{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.0,
  "eval_steps": 200,
  "global_step": 210,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02857142857142857,
      "grad_norm": 19.810293087197298,
      "learning_rate": 5e-08,
      "loss": 1.6319,
      "step": 1
    },
    {
      "epoch": 0.05714285714285714,
      "grad_norm": 20.10065322014341,
      "learning_rate": 1e-07,
      "loss": 1.5294,
      "step": 2
    },
    {
      "epoch": 0.08571428571428572,
      "grad_norm": 22.139565739974667,
      "learning_rate": 1.5e-07,
      "loss": 1.568,
      "step": 3
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 17.607582690653754,
      "learning_rate": 2e-07,
      "loss": 1.6628,
      "step": 4
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 24.910794234048186,
      "learning_rate": 2.5e-07,
      "loss": 1.5738,
      "step": 5
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 18.006572442480767,
      "learning_rate": 3e-07,
      "loss": 1.5898,
      "step": 6
    },
    {
      "epoch": 0.2,
      "grad_norm": 22.152647510768777,
      "learning_rate": 3.5e-07,
      "loss": 1.5891,
      "step": 7
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 17.044462125735237,
      "learning_rate": 4e-07,
      "loss": 1.5277,
      "step": 8
    },
    {
      "epoch": 0.2571428571428571,
      "grad_norm": 19.26164281649356,
      "learning_rate": 4.5e-07,
      "loss": 1.5504,
      "step": 9
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 20.409183726640773,
      "learning_rate": 5e-07,
      "loss": 1.5509,
      "step": 10
    },
    {
      "epoch": 0.3142857142857143,
      "grad_norm": 16.865224915691606,
      "learning_rate": 5.5e-07,
      "loss": 1.5752,
      "step": 11
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 15.722428651072727,
      "learning_rate": 6e-07,
      "loss": 1.6208,
      "step": 12
    },
    {
      "epoch": 0.37142857142857144,
      "grad_norm": 13.23419740428603,
      "learning_rate": 6.5e-07,
      "loss": 1.515,
      "step": 13
    },
    {
      "epoch": 0.4,
      "grad_norm": 12.45173447148941,
      "learning_rate": 7e-07,
      "loss": 1.4752,
      "step": 14
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 10.684936055938318,
      "learning_rate": 7.5e-07,
      "loss": 1.5073,
      "step": 15
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 8.612999241840305,
      "learning_rate": 8e-07,
      "loss": 1.4206,
      "step": 16
    },
    {
      "epoch": 0.4857142857142857,
      "grad_norm": 9.64409059662565,
      "learning_rate": 8.499999999999999e-07,
      "loss": 1.3659,
      "step": 17
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 9.53721910636901,
      "learning_rate": 9e-07,
      "loss": 1.4999,
      "step": 18
    },
    {
      "epoch": 0.5428571428571428,
      "grad_norm": 15.44618942298652,
      "learning_rate": 9.499999999999999e-07,
      "loss": 1.452,
      "step": 19
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 7.797878397007685,
      "learning_rate": 1e-06,
      "loss": 1.4059,
      "step": 20
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.611751272514192,
      "learning_rate": 9.999316524962345e-07,
      "loss": 1.3416,
      "step": 21
    },
    {
      "epoch": 0.6285714285714286,
      "grad_norm": 8.60172100284345,
      "learning_rate": 9.99726628670463e-07,
      "loss": 1.3334,
      "step": 22
    },
    {
      "epoch": 0.6571428571428571,
      "grad_norm": 7.707986778608113,
      "learning_rate": 9.993849845741523e-07,
      "loss": 1.3032,
      "step": 23
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 7.692312413668571,
      "learning_rate": 9.989068136093872e-07,
      "loss": 1.3742,
      "step": 24
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 6.364211690724886,
      "learning_rate": 9.982922465033348e-07,
      "loss": 1.2641,
      "step": 25
    },
    {
      "epoch": 0.7428571428571429,
      "grad_norm": 6.85599706491363,
      "learning_rate": 9.975414512725056e-07,
      "loss": 1.2624,
      "step": 26
    },
    {
      "epoch": 0.7714285714285715,
      "grad_norm": 6.401247246252304,
      "learning_rate": 9.966546331768192e-07,
      "loss": 1.258,
      "step": 27
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.111986647021705,
      "learning_rate": 9.956320346634875e-07,
      "loss": 1.3228,
      "step": 28
    },
    {
      "epoch": 0.8285714285714286,
      "grad_norm": 6.326533367398195,
      "learning_rate": 9.944739353007341e-07,
      "loss": 1.2034,
      "step": 29
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 6.561308869602955,
      "learning_rate": 9.931806517013612e-07,
      "loss": 1.1795,
      "step": 30
    },
    {
      "epoch": 0.8857142857142857,
      "grad_norm": 5.9273022974036795,
      "learning_rate": 9.917525374361911e-07,
      "loss": 1.2288,
      "step": 31
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 5.348130355740025,
      "learning_rate": 9.901899829374047e-07,
      "loss": 1.1764,
      "step": 32
    },
    {
      "epoch": 0.9428571428571428,
      "grad_norm": 5.274551340001603,
      "learning_rate": 9.884934153917996e-07,
      "loss": 1.1402,
      "step": 33
    },
    {
      "epoch": 0.9714285714285714,
      "grad_norm": 5.681514700875701,
      "learning_rate": 9.866632986240029e-07,
      "loss": 1.1604,
      "step": 34
    },
    {
      "epoch": 1.0,
      "grad_norm": 5.77664764652199,
      "learning_rate": 9.847001329696652e-07,
      "loss": 1.1567,
      "step": 35
    },
    {
      "epoch": 1.0285714285714285,
      "grad_norm": 5.836094968508217,
      "learning_rate": 9.826044551386742e-07,
      "loss": 1.084,
      "step": 36
    },
    {
      "epoch": 1.0571428571428572,
      "grad_norm": 5.827344070494953,
      "learning_rate": 9.803768380684242e-07,
      "loss": 1.1305,
      "step": 37
    },
    {
      "epoch": 1.0857142857142856,
      "grad_norm": 5.1906563321747115,
      "learning_rate": 9.780178907671788e-07,
      "loss": 1.0574,
      "step": 38
    },
    {
      "epoch": 1.1142857142857143,
      "grad_norm": 5.562551195695788,
      "learning_rate": 9.755282581475767e-07,
      "loss": 1.1391,
      "step": 39
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 5.717325403158778,
      "learning_rate": 9.729086208503173e-07,
      "loss": 1.1267,
      "step": 40
    },
    {
      "epoch": 1.1714285714285715,
      "grad_norm": 5.969025543565558,
      "learning_rate": 9.701596950580807e-07,
      "loss": 1.1055,
      "step": 41
    },
    {
      "epoch": 1.2,
      "grad_norm": 4.274011328484305,
      "learning_rate": 9.672822322997304e-07,
      "loss": 1.0392,
      "step": 42
    },
    {
      "epoch": 1.2285714285714286,
      "grad_norm": 4.4561906748259545,
      "learning_rate": 9.642770192448535e-07,
      "loss": 1.0393,
      "step": 43
    },
    {
      "epoch": 1.2571428571428571,
      "grad_norm": 4.618703002376858,
      "learning_rate": 9.611448774886923e-07,
      "loss": 1.0795,
      "step": 44
    },
    {
      "epoch": 1.2857142857142856,
      "grad_norm": 4.473713036590507,
      "learning_rate": 9.578866633275286e-07,
      "loss": 1.0194,
      "step": 45
    },
    {
      "epoch": 1.3142857142857143,
      "grad_norm": 4.3150195602071415,
      "learning_rate": 9.545032675245813e-07,
      "loss": 1.0409,
      "step": 46
    },
    {
      "epoch": 1.342857142857143,
      "grad_norm": 4.847108292314838,
      "learning_rate": 9.509956150664795e-07,
      "loss": 1.0475,
      "step": 47
    },
    {
      "epoch": 1.3714285714285714,
      "grad_norm": 4.716522743464805,
      "learning_rate": 9.473646649103817e-07,
      "loss": 1.046,
      "step": 48
    },
    {
      "epoch": 1.4,
      "grad_norm": 4.942648131785751,
      "learning_rate": 9.436114097218058e-07,
      "loss": 0.9876,
      "step": 49
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 5.1753554721234565,
      "learning_rate": 9.397368756032444e-07,
      "loss": 1.0394,
      "step": 50
    },
    {
      "epoch": 1.457142857142857,
      "grad_norm": 4.996816148990504,
      "learning_rate": 9.357421218136386e-07,
      "loss": 0.9647,
      "step": 51
    },
    {
      "epoch": 1.4857142857142858,
      "grad_norm": 5.982295572473842,
      "learning_rate": 9.316282404787869e-07,
      "loss": 1.0986,
      "step": 52
    },
    {
      "epoch": 1.5142857142857142,
      "grad_norm": 4.701714221899932,
      "learning_rate": 9.273963562927694e-07,
      "loss": 1.0493,
      "step": 53
    },
    {
      "epoch": 1.5428571428571427,
      "grad_norm": 4.514923820929648,
      "learning_rate": 9.230476262104676e-07,
      "loss": 0.9847,
      "step": 54
    },
    {
      "epoch": 1.5714285714285714,
      "grad_norm": 4.701919910281847,
      "learning_rate": 9.185832391312642e-07,
      "loss": 0.9794,
      "step": 55
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.175054800986126,
      "learning_rate": 9.1400441557401e-07,
      "loss": 1.0607,
      "step": 56
    },
    {
      "epoch": 1.6285714285714286,
      "grad_norm": 4.080487196055978,
      "learning_rate": 9.093124073433462e-07,
      "loss": 0.9856,
      "step": 57
    },
    {
      "epoch": 1.657142857142857,
      "grad_norm": 4.168373937047882,
      "learning_rate": 9.045084971874737e-07,
      "loss": 0.9107,
      "step": 58
    },
    {
      "epoch": 1.6857142857142857,
      "grad_norm": 4.223886586576581,
      "learning_rate": 8.995939984474623e-07,
      "loss": 0.9733,
      "step": 59
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 4.513527637157093,
      "learning_rate": 8.945702546981968e-07,
      "loss": 0.9476,
      "step": 60
    },
    {
      "epoch": 1.7428571428571429,
      "grad_norm": 6.610164639785705,
      "learning_rate": 8.894386393810562e-07,
      "loss": 0.9161,
      "step": 61
    },
    {
      "epoch": 1.7714285714285714,
      "grad_norm": 4.703651247017542,
      "learning_rate": 8.842005554284295e-07,
      "loss": 1.0058,
      "step": 62
    },
    {
      "epoch": 1.8,
      "grad_norm": 4.843849342547474,
      "learning_rate": 8.788574348801674e-07,
      "loss": 0.8815,
      "step": 63
    },
    {
      "epoch": 1.8285714285714287,
      "grad_norm": 4.850275003883815,
      "learning_rate": 8.734107384920769e-07,
      "loss": 0.9945,
      "step": 64
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 5.453604273607254,
      "learning_rate": 8.678619553365658e-07,
      "loss": 0.9822,
      "step": 65
    },
    {
      "epoch": 1.8857142857142857,
      "grad_norm": 5.3760203289286155,
      "learning_rate": 8.622126023955445e-07,
      "loss": 0.9781,
      "step": 66
    },
    {
      "epoch": 1.9142857142857141,
      "grad_norm": 3.7890108184255062,
      "learning_rate": 8.564642241456986e-07,
      "loss": 0.9188,
      "step": 67
    },
    {
      "epoch": 1.9428571428571428,
      "grad_norm": 4.482132592462384,
      "learning_rate": 8.506183921362442e-07,
      "loss": 0.9193,
      "step": 68
    },
    {
      "epoch": 1.9714285714285715,
      "grad_norm": 5.311797532408616,
      "learning_rate": 8.446767045592829e-07,
      "loss": 0.9563,
      "step": 69
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.950363360704343,
      "learning_rate": 8.386407858128706e-07,
      "loss": 0.9172,
      "step": 70
    },
    {
      "epoch": 2.0285714285714285,
      "grad_norm": 4.784370125705732,
      "learning_rate": 8.325122860569241e-07,
      "loss": 0.8843,
      "step": 71
    },
    {
      "epoch": 2.057142857142857,
      "grad_norm": 3.63025112886302,
      "learning_rate": 8.262928807620843e-07,
      "loss": 0.8547,
      "step": 72
    },
    {
      "epoch": 2.085714285714286,
      "grad_norm": 5.218576028144485,
      "learning_rate": 8.199842702516582e-07,
      "loss": 0.9414,
      "step": 73
    },
    {
      "epoch": 2.1142857142857143,
      "grad_norm": 4.677814392084542,
      "learning_rate": 8.135881792367685e-07,
      "loss": 0.8748,
      "step": 74
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 4.400225941517684,
      "learning_rate": 8.071063563448339e-07,
      "loss": 0.8646,
      "step": 75
    },
    {
      "epoch": 2.1714285714285713,
      "grad_norm": 4.629375190268374,
      "learning_rate": 8.005405736415125e-07,
      "loss": 0.8744,
      "step": 76
    },
    {
      "epoch": 2.2,
      "grad_norm": 4.1365979610935,
      "learning_rate": 7.938926261462365e-07,
      "loss": 0.8735,
      "step": 77
    },
    {
      "epoch": 2.2285714285714286,
      "grad_norm": 4.796843279900597,
      "learning_rate": 7.871643313414718e-07,
      "loss": 0.9161,
      "step": 78
    },
    {
      "epoch": 2.257142857142857,
      "grad_norm": 4.297808554492733,
      "learning_rate": 7.803575286758363e-07,
      "loss": 0.8119,
      "step": 79
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 4.229241471681201,
      "learning_rate": 7.734740790612136e-07,
      "loss": 0.8698,
      "step": 80
    },
    {
      "epoch": 2.314285714285714,
      "grad_norm": 5.474712766868538,
      "learning_rate": 7.665158643639969e-07,
      "loss": 0.8673,
      "step": 81
    },
    {
      "epoch": 2.342857142857143,
      "grad_norm": 5.8097182446336495,
      "learning_rate": 7.594847868906076e-07,
      "loss": 0.8561,
      "step": 82
    },
    {
      "epoch": 2.3714285714285714,
      "grad_norm": 6.090812187892629,
      "learning_rate": 7.523827688674219e-07,
      "loss": 0.9032,
      "step": 83
    },
    {
      "epoch": 2.4,
      "grad_norm": 4.2456911319822614,
      "learning_rate": 7.452117519152541e-07,
      "loss": 0.8431,
      "step": 84
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 4.562925662849317,
      "learning_rate": 7.379736965185368e-07,
      "loss": 0.844,
      "step": 85
    },
    {
      "epoch": 2.4571428571428573,
      "grad_norm": 4.517660029971984,
      "learning_rate": 7.306705814893439e-07,
      "loss": 0.8071,
      "step": 86
    },
    {
      "epoch": 2.4857142857142858,
      "grad_norm": 4.970820782182107,
      "learning_rate": 7.233044034264033e-07,
      "loss": 0.8726,
      "step": 87
    },
    {
      "epoch": 2.5142857142857142,
      "grad_norm": 4.817397128446438,
      "learning_rate": 7.158771761692464e-07,
      "loss": 0.8651,
      "step": 88
    },
    {
      "epoch": 2.5428571428571427,
      "grad_norm": 4.521311016914525,
      "learning_rate": 7.083909302476452e-07,
      "loss": 0.8023,
      "step": 89
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 5.57347106069374,
      "learning_rate": 7.008477123264847e-07,
      "loss": 0.8107,
      "step": 90
    },
    {
      "epoch": 2.6,
      "grad_norm": 4.958424900327067,
      "learning_rate": 6.932495846462261e-07,
      "loss": 0.8321,
      "step": 91
    },
    {
      "epoch": 2.6285714285714286,
      "grad_norm": 4.34206205323412,
      "learning_rate": 6.855986244591103e-07,
      "loss": 0.8003,
      "step": 92
    },
    {
      "epoch": 2.657142857142857,
      "grad_norm": 4.417137471491474,
      "learning_rate": 6.778969234612583e-07,
      "loss": 0.81,
      "step": 93
    },
    {
      "epoch": 2.685714285714286,
      "grad_norm": 4.714391490425693,
      "learning_rate": 6.701465872208216e-07,
      "loss": 0.8362,
      "step": 94
    },
    {
      "epoch": 2.7142857142857144,
      "grad_norm": 3.835031010544762,
      "learning_rate": 6.623497346023417e-07,
      "loss": 0.7408,
      "step": 95
    },
    {
      "epoch": 2.742857142857143,
      "grad_norm": 4.43328902038276,
      "learning_rate": 6.545084971874736e-07,
      "loss": 0.803,
      "step": 96
    },
    {
      "epoch": 2.7714285714285714,
      "grad_norm": 5.573988325671216,
      "learning_rate": 6.466250186922324e-07,
      "loss": 0.8499,
      "step": 97
    },
    {
      "epoch": 2.8,
      "grad_norm": 5.080352906047553,
      "learning_rate": 6.387014543809223e-07,
      "loss": 0.8536,
      "step": 98
    },
    {
      "epoch": 2.8285714285714287,
      "grad_norm": 4.433727646568338,
      "learning_rate": 6.307399704769098e-07,
      "loss": 0.8167,
      "step": 99
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 4.72983530688845,
      "learning_rate": 6.227427435703995e-07,
      "loss": 0.8247,
      "step": 100
    },
    {
      "epoch": 2.8857142857142857,
      "grad_norm": 4.031675844503338,
      "learning_rate": 6.147119600233758e-07,
      "loss": 0.7923,
      "step": 101
    },
    {
      "epoch": 2.914285714285714,
      "grad_norm": 4.657562682714027,
      "learning_rate": 6.066498153718734e-07,
      "loss": 0.8259,
      "step": 102
    },
    {
      "epoch": 2.942857142857143,
      "grad_norm": 3.742914618609364,
      "learning_rate": 5.985585137257401e-07,
      "loss": 0.7416,
      "step": 103
    },
    {
      "epoch": 2.9714285714285715,
      "grad_norm": 5.392956655820792,
      "learning_rate": 5.90440267166055e-07,
      "loss": 0.86,
      "step": 104
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.370818695376769,
      "learning_rate": 5.82297295140367e-07,
      "loss": 0.7756,
      "step": 105
    },
    {
      "epoch": 3.0285714285714285,
      "grad_norm": 4.1080059715561426,
      "learning_rate": 5.741318238559209e-07,
      "loss": 0.7417,
      "step": 106
    },
    {
      "epoch": 3.057142857142857,
      "grad_norm": 5.242835158751313,
      "learning_rate": 5.659460856710345e-07,
      "loss": 0.7413,
      "step": 107
    },
    {
      "epoch": 3.085714285714286,
      "grad_norm": 4.213242126183361,
      "learning_rate": 5.577423184847931e-07,
      "loss": 0.774,
      "step": 108
    },
    {
      "epoch": 3.1142857142857143,
      "grad_norm": 5.14212832829309,
      "learning_rate": 5.495227651252315e-07,
      "loss": 0.7596,
      "step": 109
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 4.536657005455878,
      "learning_rate": 5.412896727361662e-07,
      "loss": 0.7494,
      "step": 110
    },
    {
      "epoch": 3.1714285714285713,
      "grad_norm": 5.180517603140962,
      "learning_rate": 5.330452921628497e-07,
      "loss": 0.8146,
      "step": 111
    },
    {
      "epoch": 3.2,
      "grad_norm": 4.499166275543811,
      "learning_rate": 5.247918773366111e-07,
      "loss": 0.7437,
      "step": 112
    },
    {
      "epoch": 3.2285714285714286,
      "grad_norm": 4.78892229810655,
      "learning_rate": 5.165316846586541e-07,
      "loss": 0.732,
      "step": 113
    },
    {
      "epoch": 3.257142857142857,
      "grad_norm": 4.643947668361775,
      "learning_rate": 5.082669723831793e-07,
      "loss": 0.769,
      "step": 114
    },
    {
      "epoch": 3.2857142857142856,
      "grad_norm": 4.79036988016491,
      "learning_rate": 5e-07,
      "loss": 0.7586,
      "step": 115
    },
    {
      "epoch": 3.314285714285714,
      "grad_norm": 5.1592671986124845,
      "learning_rate": 4.917330276168208e-07,
      "loss": 0.7513,
      "step": 116
    },
    {
      "epoch": 3.342857142857143,
      "grad_norm": 3.944922328543468,
      "learning_rate": 4.834683153413459e-07,
      "loss": 0.7184,
      "step": 117
    },
    {
      "epoch": 3.3714285714285714,
      "grad_norm": 4.429646542430914,
      "learning_rate": 4.752081226633888e-07,
      "loss": 0.6937,
      "step": 118
    },
    {
      "epoch": 3.4,
      "grad_norm": 5.016602862879482,
      "learning_rate": 4.669547078371503e-07,
      "loss": 0.7638,
      "step": 119
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 4.443230989038584,
      "learning_rate": 4.5871032726383385e-07,
      "loss": 0.7029,
      "step": 120
    },
    {
      "epoch": 3.4571428571428573,
      "grad_norm": 4.244778886390162,
      "learning_rate": 4.5047723487476864e-07,
      "loss": 0.7298,
      "step": 121
    },
    {
      "epoch": 3.4857142857142858,
      "grad_norm": 5.068619137512623,
      "learning_rate": 4.4225768151520694e-07,
      "loss": 0.7716,
      "step": 122
    },
    {
      "epoch": 3.5142857142857142,
      "grad_norm": 4.652036232274711,
      "learning_rate": 4.340539143289655e-07,
      "loss": 0.7276,
      "step": 123
    },
    {
      "epoch": 3.5428571428571427,
      "grad_norm": 4.95389292196556,
      "learning_rate": 4.258681761440789e-07,
      "loss": 0.7345,
      "step": 124
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 4.275659541196405,
      "learning_rate": 4.1770270485963294e-07,
      "loss": 0.7323,
      "step": 125
    },
    {
      "epoch": 3.6,
      "grad_norm": 4.607417501587952,
      "learning_rate": 4.095597328339452e-07,
      "loss": 0.7606,
      "step": 126
    },
    {
      "epoch": 3.6285714285714286,
      "grad_norm": 4.794919375541705,
      "learning_rate": 4.0144148627425986e-07,
      "loss": 0.7149,
      "step": 127
    },
    {
      "epoch": 3.657142857142857,
      "grad_norm": 4.102606927748694,
      "learning_rate": 3.9335018462812664e-07,
      "loss": 0.6956,
      "step": 128
    },
    {
      "epoch": 3.685714285714286,
      "grad_norm": 4.243593175311764,
      "learning_rate": 3.8528803997662423e-07,
      "loss": 0.6894,
      "step": 129
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 4.76702832294717,
      "learning_rate": 3.772572564296004e-07,
      "loss": 0.7188,
      "step": 130
    },
    {
      "epoch": 3.742857142857143,
      "grad_norm": 4.651784836564843,
      "learning_rate": 3.692600295230901e-07,
      "loss": 0.7059,
      "step": 131
    },
    {
      "epoch": 3.7714285714285714,
      "grad_norm": 4.977499267065962,
      "learning_rate": 3.612985456190778e-07,
      "loss": 0.7022,
      "step": 132
    },
    {
      "epoch": 3.8,
      "grad_norm": 4.629091116356692,
      "learning_rate": 3.533749813077677e-07,
      "loss": 0.692,
      "step": 133
    },
    {
      "epoch": 3.8285714285714287,
      "grad_norm": 5.180269168283943,
      "learning_rate": 3.454915028125263e-07,
      "loss": 0.7245,
      "step": 134
    },
    {
      "epoch": 3.857142857142857,
      "grad_norm": 4.804587126213361,
      "learning_rate": 3.3765026539765827e-07,
      "loss": 0.6965,
      "step": 135
    },
    {
      "epoch": 3.8857142857142857,
      "grad_norm": 5.24711927553455,
      "learning_rate": 3.2985341277917846e-07,
      "loss": 0.7271,
      "step": 136
    },
    {
      "epoch": 3.914285714285714,
      "grad_norm": 4.306919498942911,
      "learning_rate": 3.221030765387417e-07,
      "loss": 0.7287,
      "step": 137
    },
    {
      "epoch": 3.942857142857143,
      "grad_norm": 5.1155454022030815,
      "learning_rate": 3.1440137554088953e-07,
      "loss": 0.7441,
      "step": 138
    },
    {
      "epoch": 3.9714285714285715,
      "grad_norm": 4.184043691688485,
      "learning_rate": 3.06750415353774e-07,
      "loss": 0.7349,
      "step": 139
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.262350945325998,
      "learning_rate": 2.9915228767351535e-07,
      "loss": 0.7377,
      "step": 140
    },
    {
      "epoch": 4.0285714285714285,
      "grad_norm": 4.103707140569048,
      "learning_rate": 2.916090697523549e-07,
      "loss": 0.6864,
      "step": 141
    },
    {
      "epoch": 4.057142857142857,
      "grad_norm": 4.9358404348712765,
      "learning_rate": 2.841228238307536e-07,
      "loss": 0.7073,
      "step": 142
    },
    {
      "epoch": 4.085714285714285,
      "grad_norm": 4.224580170912438,
      "learning_rate": 2.7669559657359673e-07,
      "loss": 0.6669,
      "step": 143
    },
    {
      "epoch": 4.114285714285714,
      "grad_norm": 4.571605759259632,
      "learning_rate": 2.6932941851065615e-07,
      "loss": 0.6673,
      "step": 144
    },
    {
      "epoch": 4.142857142857143,
      "grad_norm": 4.161244011003443,
      "learning_rate": 2.620263034814632e-07,
      "loss": 0.6644,
      "step": 145
    },
    {
      "epoch": 4.171428571428572,
      "grad_norm": 4.128269627797568,
      "learning_rate": 2.547882480847461e-07,
      "loss": 0.6568,
      "step": 146
    },
    {
      "epoch": 4.2,
      "grad_norm": 4.562112299679241,
      "learning_rate": 2.476172311325783e-07,
      "loss": 0.7048,
      "step": 147
    },
    {
      "epoch": 4.228571428571429,
      "grad_norm": 5.336142167342819,
      "learning_rate": 2.4051521310939254e-07,
      "loss": 0.7082,
      "step": 148
    },
    {
      "epoch": 4.257142857142857,
      "grad_norm": 4.796197499964798,
      "learning_rate": 2.3348413563600323e-07,
      "loss": 0.6497,
      "step": 149
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 4.825647232190606,
      "learning_rate": 2.2652592093878665e-07,
      "loss": 0.6851,
      "step": 150
    },
    {
      "epoch": 4.314285714285714,
      "grad_norm": 4.462461397938811,
      "learning_rate": 2.1964247132416368e-07,
      "loss": 0.6974,
      "step": 151
    },
    {
      "epoch": 4.3428571428571425,
      "grad_norm": 4.829646877353178,
      "learning_rate": 2.128356686585282e-07,
      "loss": 0.6439,
      "step": 152
    },
    {
      "epoch": 4.371428571428572,
      "grad_norm": 5.043886293339906,
      "learning_rate": 2.0610737385376348e-07,
      "loss": 0.7215,
      "step": 153
    },
    {
      "epoch": 4.4,
      "grad_norm": 4.598743427591961,
      "learning_rate": 1.9945942635848745e-07,
      "loss": 0.644,
      "step": 154
    },
    {
      "epoch": 4.428571428571429,
      "grad_norm": 3.8253679810994012,
      "learning_rate": 1.9289364365516607e-07,
      "loss": 0.6397,
      "step": 155
    },
    {
      "epoch": 4.457142857142857,
      "grad_norm": 4.807293265176188,
      "learning_rate": 1.8641182076323148e-07,
      "loss": 0.689,
      "step": 156
    },
    {
      "epoch": 4.485714285714286,
      "grad_norm": 3.8062822348294363,
      "learning_rate": 1.8001572974834168e-07,
      "loss": 0.6466,
      "step": 157
    },
    {
      "epoch": 4.514285714285714,
      "grad_norm": 4.883594185819761,
      "learning_rate": 1.7370711923791564e-07,
      "loss": 0.6691,
      "step": 158
    },
    {
      "epoch": 4.542857142857143,
      "grad_norm": 5.143361513426633,
      "learning_rate": 1.674877139430758e-07,
      "loss": 0.6432,
      "step": 159
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 4.99547680018691,
      "learning_rate": 1.6135921418712955e-07,
      "loss": 0.6685,
      "step": 160
    },
    {
      "epoch": 4.6,
      "grad_norm": 4.6838462077186795,
      "learning_rate": 1.553232954407171e-07,
      "loss": 0.657,
      "step": 161
    },
    {
      "epoch": 4.628571428571428,
      "grad_norm": 4.841965011015589,
      "learning_rate": 1.493816078637557e-07,
      "loss": 0.6671,
      "step": 162
    },
    {
      "epoch": 4.6571428571428575,
      "grad_norm": 5.2138374952108535,
      "learning_rate": 1.435357758543015e-07,
      "loss": 0.7393,
      "step": 163
    },
    {
      "epoch": 4.685714285714286,
      "grad_norm": 4.7664344007331225,
      "learning_rate": 1.3778739760445552e-07,
      "loss": 0.6904,
      "step": 164
    },
    {
      "epoch": 4.714285714285714,
      "grad_norm": 4.757741131293643,
      "learning_rate": 1.321380446634342e-07,
      "loss": 0.7093,
      "step": 165
    },
    {
      "epoch": 4.742857142857143,
      "grad_norm": 5.064771039141748,
      "learning_rate": 1.2658926150792322e-07,
      "loss": 0.6588,
      "step": 166
    },
    {
      "epoch": 4.771428571428571,
      "grad_norm": 4.745894552114562,
      "learning_rate": 1.2114256511983274e-07,
      "loss": 0.6914,
      "step": 167
    },
    {
      "epoch": 4.8,
      "grad_norm": 4.5842953923274425,
      "learning_rate": 1.1579944457157059e-07,
      "loss": 0.6819,
      "step": 168
    },
    {
      "epoch": 4.828571428571428,
      "grad_norm": 6.02007293004415,
      "learning_rate": 1.1056136061894384e-07,
      "loss": 0.7107,
      "step": 169
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 4.04792450108683,
      "learning_rate": 1.0542974530180327e-07,
      "loss": 0.6577,
      "step": 170
    },
    {
      "epoch": 4.885714285714286,
      "grad_norm": 4.334397719277511,
      "learning_rate": 1.0040600155253764e-07,
      "loss": 0.6926,
      "step": 171
    },
    {
      "epoch": 4.914285714285715,
      "grad_norm": 4.763783176385905,
      "learning_rate": 9.549150281252632e-08,
      "loss": 0.6824,
      "step": 172
    },
    {
      "epoch": 4.942857142857143,
      "grad_norm": 3.8987128530052075,
      "learning_rate": 9.068759265665382e-08,
      "loss": 0.6293,
      "step": 173
    },
    {
      "epoch": 4.9714285714285715,
      "grad_norm": 5.260545837340523,
      "learning_rate": 8.599558442598998e-08,
      "loss": 0.7605,
      "step": 174
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.28845521901137,
      "learning_rate": 8.141676086873573e-08,
      "loss": 0.6685,
      "step": 175
    },
    {
      "epoch": 5.0285714285714285,
      "grad_norm": 4.317069591639633,
      "learning_rate": 7.695237378953224e-08,
      "loss": 0.6178,
      "step": 176
    },
    {
      "epoch": 5.057142857142857,
      "grad_norm": 4.673363595181425,
      "learning_rate": 7.260364370723043e-08,
      "loss": 0.6338,
      "step": 177
    },
    {
      "epoch": 5.085714285714285,
      "grad_norm": 4.179297774765176,
      "learning_rate": 6.837175952121304e-08,
      "loss": 0.656,
      "step": 178
    },
    {
      "epoch": 5.114285714285714,
      "grad_norm": 4.2740125604739685,
      "learning_rate": 6.42578781863613e-08,
      "loss": 0.7062,
      "step": 179
    },
    {
      "epoch": 5.142857142857143,
      "grad_norm": 3.7179175942036378,
      "learning_rate": 6.026312439675551e-08,
      "loss": 0.6132,
      "step": 180
    },
    {
      "epoch": 5.171428571428572,
      "grad_norm": 4.586691811463999,
      "learning_rate": 5.638859027819409e-08,
      "loss": 0.6846,
      "step": 181
    },
    {
      "epoch": 5.2,
      "grad_norm": 4.74824620606461,
      "learning_rate": 5.263533508961826e-08,
      "loss": 0.695,
      "step": 182
    },
    {
      "epoch": 5.228571428571429,
      "grad_norm": 4.552865468215117,
      "learning_rate": 4.9004384933520547e-08,
      "loss": 0.6363,
      "step": 183
    },
    {
      "epoch": 5.257142857142857,
      "grad_norm": 3.868803646466012,
      "learning_rate": 4.549673247541874e-08,
      "loss": 0.6197,
      "step": 184
    },
    {
      "epoch": 5.285714285714286,
      "grad_norm": 4.257422013394126,
      "learning_rate": 4.2113336672471245e-08,
      "loss": 0.6325,
      "step": 185
    },
    {
      "epoch": 5.314285714285714,
      "grad_norm": 4.138411901026596,
      "learning_rate": 3.8855122511307626e-08,
      "loss": 0.6074,
      "step": 186
    },
    {
      "epoch": 5.3428571428571425,
      "grad_norm": 4.386100060157392,
      "learning_rate": 3.572298075514652e-08,
      "loss": 0.6298,
      "step": 187
    },
    {
      "epoch": 5.371428571428572,
      "grad_norm": 3.899133464905973,
      "learning_rate": 3.271776770026963e-08,
      "loss": 0.5991,
      "step": 188
    },
    {
      "epoch": 5.4,
      "grad_norm": 5.227219263192165,
      "learning_rate": 2.9840304941919416e-08,
      "loss": 0.6917,
      "step": 189
    },
    {
      "epoch": 5.428571428571429,
      "grad_norm": 4.194540185559828,
      "learning_rate": 2.7091379149682682e-08,
      "loss": 0.6592,
      "step": 190
    },
    {
      "epoch": 5.457142857142857,
      "grad_norm": 4.103130224713111,
      "learning_rate": 2.4471741852423233e-08,
      "loss": 0.6467,
      "step": 191
    },
    {
      "epoch": 5.485714285714286,
      "grad_norm": 4.401362210384649,
      "learning_rate": 2.1982109232821176e-08,
      "loss": 0.6471,
      "step": 192
    },
    {
      "epoch": 5.514285714285714,
      "grad_norm": 4.4025758572699365,
      "learning_rate": 1.9623161931575926e-08,
      "loss": 0.6537,
      "step": 193
    },
    {
      "epoch": 5.542857142857143,
      "grad_norm": 4.940644969360135,
      "learning_rate": 1.7395544861325718e-08,
      "loss": 0.6647,
      "step": 194
    },
    {
      "epoch": 5.571428571428571,
      "grad_norm": 4.562635610658606,
      "learning_rate": 1.5299867030334813e-08,
      "loss": 0.6422,
      "step": 195
    },
    {
      "epoch": 5.6,
      "grad_norm": 4.58494040795585,
      "learning_rate": 1.3336701375997127e-08,
      "loss": 0.7141,
      "step": 196
    },
    {
      "epoch": 5.628571428571428,
      "grad_norm": 4.240912508697285,
      "learning_rate": 1.1506584608200364e-08,
      "loss": 0.6423,
      "step": 197
    },
    {
      "epoch": 5.6571428571428575,
      "grad_norm": 4.474353548561071,
      "learning_rate": 9.810017062595321e-09,
      "loss": 0.6362,
      "step": 198
    },
    {
      "epoch": 5.685714285714286,
      "grad_norm": 5.263243471264244,
      "learning_rate": 8.247462563808816e-09,
      "loss": 0.6436,
      "step": 199
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 5.629514926131208,
      "learning_rate": 6.819348298638839e-09,
      "loss": 0.6434,
      "step": 200
    },
    {
      "epoch": 5.714285714285714,
      "eval_loss": 1.0673894882202148,
      "eval_runtime": 1.3433,
      "eval_samples_per_second": 17.122,
      "eval_steps_per_second": 0.744,
      "step": 200
    },
    {
      "epoch": 5.742857142857143,
      "grad_norm": 4.388069390318575,
      "learning_rate": 5.526064699265753e-09,
      "loss": 0.6946,
      "step": 201
    },
    {
      "epoch": 5.771428571428571,
      "grad_norm": 3.85292435393061,
      "learning_rate": 4.367965336512403e-09,
      "loss": 0.6113,
      "step": 202
    },
    {
      "epoch": 5.8,
      "grad_norm": 4.3104727339441755,
      "learning_rate": 3.3453668231809283e-09,
      "loss": 0.6493,
      "step": 203
    },
    {
      "epoch": 5.828571428571428,
      "grad_norm": 4.21000584494849,
      "learning_rate": 2.458548727494292e-09,
      "loss": 0.653,
      "step": 204
    },
    {
      "epoch": 5.857142857142857,
      "grad_norm": 5.114411845255924,
      "learning_rate": 1.7077534966650765e-09,
      "loss": 0.6527,
      "step": 205
    },
    {
      "epoch": 5.885714285714286,
      "grad_norm": 4.988848182549562,
      "learning_rate": 1.0931863906127325e-09,
      "loss": 0.6935,
      "step": 206
    },
    {
      "epoch": 5.914285714285715,
      "grad_norm": 4.005152725886777,
      "learning_rate": 6.150154258476314e-10,
      "loss": 0.6559,
      "step": 207
    },
    {
      "epoch": 5.942857142857143,
      "grad_norm": 6.395756261520298,
      "learning_rate": 2.733713295369755e-10,
      "loss": 0.693,
      "step": 208
    },
    {
      "epoch": 5.9714285714285715,
      "grad_norm": 5.465002272924693,
      "learning_rate": 6.834750376549791e-11,
      "loss": 0.7231,
      "step": 209
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.273940345648053,
      "learning_rate": 0.0,
      "loss": 0.6288,
      "step": 210
    },
    {
      "epoch": 6.0,
      "step": 210,
      "total_flos": 18587454013440.0,
      "train_loss": 0.8877790099098569,
      "train_runtime": 1031.8496,
      "train_samples_per_second": 12.932,
      "train_steps_per_second": 0.204
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 210,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 18587454013440.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}