{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 0,
  "global_step": 216,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004629629629629629,
      "grad_norm": 0.447265625,
      "learning_rate": 9.953703703703704e-06,
      "loss": 1.6465,
      "step": 1
    },
    {
      "epoch": 0.009259259259259259,
      "grad_norm": 0.43359375,
      "learning_rate": 9.907407407407408e-06,
      "loss": 1.6266,
      "step": 2
    },
    {
      "epoch": 0.013888888888888888,
      "grad_norm": 0.44921875,
      "learning_rate": 9.861111111111112e-06,
      "loss": 1.6816,
      "step": 3
    },
    {
      "epoch": 0.018518518518518517,
      "grad_norm": 0.416015625,
      "learning_rate": 9.814814814814815e-06,
      "loss": 1.6317,
      "step": 4
    },
    {
      "epoch": 0.023148148148148147,
      "grad_norm": 0.42578125,
      "learning_rate": 9.768518518518519e-06,
      "loss": 1.5986,
      "step": 5
    },
    {
      "epoch": 0.027777777777777776,
      "grad_norm": 0.408203125,
      "learning_rate": 9.722222222222223e-06,
      "loss": 1.6657,
      "step": 6
    },
    {
      "epoch": 0.032407407407407406,
      "grad_norm": 0.384765625,
      "learning_rate": 9.675925925925926e-06,
      "loss": 1.649,
      "step": 7
    },
    {
      "epoch": 0.037037037037037035,
      "grad_norm": 0.357421875,
      "learning_rate": 9.62962962962963e-06,
      "loss": 1.6472,
      "step": 8
    },
    {
      "epoch": 0.041666666666666664,
      "grad_norm": 0.322265625,
      "learning_rate": 9.583333333333335e-06,
      "loss": 1.5746,
      "step": 9
    },
    {
      "epoch": 0.046296296296296294,
      "grad_norm": 0.310546875,
      "learning_rate": 9.537037037037037e-06,
      "loss": 1.585,
      "step": 10
    },
    {
      "epoch": 0.05092592592592592,
      "grad_norm": 0.29296875,
      "learning_rate": 9.490740740740741e-06,
      "loss": 1.5537,
      "step": 11
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 0.275390625,
      "learning_rate": 9.444444444444445e-06,
      "loss": 1.5633,
      "step": 12
    },
    {
      "epoch": 0.06018518518518518,
      "grad_norm": 0.26171875,
      "learning_rate": 9.398148148148148e-06,
      "loss": 1.522,
      "step": 13
    },
    {
      "epoch": 0.06481481481481481,
      "grad_norm": 0.263671875,
      "learning_rate": 9.351851851851854e-06,
      "loss": 1.5383,
      "step": 14
    },
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 0.271484375,
      "learning_rate": 9.305555555555557e-06,
      "loss": 1.5274,
      "step": 15
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 0.283203125,
      "learning_rate": 9.25925925925926e-06,
      "loss": 1.602,
      "step": 16
    },
    {
      "epoch": 0.0787037037037037,
      "grad_norm": 0.29296875,
      "learning_rate": 9.212962962962963e-06,
      "loss": 1.6071,
      "step": 17
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.21875,
      "learning_rate": 9.166666666666666e-06,
      "loss": 1.4715,
      "step": 18
    },
    {
      "epoch": 0.08796296296296297,
      "grad_norm": 0.208984375,
      "learning_rate": 9.120370370370372e-06,
      "loss": 1.4999,
      "step": 19
    },
    {
      "epoch": 0.09259259259259259,
      "grad_norm": 0.20703125,
      "learning_rate": 9.074074074074075e-06,
      "loss": 1.4785,
      "step": 20
    },
    {
      "epoch": 0.09722222222222222,
      "grad_norm": 0.2158203125,
      "learning_rate": 9.027777777777779e-06,
      "loss": 1.4031,
      "step": 21
    },
    {
      "epoch": 0.10185185185185185,
      "grad_norm": 0.1923828125,
      "learning_rate": 8.981481481481483e-06,
      "loss": 1.4511,
      "step": 22
    },
    {
      "epoch": 0.10648148148148148,
      "grad_norm": 0.197265625,
      "learning_rate": 8.935185185185186e-06,
      "loss": 1.4559,
      "step": 23
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.2001953125,
      "learning_rate": 8.888888888888888e-06,
      "loss": 1.4542,
      "step": 24
    },
    {
      "epoch": 0.11574074074074074,
      "grad_norm": 0.212890625,
      "learning_rate": 8.842592592592594e-06,
      "loss": 1.4135,
      "step": 25
    },
    {
      "epoch": 0.12037037037037036,
      "grad_norm": 0.193359375,
      "learning_rate": 8.796296296296297e-06,
      "loss": 1.4799,
      "step": 26
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.1806640625,
      "learning_rate": 8.750000000000001e-06,
      "loss": 1.436,
      "step": 27
    },
    {
      "epoch": 0.12962962962962962,
      "grad_norm": 0.1962890625,
      "learning_rate": 8.703703703703705e-06,
      "loss": 1.4308,
      "step": 28
    },
    {
      "epoch": 0.13425925925925927,
      "grad_norm": 0.224609375,
      "learning_rate": 8.657407407407408e-06,
      "loss": 1.4336,
      "step": 29
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 0.166015625,
      "learning_rate": 8.611111111111112e-06,
      "loss": 1.3854,
      "step": 30
    },
    {
      "epoch": 0.14351851851851852,
      "grad_norm": 0.1826171875,
      "learning_rate": 8.564814814814816e-06,
      "loss": 1.452,
      "step": 31
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 0.1748046875,
      "learning_rate": 8.518518518518519e-06,
      "loss": 1.4227,
      "step": 32
    },
    {
      "epoch": 0.1527777777777778,
      "grad_norm": 0.171875,
      "learning_rate": 8.472222222222223e-06,
      "loss": 1.4428,
      "step": 33
    },
    {
      "epoch": 0.1574074074074074,
      "grad_norm": 0.166015625,
      "learning_rate": 8.425925925925926e-06,
      "loss": 1.3733,
      "step": 34
    },
    {
      "epoch": 0.16203703703703703,
      "grad_norm": 0.1533203125,
      "learning_rate": 8.37962962962963e-06,
      "loss": 1.3572,
      "step": 35
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.1494140625,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.3667,
      "step": 36
    },
    {
      "epoch": 0.1712962962962963,
      "grad_norm": 0.1552734375,
      "learning_rate": 8.287037037037037e-06,
      "loss": 1.3592,
      "step": 37
    },
    {
      "epoch": 0.17592592592592593,
      "grad_norm": 0.1611328125,
      "learning_rate": 8.240740740740741e-06,
      "loss": 1.3731,
      "step": 38
    },
    {
      "epoch": 0.18055555555555555,
      "grad_norm": 0.16015625,
      "learning_rate": 8.194444444444445e-06,
      "loss": 1.4005,
      "step": 39
    },
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 0.2099609375,
      "learning_rate": 8.148148148148148e-06,
      "loss": 1.304,
      "step": 40
    },
    {
      "epoch": 0.18981481481481483,
      "grad_norm": 0.1611328125,
      "learning_rate": 8.101851851851854e-06,
      "loss": 1.3128,
      "step": 41
    },
    {
      "epoch": 0.19444444444444445,
      "grad_norm": 0.154296875,
      "learning_rate": 8.055555555555557e-06,
      "loss": 1.2895,
      "step": 42
    },
    {
      "epoch": 0.19907407407407407,
      "grad_norm": 0.1494140625,
      "learning_rate": 8.00925925925926e-06,
      "loss": 1.2958,
      "step": 43
    },
    {
      "epoch": 0.2037037037037037,
      "grad_norm": 0.16015625,
      "learning_rate": 7.962962962962963e-06,
      "loss": 1.3677,
      "step": 44
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 0.158203125,
      "learning_rate": 7.916666666666667e-06,
      "loss": 1.3493,
      "step": 45
    },
    {
      "epoch": 0.21296296296296297,
      "grad_norm": 0.134765625,
      "learning_rate": 7.870370370370372e-06,
      "loss": 1.2962,
      "step": 46
    },
    {
      "epoch": 0.2175925925925926,
      "grad_norm": 0.1552734375,
      "learning_rate": 7.824074074074076e-06,
      "loss": 1.2977,
      "step": 47
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.1279296875,
      "learning_rate": 7.77777777777778e-06,
      "loss": 1.321,
      "step": 48
    },
    {
      "epoch": 0.22685185185185186,
      "grad_norm": 0.140625,
      "learning_rate": 7.731481481481483e-06,
      "loss": 1.3206,
      "step": 49
    },
    {
      "epoch": 0.23148148148148148,
      "grad_norm": 0.13671875,
      "learning_rate": 7.685185185185185e-06,
      "loss": 1.3124,
      "step": 50
    },
    {
      "epoch": 0.2361111111111111,
      "grad_norm": 0.1669921875,
      "learning_rate": 7.638888888888888e-06,
      "loss": 1.3489,
      "step": 51
    },
    {
      "epoch": 0.24074074074074073,
      "grad_norm": 0.1328125,
      "learning_rate": 7.592592592592594e-06,
      "loss": 1.3231,
      "step": 52
    },
    {
      "epoch": 0.24537037037037038,
      "grad_norm": 0.14453125,
      "learning_rate": 7.546296296296297e-06,
      "loss": 1.2798,
      "step": 53
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1396484375,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.3176,
      "step": 54
    },
    {
      "epoch": 0.25462962962962965,
      "grad_norm": 0.1572265625,
      "learning_rate": 7.453703703703704e-06,
      "loss": 1.3257,
      "step": 55
    },
    {
      "epoch": 0.25925925925925924,
      "grad_norm": 0.130859375,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 1.287,
      "step": 56
    },
    {
      "epoch": 0.2638888888888889,
      "grad_norm": 0.1240234375,
      "learning_rate": 7.361111111111112e-06,
      "loss": 1.2634,
      "step": 57
    },
    {
      "epoch": 0.26851851851851855,
      "grad_norm": 0.1376953125,
      "learning_rate": 7.314814814814816e-06,
      "loss": 1.2709,
      "step": 58
    },
    {
      "epoch": 0.27314814814814814,
      "grad_norm": 0.14453125,
      "learning_rate": 7.268518518518519e-06,
      "loss": 1.2588,
      "step": 59
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.1513671875,
      "learning_rate": 7.222222222222223e-06,
      "loss": 1.3008,
      "step": 60
    },
    {
      "epoch": 0.2824074074074074,
      "grad_norm": 0.1376953125,
      "learning_rate": 7.1759259259259266e-06,
      "loss": 1.3249,
      "step": 61
    },
    {
      "epoch": 0.28703703703703703,
      "grad_norm": 0.17578125,
      "learning_rate": 7.129629629629629e-06,
      "loss": 1.2408,
      "step": 62
    },
    {
      "epoch": 0.2916666666666667,
      "grad_norm": 0.1279296875,
      "learning_rate": 7.083333333333335e-06,
      "loss": 1.2279,
      "step": 63
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 0.1396484375,
      "learning_rate": 7.0370370370370375e-06,
      "loss": 1.2889,
      "step": 64
    },
    {
      "epoch": 0.30092592592592593,
      "grad_norm": 0.1494140625,
      "learning_rate": 6.990740740740741e-06,
      "loss": 1.2507,
      "step": 65
    },
    {
      "epoch": 0.3055555555555556,
      "grad_norm": 0.1376953125,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.2577,
      "step": 66
    },
    {
      "epoch": 0.3101851851851852,
      "grad_norm": 0.150390625,
      "learning_rate": 6.898148148148148e-06,
      "loss": 1.2884,
      "step": 67
    },
    {
      "epoch": 0.3148148148148148,
      "grad_norm": 0.1484375,
      "learning_rate": 6.851851851851853e-06,
      "loss": 1.2481,
      "step": 68
    },
    {
      "epoch": 0.3194444444444444,
      "grad_norm": 0.130859375,
      "learning_rate": 6.8055555555555566e-06,
      "loss": 1.2665,
      "step": 69
    },
    {
      "epoch": 0.32407407407407407,
      "grad_norm": 0.1220703125,
      "learning_rate": 6.75925925925926e-06,
      "loss": 1.2544,
      "step": 70
    },
    {
      "epoch": 0.3287037037037037,
      "grad_norm": 0.130859375,
      "learning_rate": 6.712962962962963e-06,
      "loss": 1.2564,
      "step": 71
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.1416015625,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.2323,
      "step": 72
    },
    {
      "epoch": 0.33796296296296297,
      "grad_norm": 0.12353515625,
      "learning_rate": 6.620370370370371e-06,
      "loss": 1.2141,
      "step": 73
    },
    {
      "epoch": 0.3425925925925926,
      "grad_norm": 0.1494140625,
      "learning_rate": 6.574074074074075e-06,
      "loss": 1.3009,
      "step": 74
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 0.123046875,
      "learning_rate": 6.5277777777777784e-06,
      "loss": 1.2516,
      "step": 75
    },
    {
      "epoch": 0.35185185185185186,
      "grad_norm": 0.1298828125,
      "learning_rate": 6.481481481481482e-06,
      "loss": 1.2344,
      "step": 76
    },
    {
      "epoch": 0.35648148148148145,
      "grad_norm": 0.11474609375,
      "learning_rate": 6.435185185185186e-06,
      "loss": 1.2343,
      "step": 77
    },
    {
      "epoch": 0.3611111111111111,
      "grad_norm": 0.1279296875,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 1.2473,
      "step": 78
    },
    {
      "epoch": 0.36574074074074076,
      "grad_norm": 0.12255859375,
      "learning_rate": 6.342592592592594e-06,
      "loss": 1.2557,
      "step": 79
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.1953125,
      "learning_rate": 6.296296296296297e-06,
      "loss": 1.2889,
      "step": 80
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.1455078125,
      "learning_rate": 6.25e-06,
      "loss": 1.23,
      "step": 81
    },
    {
      "epoch": 0.37962962962962965,
      "grad_norm": 0.1259765625,
      "learning_rate": 6.203703703703704e-06,
      "loss": 1.2176,
      "step": 82
    },
    {
      "epoch": 0.38425925925925924,
      "grad_norm": 0.12255859375,
      "learning_rate": 6.157407407407408e-06,
      "loss": 1.2483,
      "step": 83
    },
    {
      "epoch": 0.3888888888888889,
      "grad_norm": 0.12353515625,
      "learning_rate": 6.111111111111112e-06,
      "loss": 1.1674,
      "step": 84
    },
    {
      "epoch": 0.39351851851851855,
      "grad_norm": 0.1396484375,
      "learning_rate": 6.064814814814816e-06,
      "loss": 1.2136,
      "step": 85
    },
    {
      "epoch": 0.39814814814814814,
      "grad_norm": 0.1201171875,
      "learning_rate": 6.018518518518519e-06,
      "loss": 1.1876,
      "step": 86
    },
    {
      "epoch": 0.4027777777777778,
      "grad_norm": 0.130859375,
      "learning_rate": 5.972222222222222e-06,
      "loss": 1.2409,
      "step": 87
    },
    {
      "epoch": 0.4074074074074074,
      "grad_norm": 0.158203125,
      "learning_rate": 5.925925925925926e-06,
      "loss": 1.2731,
      "step": 88
    },
    {
      "epoch": 0.41203703703703703,
      "grad_norm": 0.134765625,
      "learning_rate": 5.8796296296296295e-06,
      "loss": 1.1931,
      "step": 89
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.119140625,
      "learning_rate": 5.833333333333334e-06,
      "loss": 1.27,
      "step": 90
    },
    {
      "epoch": 0.4212962962962963,
      "grad_norm": 0.1318359375,
      "learning_rate": 5.787037037037038e-06,
      "loss": 1.1893,
      "step": 91
    },
    {
      "epoch": 0.42592592592592593,
      "grad_norm": 0.123046875,
      "learning_rate": 5.740740740740741e-06,
      "loss": 1.1905,
      "step": 92
    },
    {
      "epoch": 0.4305555555555556,
      "grad_norm": 0.1298828125,
      "learning_rate": 5.694444444444445e-06,
      "loss": 1.2595,
      "step": 93
    },
    {
      "epoch": 0.4351851851851852,
      "grad_norm": 0.125,
      "learning_rate": 5.6481481481481485e-06,
      "loss": 1.2024,
      "step": 94
    },
    {
      "epoch": 0.4398148148148148,
      "grad_norm": 0.1259765625,
      "learning_rate": 5.601851851851853e-06,
      "loss": 1.2213,
      "step": 95
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.11279296875,
      "learning_rate": 5.555555555555557e-06,
      "loss": 1.1871,
      "step": 96
    },
    {
      "epoch": 0.44907407407407407,
      "grad_norm": 0.10888671875,
      "learning_rate": 5.5092592592592595e-06,
      "loss": 1.1832,
      "step": 97
    },
    {
      "epoch": 0.4537037037037037,
      "grad_norm": 0.107421875,
      "learning_rate": 5.462962962962963e-06,
      "loss": 1.1935,
      "step": 98
    },
    {
      "epoch": 0.4583333333333333,
      "grad_norm": 0.1162109375,
      "learning_rate": 5.416666666666667e-06,
      "loss": 1.2292,
      "step": 99
    },
    {
      "epoch": 0.46296296296296297,
      "grad_norm": 0.1142578125,
      "learning_rate": 5.370370370370371e-06,
      "loss": 1.193,
      "step": 100
    },
    {
      "epoch": 0.4675925925925926,
      "grad_norm": 0.10302734375,
      "learning_rate": 5.324074074074075e-06,
      "loss": 1.1853,
      "step": 101
    },
    {
      "epoch": 0.4722222222222222,
      "grad_norm": 0.146484375,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 1.2084,
      "step": 102
    },
    {
      "epoch": 0.47685185185185186,
      "grad_norm": 0.11669921875,
      "learning_rate": 5.231481481481482e-06,
      "loss": 1.1933,
      "step": 103
    },
    {
      "epoch": 0.48148148148148145,
      "grad_norm": 0.130859375,
      "learning_rate": 5.185185185185185e-06,
      "loss": 1.2072,
      "step": 104
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 0.1376953125,
      "learning_rate": 5.138888888888889e-06,
      "loss": 1.1547,
      "step": 105
    },
    {
      "epoch": 0.49074074074074076,
      "grad_norm": 0.10693359375,
      "learning_rate": 5.092592592592593e-06,
      "loss": 1.2189,
      "step": 106
    },
    {
      "epoch": 0.49537037037037035,
      "grad_norm": 0.12109375,
      "learning_rate": 5.046296296296297e-06,
      "loss": 1.203,
      "step": 107
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1171875,
      "learning_rate": 5e-06,
      "loss": 1.2275,
      "step": 108
    },
    {
      "epoch": 0.5046296296296297,
      "grad_norm": 0.10791015625,
      "learning_rate": 4.953703703703704e-06,
      "loss": 1.1888,
      "step": 109
    },
    {
      "epoch": 0.5092592592592593,
      "grad_norm": 0.1318359375,
      "learning_rate": 4.907407407407408e-06,
      "loss": 1.2062,
      "step": 110
    },
    {
      "epoch": 0.5138888888888888,
      "grad_norm": 0.119140625,
      "learning_rate": 4.861111111111111e-06,
      "loss": 1.1897,
      "step": 111
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 0.125,
      "learning_rate": 4.814814814814815e-06,
      "loss": 1.2199,
      "step": 112
    },
    {
      "epoch": 0.5231481481481481,
      "grad_norm": 0.10302734375,
      "learning_rate": 4.768518518518519e-06,
      "loss": 1.1873,
      "step": 113
    },
    {
      "epoch": 0.5277777777777778,
      "grad_norm": 0.11865234375,
      "learning_rate": 4.722222222222222e-06,
      "loss": 1.1948,
      "step": 114
    },
    {
      "epoch": 0.5324074074074074,
      "grad_norm": 0.11279296875,
      "learning_rate": 4.675925925925927e-06,
      "loss": 1.1639,
      "step": 115
    },
    {
      "epoch": 0.5370370370370371,
      "grad_norm": 0.11767578125,
      "learning_rate": 4.62962962962963e-06,
      "loss": 1.2247,
      "step": 116
    },
    {
      "epoch": 0.5416666666666666,
      "grad_norm": 0.11669921875,
      "learning_rate": 4.583333333333333e-06,
      "loss": 1.216,
      "step": 117
    },
    {
      "epoch": 0.5462962962962963,
      "grad_norm": 0.11083984375,
      "learning_rate": 4.537037037037038e-06,
      "loss": 1.202,
      "step": 118
    },
    {
      "epoch": 0.5509259259259259,
      "grad_norm": 0.16015625,
      "learning_rate": 4.490740740740741e-06,
      "loss": 1.1947,
      "step": 119
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.11181640625,
      "learning_rate": 4.444444444444444e-06,
      "loss": 1.2281,
      "step": 120
    },
    {
      "epoch": 0.5601851851851852,
      "grad_norm": 0.1416015625,
      "learning_rate": 4.398148148148149e-06,
      "loss": 1.1904,
      "step": 121
    },
    {
      "epoch": 0.5648148148148148,
      "grad_norm": 0.10693359375,
      "learning_rate": 4.351851851851852e-06,
      "loss": 1.1804,
      "step": 122
    },
    {
      "epoch": 0.5694444444444444,
      "grad_norm": 0.1044921875,
      "learning_rate": 4.305555555555556e-06,
      "loss": 1.1783,
      "step": 123
    },
    {
      "epoch": 0.5740740740740741,
      "grad_norm": 0.134765625,
      "learning_rate": 4.2592592592592596e-06,
      "loss": 1.2135,
      "step": 124
    },
    {
      "epoch": 0.5787037037037037,
      "grad_norm": 0.1123046875,
      "learning_rate": 4.212962962962963e-06,
      "loss": 1.2321,
      "step": 125
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.1240234375,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.1895,
      "step": 126
    },
    {
      "epoch": 0.5879629629629629,
      "grad_norm": 0.10791015625,
      "learning_rate": 4.1203703703703705e-06,
      "loss": 1.1869,
      "step": 127
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 0.298828125,
      "learning_rate": 4.074074074074074e-06,
      "loss": 1.3335,
      "step": 128
    },
    {
      "epoch": 0.5972222222222222,
      "grad_norm": 0.123046875,
      "learning_rate": 4.027777777777779e-06,
      "loss": 1.1667,
      "step": 129
    },
    {
      "epoch": 0.6018518518518519,
      "grad_norm": 0.1181640625,
      "learning_rate": 3.9814814814814814e-06,
      "loss": 1.1843,
      "step": 130
    },
    {
      "epoch": 0.6064814814814815,
      "grad_norm": 0.1083984375,
      "learning_rate": 3.935185185185186e-06,
      "loss": 1.1514,
      "step": 131
    },
    {
      "epoch": 0.6111111111111112,
      "grad_norm": 0.1279296875,
      "learning_rate": 3.88888888888889e-06,
      "loss": 1.1706,
      "step": 132
    },
    {
      "epoch": 0.6157407407407407,
      "grad_norm": 0.10302734375,
      "learning_rate": 3.842592592592592e-06,
      "loss": 1.178,
      "step": 133
    },
    {
      "epoch": 0.6203703703703703,
      "grad_norm": 0.1064453125,
      "learning_rate": 3.796296296296297e-06,
      "loss": 1.1989,
      "step": 134
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.125,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 1.22,
      "step": 135
    },
    {
      "epoch": 0.6296296296296297,
      "grad_norm": 0.119140625,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 1.1553,
      "step": 136
    },
    {
      "epoch": 0.6342592592592593,
      "grad_norm": 0.1298828125,
      "learning_rate": 3.657407407407408e-06,
      "loss": 1.2132,
      "step": 137
    },
    {
      "epoch": 0.6388888888888888,
      "grad_norm": 0.1298828125,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 1.1797,
      "step": 138
    },
    {
      "epoch": 0.6435185185185185,
      "grad_norm": 0.1318359375,
      "learning_rate": 3.5648148148148147e-06,
      "loss": 1.1859,
      "step": 139
    },
    {
      "epoch": 0.6481481481481481,
      "grad_norm": 0.12890625,
      "learning_rate": 3.5185185185185187e-06,
      "loss": 1.2317,
      "step": 140
    },
    {
      "epoch": 0.6527777777777778,
      "grad_norm": 0.11474609375,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 1.1821,
      "step": 141
    },
    {
      "epoch": 0.6574074074074074,
      "grad_norm": 0.103515625,
      "learning_rate": 3.4259259259259265e-06,
      "loss": 1.17,
      "step": 142
    },
    {
      "epoch": 0.6620370370370371,
      "grad_norm": 0.15234375,
      "learning_rate": 3.37962962962963e-06,
      "loss": 1.1853,
      "step": 143
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.126953125,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.1787,
      "step": 144
    },
    {
      "epoch": 0.6712962962962963,
      "grad_norm": 0.1259765625,
      "learning_rate": 3.2870370370370374e-06,
      "loss": 1.1586,
      "step": 145
    },
    {
      "epoch": 0.6759259259259259,
      "grad_norm": 0.115234375,
      "learning_rate": 3.240740740740741e-06,
      "loss": 1.1641,
      "step": 146
    },
    {
      "epoch": 0.6805555555555556,
      "grad_norm": 0.197265625,
      "learning_rate": 3.1944444444444443e-06,
      "loss": 1.2853,
      "step": 147
    },
    {
      "epoch": 0.6851851851851852,
      "grad_norm": 0.11181640625,
      "learning_rate": 3.1481481481481483e-06,
      "loss": 1.1666,
      "step": 148
    },
    {
      "epoch": 0.6898148148148148,
      "grad_norm": 0.107421875,
      "learning_rate": 3.101851851851852e-06,
      "loss": 1.1545,
      "step": 149
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.1201171875,
      "learning_rate": 3.055555555555556e-06,
      "loss": 1.1556,
      "step": 150
    },
    {
      "epoch": 0.6990740740740741,
      "grad_norm": 0.1318359375,
      "learning_rate": 3.0092592592592597e-06,
      "loss": 1.1595,
      "step": 151
    },
    {
      "epoch": 0.7037037037037037,
      "grad_norm": 0.11083984375,
      "learning_rate": 2.962962962962963e-06,
      "loss": 1.2139,
      "step": 152
    },
    {
      "epoch": 0.7083333333333334,
      "grad_norm": 0.1396484375,
      "learning_rate": 2.916666666666667e-06,
      "loss": 1.171,
      "step": 153
    },
    {
      "epoch": 0.7129629629629629,
      "grad_norm": 0.11181640625,
      "learning_rate": 2.8703703703703706e-06,
      "loss": 1.2093,
      "step": 154
    },
    {
      "epoch": 0.7175925925925926,
      "grad_norm": 0.1337890625,
      "learning_rate": 2.8240740740740743e-06,
      "loss": 1.1798,
      "step": 155
    },
    {
      "epoch": 0.7222222222222222,
      "grad_norm": 0.1044921875,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 1.1516,
      "step": 156
    },
    {
      "epoch": 0.7268518518518519,
      "grad_norm": 0.1279296875,
      "learning_rate": 2.7314814814814816e-06,
      "loss": 1.1745,
      "step": 157
    },
    {
      "epoch": 0.7314814814814815,
      "grad_norm": 0.130859375,
      "learning_rate": 2.6851851851851856e-06,
      "loss": 1.1929,
      "step": 158
    },
    {
      "epoch": 0.7361111111111112,
      "grad_norm": 0.11669921875,
      "learning_rate": 2.6388888888888893e-06,
      "loss": 1.1423,
      "step": 159
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.1279296875,
      "learning_rate": 2.5925925925925925e-06,
      "loss": 1.1756,
      "step": 160
    },
    {
      "epoch": 0.7453703703703703,
      "grad_norm": 0.11474609375,
      "learning_rate": 2.5462962962962966e-06,
      "loss": 1.1802,
      "step": 161
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.11572265625,
      "learning_rate": 2.5e-06,
      "loss": 1.1388,
      "step": 162
    },
    {
      "epoch": 0.7546296296296297,
      "grad_norm": 0.1416015625,
      "learning_rate": 2.453703703703704e-06,
      "loss": 1.1931,
      "step": 163
    },
    {
      "epoch": 0.7592592592592593,
      "grad_norm": 0.11376953125,
      "learning_rate": 2.4074074074074075e-06,
      "loss": 1.1773,
      "step": 164
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 0.12890625,
      "learning_rate": 2.361111111111111e-06,
      "loss": 1.1975,
      "step": 165
    },
    {
      "epoch": 0.7685185185185185,
      "grad_norm": 0.11279296875,
      "learning_rate": 2.314814814814815e-06,
      "loss": 1.1671,
      "step": 166
    },
    {
      "epoch": 0.7731481481481481,
      "grad_norm": 0.1201171875,
      "learning_rate": 2.268518518518519e-06,
      "loss": 1.1767,
      "step": 167
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.138671875,
      "learning_rate": 2.222222222222222e-06,
      "loss": 1.1994,
      "step": 168
    },
    {
      "epoch": 0.7824074074074074,
      "grad_norm": 0.12109375,
      "learning_rate": 2.175925925925926e-06,
      "loss": 1.1617,
      "step": 169
    },
    {
      "epoch": 0.7870370370370371,
      "grad_norm": 0.12060546875,
      "learning_rate": 2.1296296296296298e-06,
      "loss": 1.1678,
      "step": 170
    },
    {
      "epoch": 0.7916666666666666,
      "grad_norm": 0.1083984375,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.1615,
      "step": 171
    },
    {
      "epoch": 0.7962962962962963,
      "grad_norm": 0.119140625,
      "learning_rate": 2.037037037037037e-06,
      "loss": 1.1803,
      "step": 172
    },
    {
      "epoch": 0.8009259259259259,
      "grad_norm": 0.107421875,
      "learning_rate": 1.9907407407407407e-06,
      "loss": 1.1319,
      "step": 173
    },
    {
      "epoch": 0.8055555555555556,
      "grad_norm": 0.142578125,
      "learning_rate": 1.944444444444445e-06,
      "loss": 1.193,
      "step": 174
    },
    {
      "epoch": 0.8101851851851852,
      "grad_norm": 0.11767578125,
      "learning_rate": 1.8981481481481484e-06,
      "loss": 1.2081,
      "step": 175
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 0.10595703125,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 1.2131,
      "step": 176
    },
    {
      "epoch": 0.8194444444444444,
      "grad_norm": 0.109375,
      "learning_rate": 1.8055555555555557e-06,
      "loss": 1.1399,
      "step": 177
    },
    {
      "epoch": 0.8240740740740741,
      "grad_norm": 0.111328125,
      "learning_rate": 1.7592592592592594e-06,
      "loss": 1.1414,
      "step": 178
    },
    {
      "epoch": 0.8287037037037037,
      "grad_norm": 0.1357421875,
      "learning_rate": 1.7129629629629632e-06,
      "loss": 1.203,
      "step": 179
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.1357421875,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.2028,
      "step": 180
    },
    {
      "epoch": 0.8379629629629629,
      "grad_norm": 0.2109375,
      "learning_rate": 1.6203703703703705e-06,
      "loss": 1.2572,
      "step": 181
    },
    {
      "epoch": 0.8425925925925926,
      "grad_norm": 0.12890625,
      "learning_rate": 1.5740740740740742e-06,
      "loss": 1.1873,
      "step": 182
    },
    {
      "epoch": 0.8472222222222222,
      "grad_norm": 0.12255859375,
      "learning_rate": 1.527777777777778e-06,
      "loss": 1.1527,
      "step": 183
    },
    {
      "epoch": 0.8518518518518519,
      "grad_norm": 0.10693359375,
      "learning_rate": 1.4814814814814815e-06,
      "loss": 1.162,
      "step": 184
    },
    {
      "epoch": 0.8564814814814815,
      "grad_norm": 0.14453125,
      "learning_rate": 1.4351851851851853e-06,
      "loss": 1.1484,
      "step": 185
    },
    {
      "epoch": 0.8611111111111112,
      "grad_norm": 0.1181640625,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 1.1775,
      "step": 186
    },
    {
      "epoch": 0.8657407407407407,
      "grad_norm": 0.1298828125,
      "learning_rate": 1.3425925925925928e-06,
      "loss": 1.1874,
      "step": 187
    },
    {
      "epoch": 0.8703703703703703,
      "grad_norm": 0.1083984375,
      "learning_rate": 1.2962962962962962e-06,
      "loss": 1.1923,
      "step": 188
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.11181640625,
      "learning_rate": 1.25e-06,
      "loss": 1.1916,
      "step": 189
    },
    {
      "epoch": 0.8796296296296297,
      "grad_norm": 0.1123046875,
      "learning_rate": 1.2037037037037037e-06,
      "loss": 1.1725,
      "step": 190
    },
    {
      "epoch": 0.8842592592592593,
      "grad_norm": 0.111328125,
      "learning_rate": 1.1574074074074076e-06,
      "loss": 1.1421,
      "step": 191
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.1181640625,
      "learning_rate": 1.111111111111111e-06,
      "loss": 1.2,
      "step": 192
    },
    {
      "epoch": 0.8935185185185185,
      "grad_norm": 0.11181640625,
      "learning_rate": 1.0648148148148149e-06,
      "loss": 1.1657,
      "step": 193
    },
    {
      "epoch": 0.8981481481481481,
      "grad_norm": 0.109375,
      "learning_rate": 1.0185185185185185e-06,
      "loss": 1.2042,
      "step": 194
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 0.130859375,
      "learning_rate": 9.722222222222224e-07,
      "loss": 1.1444,
      "step": 195
    },
    {
      "epoch": 0.9074074074074074,
      "grad_norm": 0.1240234375,
      "learning_rate": 9.259259259259259e-07,
      "loss": 1.151,
      "step": 196
    },
    {
      "epoch": 0.9120370370370371,
      "grad_norm": 0.10888671875,
      "learning_rate": 8.796296296296297e-07,
      "loss": 1.2003,
      "step": 197
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.11767578125,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.1689,
      "step": 198
    },
    {
      "epoch": 0.9212962962962963,
      "grad_norm": 0.11962890625,
      "learning_rate": 7.870370370370371e-07,
      "loss": 1.1574,
      "step": 199
    },
    {
      "epoch": 0.9259259259259259,
      "grad_norm": 0.1044921875,
      "learning_rate": 7.407407407407407e-07,
      "loss": 1.1167,
      "step": 200
    },
    {
      "epoch": 0.9305555555555556,
      "grad_norm": 0.11083984375,
      "learning_rate": 6.944444444444446e-07,
      "loss": 1.1584,
      "step": 201
    },
    {
      "epoch": 0.9351851851851852,
      "grad_norm": 0.11767578125,
      "learning_rate": 6.481481481481481e-07,
      "loss": 1.1448,
      "step": 202
    },
    {
      "epoch": 0.9398148148148148,
      "grad_norm": 0.15234375,
      "learning_rate": 6.018518518518519e-07,
      "loss": 1.1669,
      "step": 203
    },
    {
      "epoch": 0.9444444444444444,
      "grad_norm": 0.1474609375,
      "learning_rate": 5.555555555555555e-07,
      "loss": 1.1818,
      "step": 204
    },
    {
      "epoch": 0.9490740740740741,
      "grad_norm": 0.125,
      "learning_rate": 5.092592592592593e-07,
      "loss": 1.1887,
      "step": 205
    },
    {
      "epoch": 0.9537037037037037,
      "grad_norm": 0.11083984375,
      "learning_rate": 4.6296296296296297e-07,
      "loss": 1.1684,
      "step": 206
    },
    {
      "epoch": 0.9583333333333334,
      "grad_norm": 0.11279296875,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 1.1349,
      "step": 207
    },
    {
      "epoch": 0.9629629629629629,
      "grad_norm": 0.1240234375,
      "learning_rate": 3.7037037037037036e-07,
      "loss": 1.174,
      "step": 208
    },
    {
      "epoch": 0.9675925925925926,
      "grad_norm": 0.1259765625,
      "learning_rate": 3.2407407407407406e-07,
      "loss": 1.1853,
      "step": 209
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.11474609375,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 1.1687,
      "step": 210
    },
    {
      "epoch": 0.9768518518518519,
      "grad_norm": 0.11669921875,
      "learning_rate": 2.3148148148148148e-07,
      "loss": 1.1681,
      "step": 211
    },
    {
      "epoch": 0.9814814814814815,
      "grad_norm": 0.1181640625,
      "learning_rate": 1.8518518518518518e-07,
      "loss": 1.1854,
      "step": 212
    },
    {
      "epoch": 0.9861111111111112,
      "grad_norm": 0.16015625,
      "learning_rate": 1.3888888888888888e-07,
      "loss": 1.2004,
      "step": 213
    },
    {
      "epoch": 0.9907407407407407,
      "grad_norm": 0.1435546875,
      "learning_rate": 9.259259259259259e-08,
      "loss": 1.1464,
      "step": 214
    },
    {
      "epoch": 0.9953703703703703,
      "grad_norm": 0.130859375,
      "learning_rate": 4.6296296296296295e-08,
      "loss": 1.1237,
      "step": 215
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0,
      "loss": 1.1544,
      "step": 216
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 216,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.568297845164605e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}