176470588236e-05,
      "loss": 0.2252,
      "step": 178
    },
    {
      "epoch": 0.7664927070788171,
      "grad_norm": 0.2751845121383667,
      "learning_rate": 2.7450980392156865e-05,
      "loss": 0.2201,
      "step": 179
    },
    {
      "epoch": 0.7707747892412685,
      "grad_norm": 0.28968045115470886,
      "learning_rate": 2.696078431372549e-05,
      "loss": 0.2062,
      "step": 180
    },
    {
      "epoch": 0.77505687140372,
      "grad_norm": 0.4162120223045349,
      "learning_rate": 2.647058823529412e-05,
      "loss": 0.263,
      "step": 181
    },
    {
      "epoch": 0.7793389535661716,
      "grad_norm": 0.30583426356315613,
      "learning_rate": 2.5980392156862747e-05,
      "loss": 0.2186,
      "step": 182
    },
    {
      "epoch": 0.7836210357286231,
      "grad_norm": 0.2964443564414978,
      "learning_rate": 2.5490196078431373e-05,
      "loss": 0.2484,
      "step": 183
    },
    {
      "epoch": 0.7879031178910746,
      "grad_norm": 0.3738739788532257,
      "learning_rate": 2.5e-05,
      "loss": 0.2564,
      "step": 184
    },
    {
      "epoch": 0.792185200053526,
      "grad_norm": 0.313554048538208,
      "learning_rate": 2.4509803921568626e-05,
      "loss": 0.2685,
      "step": 185
    },
    {
      "epoch": 0.7964672822159775,
      "grad_norm": 0.29994913935661316,
      "learning_rate": 2.401960784313726e-05,
      "loss": 0.2649,
      "step": 186
    },
    {
      "epoch": 0.800749364378429,
      "grad_norm": 0.29549285769462585,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.2323,
      "step": 187
    },
    {
      "epoch": 0.8050314465408805,
      "grad_norm": 0.2795826196670532,
      "learning_rate": 2.303921568627451e-05,
      "loss": 0.2473,
      "step": 188
    },
    {
      "epoch": 0.809313528703332,
      "grad_norm": 0.2648061215877533,
      "learning_rate": 2.2549019607843138e-05,
      "loss": 0.1964,
      "step": 189
    },
    {
      "epoch": 0.8135956108657835,
      "grad_norm": 0.3009537160396576,
      "learning_rate": 2.2058823529411766e-05,
      "loss": 0.2255,
      "step": 190
    },
    {
      "epoch": 0.817877693028235,
      "grad_norm": 0.2910759150981903,
      "learning_rate": 2.1568627450980395e-05,
      "loss": 0.1923,
      "step": 191
    },
    {
      "epoch": 0.8221597751906865,
      "grad_norm": 0.27787894010543823,
      "learning_rate": 2.107843137254902e-05,
      "loss": 0.2343,
      "step": 192
    },
    {
      "epoch": 0.8264418573531379,
      "grad_norm": 0.2684335708618164,
      "learning_rate": 2.058823529411765e-05,
      "loss": 0.2077,
      "step": 193
    },
    {
      "epoch": 0.8307239395155894,
      "grad_norm": 0.2913365364074707,
      "learning_rate": 2.0098039215686277e-05,
      "loss": 0.2239,
      "step": 194
    },
    {
      "epoch": 0.8350060216780409,
      "grad_norm": 0.3032510578632355,
      "learning_rate": 1.9607843137254903e-05,
      "loss": 0.2636,
      "step": 195
    },
    {
      "epoch": 0.8392881038404925,
      "grad_norm": 0.3184592127799988,
      "learning_rate": 1.9117647058823528e-05,
      "loss": 0.2412,
      "step": 196
    },
    {
      "epoch": 0.843570186002944,
      "grad_norm": 0.31382012367248535,
      "learning_rate": 1.862745098039216e-05,
      "loss": 0.2739,
      "step": 197
    },
    {
      "epoch": 0.8478522681653954,
      "grad_norm": 0.2534014880657196,
      "learning_rate": 1.8137254901960785e-05,
      "loss": 0.2059,
      "step": 198
    },
    {
      "epoch": 0.8521343503278469,
      "grad_norm": 0.31502148509025574,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.2359,
      "step": 199
    },
    {
      "epoch": 0.8564164324902984,
      "grad_norm": 0.31098228693008423,
      "learning_rate": 1.715686274509804e-05,
      "loss": 0.2053,
      "step": 200
    },
    {
      "epoch": 0.8606985146527499,
      "grad_norm": 0.28679826855659485,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.2443,
      "step": 201
    },
    {
      "epoch": 0.8649805968152013,
      "grad_norm": 0.2990693151950836,
      "learning_rate": 1.6176470588235296e-05,
      "loss": 0.2306,
      "step": 202
    },
    {
      "epoch": 0.8692626789776529,
      "grad_norm": 0.2778109014034271,
      "learning_rate": 1.568627450980392e-05,
      "loss": 0.2373,
      "step": 203
    },
    {
      "epoch": 0.8735447611401044,
      "grad_norm": 0.27156633138656616,
      "learning_rate": 1.5196078431372548e-05,
      "loss": 0.2322,
      "step": 204
    },
    {
      "epoch": 0.8778268433025559,
      "grad_norm": 0.2727411985397339,
      "learning_rate": 1.4705882352941177e-05,
      "loss": 0.1994,
      "step": 205
    },
    {
      "epoch": 0.8821089254650074,
      "grad_norm": 0.30019769072532654,
      "learning_rate": 1.4215686274509804e-05,
      "loss": 0.227,
      "step": 206
    },
    {
      "epoch": 0.8863910076274588,
      "grad_norm": 0.2669230103492737,
      "learning_rate": 1.3725490196078432e-05,
      "loss": 0.1886,
      "step": 207
    },
    {
      "epoch": 0.8906730897899103,
      "grad_norm": 0.30478113889694214,
      "learning_rate": 1.323529411764706e-05,
      "loss": 0.2578,
      "step": 208
    },
    {
      "epoch": 0.8949551719523618,
      "grad_norm": 0.2967156767845154,
      "learning_rate": 1.2745098039215686e-05,
      "loss": 0.2523,
      "step": 209
    },
    {
      "epoch": 0.8992372541148134,
      "grad_norm": 0.3051101863384247,
      "learning_rate": 1.2254901960784313e-05,
      "loss": 0.2241,
      "step": 210
    },
    {
      "epoch": 0.9035193362772648,
      "grad_norm": 0.3345109820365906,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.2424,
      "step": 211
    },
    {
      "epoch": 0.9078014184397163,
      "grad_norm": 0.2779349088668823,
      "learning_rate": 1.1274509803921569e-05,
      "loss": 0.1901,
      "step": 212
    },
    {
      "epoch": 0.9120835006021678,
      "grad_norm": 0.33007028698921204,
      "learning_rate": 1.0784313725490197e-05,
      "loss": 0.2382,
      "step": 213
    },
    {
      "epoch": 0.9163655827646193,
      "grad_norm": 0.30623412132263184,
      "learning_rate": 1.0294117647058824e-05,
      "loss": 0.2498,
      "step": 214
    },
    {
      "epoch": 0.9206476649270708,
      "grad_norm": 0.32106027007102966,
      "learning_rate": 9.803921568627451e-06,
      "loss": 0.2347,
      "step": 215
    },
    {
      "epoch": 0.9249297470895222,
      "grad_norm": 0.2883519232273102,
      "learning_rate": 9.31372549019608e-06,
      "loss": 0.2447,
      "step": 216
    },
    {
      "epoch": 0.9292118292519738,
      "grad_norm": 0.2879474461078644,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.2311,
      "step": 217
    },
    {
      "epoch": 0.9334939114144253,
      "grad_norm": 0.30156365036964417,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.2254,
      "step": 218
    },
    {
      "epoch": 0.9377759935768768,
      "grad_norm": 0.3522617816925049,
      "learning_rate": 7.84313725490196e-06,
      "loss": 0.2696,
      "step": 219
    },
    {
      "epoch": 0.9420580757393282,
      "grad_norm": 0.26573362946510315,
      "learning_rate": 7.3529411764705884e-06,
      "loss": 0.1944,
      "step": 220
    },
    {
      "epoch": 0.9463401579017797,
      "grad_norm": 0.34632617235183716,
      "learning_rate": 6.862745098039216e-06,
      "loss": 0.2886,
      "step": 221
    },
    {
      "epoch": 0.9506222400642312,
      "grad_norm": 0.29579487442970276,
      "learning_rate": 6.372549019607843e-06,
      "loss": 0.2386,
      "step": 222
    },
    {
      "epoch": 0.9549043222266828,
      "grad_norm": 0.3460155427455902,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.2346,
      "step": 223
    },
    {
      "epoch": 0.9591864043891342,
      "grad_norm": 0.2727835476398468,
      "learning_rate": 5.392156862745099e-06,
      "loss": 0.2215,
      "step": 224
    },
    {
      "epoch": 0.9634684865515857,
      "grad_norm": 0.27711188793182373,
      "learning_rate": 4.901960784313726e-06,
      "loss": 0.2255,
      "step": 225
    },
    {
      "epoch": 0.9677505687140372,
      "grad_norm": 0.32168129086494446,
      "learning_rate": 4.411764705882353e-06,
      "loss": 0.2691,
      "step": 226
    },
    {
      "epoch": 0.9720326508764887,
      "grad_norm": 0.2886391580104828,
      "learning_rate": 3.92156862745098e-06,
      "loss": 0.2074,
      "step": 227
    },
    {
      "epoch": 0.9763147330389402,
      "grad_norm": 0.317636638879776,
      "learning_rate": 3.431372549019608e-06,
      "loss": 0.2577,
      "step": 228
    },
    {
      "epoch": 0.9805968152013916,
      "grad_norm": 0.3026321828365326,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 0.242,
      "step": 229
    },
    {
      "epoch": 0.9848788973638432,
      "grad_norm": 0.3242248594760895,
      "learning_rate": 2.450980392156863e-06,
      "loss": 0.2338,
      "step": 230
    },
    {
      "epoch": 0.9891609795262947,
      "grad_norm": 0.31835997104644775,
      "learning_rate": 1.96078431372549e-06,
      "loss": 0.2157,
      "step": 231
    },
    {
      "epoch": 0.9934430616887462,
      "grad_norm": 0.2741211950778961,
      "learning_rate": 1.4705882352941177e-06,
      "loss": 0.2209,
      "step": 232
    },
    {
      "epoch": 0.9977251438511976,
      "grad_norm": 0.2993812561035156,
      "learning_rate": 9.80392156862745e-07,
      "loss": 0.1945,
      "step": 233
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.49857211112976074,
      "learning_rate": 4.901960784313725e-07,
      "loss": 0.2137,
      "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": 3.805299957595853e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}