{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 0,
  "global_step": 333,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003003003003003003,
      "grad_norm": 0.40234375,
      "learning_rate": 9.96996996996997e-06,
      "loss": 1.9426,
      "step": 1
    },
    {
      "epoch": 0.006006006006006006,
      "grad_norm": 0.37109375,
      "learning_rate": 9.93993993993994e-06,
      "loss": 1.8313,
      "step": 2
    },
    {
      "epoch": 0.009009009009009009,
      "grad_norm": 0.3828125,
      "learning_rate": 9.90990990990991e-06,
      "loss": 1.8864,
      "step": 3
    },
    {
      "epoch": 0.012012012012012012,
      "grad_norm": 0.376953125,
      "learning_rate": 9.879879879879881e-06,
      "loss": 1.8605,
      "step": 4
    },
    {
      "epoch": 0.015015015015015015,
      "grad_norm": 0.365234375,
      "learning_rate": 9.849849849849851e-06,
      "loss": 1.7652,
      "step": 5
    },
    {
      "epoch": 0.018018018018018018,
      "grad_norm": 0.359375,
      "learning_rate": 9.81981981981982e-06,
      "loss": 1.7763,
      "step": 6
    },
    {
      "epoch": 0.021021021021021023,
      "grad_norm": 0.341796875,
      "learning_rate": 9.78978978978979e-06,
      "loss": 1.8606,
      "step": 7
    },
    {
      "epoch": 0.024024024024024024,
      "grad_norm": 0.33203125,
      "learning_rate": 9.75975975975976e-06,
      "loss": 1.7653,
      "step": 8
    },
    {
      "epoch": 0.02702702702702703,
      "grad_norm": 0.39453125,
      "learning_rate": 9.729729729729732e-06,
      "loss": 1.7484,
      "step": 9
    },
    {
      "epoch": 0.03003003003003003,
      "grad_norm": 0.3359375,
      "learning_rate": 9.699699699699701e-06,
      "loss": 1.7719,
      "step": 10
    },
    {
      "epoch": 0.03303303303303303,
      "grad_norm": 0.298828125,
      "learning_rate": 9.669669669669671e-06,
      "loss": 1.7363,
      "step": 11
    },
    {
      "epoch": 0.036036036036036036,
      "grad_norm": 0.287109375,
      "learning_rate": 9.63963963963964e-06,
      "loss": 1.6763,
      "step": 12
    },
    {
      "epoch": 0.03903903903903904,
      "grad_norm": 0.302734375,
      "learning_rate": 9.60960960960961e-06,
      "loss": 1.8133,
      "step": 13
    },
    {
      "epoch": 0.042042042042042045,
      "grad_norm": 0.279296875,
      "learning_rate": 9.57957957957958e-06,
      "loss": 1.687,
      "step": 14
    },
    {
      "epoch": 0.04504504504504504,
      "grad_norm": 0.28125,
      "learning_rate": 9.54954954954955e-06,
      "loss": 1.5893,
      "step": 15
    },
    {
      "epoch": 0.04804804804804805,
      "grad_norm": 0.2890625,
      "learning_rate": 9.51951951951952e-06,
      "loss": 1.7345,
      "step": 16
    },
    {
      "epoch": 0.05105105105105105,
      "grad_norm": 0.287109375,
      "learning_rate": 9.489489489489491e-06,
      "loss": 1.6844,
      "step": 17
    },
    {
      "epoch": 0.05405405405405406,
      "grad_norm": 0.298828125,
      "learning_rate": 9.45945945945946e-06,
      "loss": 1.6965,
      "step": 18
    },
    {
      "epoch": 0.057057057057057055,
      "grad_norm": 0.3046875,
      "learning_rate": 9.42942942942943e-06,
      "loss": 1.692,
      "step": 19
    },
    {
      "epoch": 0.06006006006006006,
      "grad_norm": 0.263671875,
      "learning_rate": 9.3993993993994e-06,
      "loss": 1.624,
      "step": 20
    },
    {
      "epoch": 0.06306306306306306,
      "grad_norm": 0.2421875,
      "learning_rate": 9.36936936936937e-06,
      "loss": 1.5257,
      "step": 21
    },
    {
      "epoch": 0.06606606606606606,
      "grad_norm": 0.2333984375,
      "learning_rate": 9.339339339339341e-06,
      "loss": 1.522,
      "step": 22
    },
    {
      "epoch": 0.06906906906906907,
      "grad_norm": 0.2578125,
      "learning_rate": 9.30930930930931e-06,
      "loss": 1.6405,
      "step": 23
    },
    {
      "epoch": 0.07207207207207207,
      "grad_norm": 0.240234375,
      "learning_rate": 9.27927927927928e-06,
      "loss": 1.6097,
      "step": 24
    },
    {
      "epoch": 0.07507507507507508,
      "grad_norm": 0.2470703125,
      "learning_rate": 9.24924924924925e-06,
      "loss": 1.528,
      "step": 25
    },
    {
      "epoch": 0.07807807807807808,
      "grad_norm": 0.2333984375,
      "learning_rate": 9.21921921921922e-06,
      "loss": 1.5835,
      "step": 26
    },
    {
      "epoch": 0.08108108108108109,
      "grad_norm": 0.23046875,
      "learning_rate": 9.189189189189191e-06,
      "loss": 1.4598,
      "step": 27
    },
    {
      "epoch": 0.08408408408408409,
      "grad_norm": 0.234375,
      "learning_rate": 9.15915915915916e-06,
      "loss": 1.5218,
      "step": 28
    },
    {
      "epoch": 0.08708708708708708,
      "grad_norm": 0.28125,
      "learning_rate": 9.129129129129129e-06,
      "loss": 1.5366,
      "step": 29
    },
    {
      "epoch": 0.09009009009009009,
      "grad_norm": 0.2080078125,
      "learning_rate": 9.0990990990991e-06,
      "loss": 1.4601,
      "step": 30
    },
    {
      "epoch": 0.09309309309309309,
      "grad_norm": 0.2138671875,
      "learning_rate": 9.06906906906907e-06,
      "loss": 1.5354,
      "step": 31
    },
    {
      "epoch": 0.0960960960960961,
      "grad_norm": 0.21875,
      "learning_rate": 9.03903903903904e-06,
      "loss": 1.4839,
      "step": 32
    },
    {
      "epoch": 0.0990990990990991,
      "grad_norm": 0.2275390625,
      "learning_rate": 9.00900900900901e-06,
      "loss": 1.5286,
      "step": 33
    },
    {
      "epoch": 0.1021021021021021,
      "grad_norm": 0.193359375,
      "learning_rate": 8.97897897897898e-06,
      "loss": 1.429,
      "step": 34
    },
    {
      "epoch": 0.10510510510510511,
      "grad_norm": 0.1943359375,
      "learning_rate": 8.94894894894895e-06,
      "loss": 1.4561,
      "step": 35
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 0.2158203125,
      "learning_rate": 8.91891891891892e-06,
      "loss": 1.3834,
      "step": 36
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.201171875,
      "learning_rate": 8.888888888888888e-06,
      "loss": 1.4897,
      "step": 37
    },
    {
      "epoch": 0.11411411411411411,
      "grad_norm": 0.1826171875,
      "learning_rate": 8.85885885885886e-06,
      "loss": 1.3678,
      "step": 38
    },
    {
      "epoch": 0.11711711711711711,
      "grad_norm": 0.193359375,
      "learning_rate": 8.82882882882883e-06,
      "loss": 1.4558,
      "step": 39
    },
    {
      "epoch": 0.12012012012012012,
      "grad_norm": 0.1826171875,
      "learning_rate": 8.798798798798799e-06,
      "loss": 1.3233,
      "step": 40
    },
    {
      "epoch": 0.12312312312312312,
      "grad_norm": 0.181640625,
      "learning_rate": 8.768768768768769e-06,
      "loss": 1.3727,
      "step": 41
    },
    {
      "epoch": 0.12612612612612611,
      "grad_norm": 0.19140625,
      "learning_rate": 8.738738738738739e-06,
      "loss": 1.4092,
      "step": 42
    },
    {
      "epoch": 0.12912912912912913,
      "grad_norm": 0.1708984375,
      "learning_rate": 8.70870870870871e-06,
      "loss": 1.3226,
      "step": 43
    },
    {
      "epoch": 0.13213213213213212,
      "grad_norm": 0.1865234375,
      "learning_rate": 8.67867867867868e-06,
      "loss": 1.3952,
      "step": 44
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 0.1689453125,
      "learning_rate": 8.64864864864865e-06,
      "loss": 1.3698,
      "step": 45
    },
    {
      "epoch": 0.13813813813813813,
      "grad_norm": 0.1669921875,
      "learning_rate": 8.618618618618619e-06,
      "loss": 1.3925,
      "step": 46
    },
    {
      "epoch": 0.14114114114114115,
      "grad_norm": 0.1728515625,
      "learning_rate": 8.588588588588589e-06,
      "loss": 1.3017,
      "step": 47
    },
    {
      "epoch": 0.14414414414414414,
      "grad_norm": 0.1748046875,
      "learning_rate": 8.55855855855856e-06,
      "loss": 1.3779,
      "step": 48
    },
    {
      "epoch": 0.14714714714714713,
      "grad_norm": 0.1865234375,
      "learning_rate": 8.52852852852853e-06,
      "loss": 1.2694,
      "step": 49
    },
    {
      "epoch": 0.15015015015015015,
      "grad_norm": 0.181640625,
      "learning_rate": 8.4984984984985e-06,
      "loss": 1.3452,
      "step": 50
    },
    {
      "epoch": 0.15315315315315314,
      "grad_norm": 0.1748046875,
      "learning_rate": 8.46846846846847e-06,
      "loss": 1.357,
      "step": 51
    },
    {
      "epoch": 0.15615615615615616,
      "grad_norm": 0.1826171875,
      "learning_rate": 8.438438438438439e-06,
      "loss": 1.3165,
      "step": 52
    },
    {
      "epoch": 0.15915915915915915,
      "grad_norm": 0.353515625,
      "learning_rate": 8.408408408408409e-06,
      "loss": 1.3536,
      "step": 53
    },
    {
      "epoch": 0.16216216216216217,
      "grad_norm": 0.1650390625,
      "learning_rate": 8.378378378378378e-06,
      "loss": 1.3193,
      "step": 54
    },
    {
      "epoch": 0.16516516516516516,
      "grad_norm": 0.1806640625,
      "learning_rate": 8.348348348348348e-06,
      "loss": 1.3051,
      "step": 55
    },
    {
      "epoch": 0.16816816816816818,
      "grad_norm": 0.1630859375,
      "learning_rate": 8.31831831831832e-06,
      "loss": 1.336,
      "step": 56
    },
    {
      "epoch": 0.17117117117117117,
      "grad_norm": 0.1591796875,
      "learning_rate": 8.288288288288289e-06,
      "loss": 1.3247,
      "step": 57
    },
    {
      "epoch": 0.17417417417417416,
      "grad_norm": 0.1923828125,
      "learning_rate": 8.258258258258259e-06,
      "loss": 1.3837,
      "step": 58
    },
    {
      "epoch": 0.17717717717717718,
      "grad_norm": 0.2255859375,
      "learning_rate": 8.228228228228229e-06,
      "loss": 1.2709,
      "step": 59
    },
    {
      "epoch": 0.18018018018018017,
      "grad_norm": 0.1630859375,
      "learning_rate": 8.198198198198198e-06,
      "loss": 1.2832,
      "step": 60
    },
    {
      "epoch": 0.1831831831831832,
      "grad_norm": 0.1611328125,
      "learning_rate": 8.16816816816817e-06,
      "loss": 1.2983,
      "step": 61
    },
    {
      "epoch": 0.18618618618618618,
      "grad_norm": 0.158203125,
      "learning_rate": 8.13813813813814e-06,
      "loss": 1.2896,
      "step": 62
    },
    {
      "epoch": 0.1891891891891892,
      "grad_norm": 0.1689453125,
      "learning_rate": 8.108108108108109e-06,
      "loss": 1.3115,
      "step": 63
    },
    {
      "epoch": 0.1921921921921922,
      "grad_norm": 0.1630859375,
      "learning_rate": 8.078078078078079e-06,
      "loss": 1.3231,
      "step": 64
    },
    {
      "epoch": 0.19519519519519518,
      "grad_norm": 0.2119140625,
      "learning_rate": 8.048048048048048e-06,
      "loss": 1.3196,
      "step": 65
    },
    {
      "epoch": 0.1981981981981982,
      "grad_norm": 0.23046875,
      "learning_rate": 8.018018018018018e-06,
      "loss": 1.2569,
      "step": 66
    },
    {
      "epoch": 0.2012012012012012,
      "grad_norm": 0.1572265625,
      "learning_rate": 7.987987987987988e-06,
      "loss": 1.2802,
      "step": 67
    },
    {
      "epoch": 0.2042042042042042,
      "grad_norm": 0.1533203125,
      "learning_rate": 7.95795795795796e-06,
      "loss": 1.2363,
      "step": 68
    },
    {
      "epoch": 0.2072072072072072,
      "grad_norm": 0.357421875,
      "learning_rate": 7.927927927927929e-06,
      "loss": 1.2558,
      "step": 69
    },
    {
      "epoch": 0.21021021021021022,
      "grad_norm": 0.2265625,
      "learning_rate": 7.897897897897899e-06,
      "loss": 1.2902,
      "step": 70
    },
    {
      "epoch": 0.2132132132132132,
      "grad_norm": 0.2158203125,
      "learning_rate": 7.867867867867868e-06,
      "loss": 1.2848,
      "step": 71
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 0.1845703125,
      "learning_rate": 7.837837837837838e-06,
      "loss": 1.3284,
      "step": 72
    },
    {
      "epoch": 0.21921921921921922,
      "grad_norm": 0.1748046875,
      "learning_rate": 7.807807807807808e-06,
      "loss": 1.2584,
      "step": 73
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.166015625,
      "learning_rate": 7.77777777777778e-06,
      "loss": 1.2774,
      "step": 74
    },
    {
      "epoch": 0.22522522522522523,
      "grad_norm": 0.1484375,
      "learning_rate": 7.747747747747749e-06,
      "loss": 1.2155,
      "step": 75
    },
    {
      "epoch": 0.22822822822822822,
      "grad_norm": 0.408203125,
      "learning_rate": 7.717717717717719e-06,
      "loss": 1.2701,
      "step": 76
    },
    {
      "epoch": 0.23123123123123124,
      "grad_norm": 0.1630859375,
      "learning_rate": 7.687687687687688e-06,
      "loss": 1.2338,
      "step": 77
    },
    {
      "epoch": 0.23423423423423423,
      "grad_norm": 0.1650390625,
      "learning_rate": 7.657657657657658e-06,
      "loss": 1.2155,
      "step": 78
    },
    {
      "epoch": 0.23723723723723725,
      "grad_norm": 0.16015625,
      "learning_rate": 7.6276276276276285e-06,
      "loss": 1.2669,
      "step": 79
    },
    {
      "epoch": 0.24024024024024024,
      "grad_norm": 0.16796875,
      "learning_rate": 7.597597597597598e-06,
      "loss": 1.3076,
      "step": 80
    },
    {
      "epoch": 0.24324324324324326,
      "grad_norm": 0.16015625,
      "learning_rate": 7.567567567567569e-06,
      "loss": 1.2316,
      "step": 81
    },
    {
      "epoch": 0.24624624624624625,
      "grad_norm": 0.1591796875,
      "learning_rate": 7.5375375375375385e-06,
      "loss": 1.2527,
      "step": 82
    },
    {
      "epoch": 0.24924924924924924,
      "grad_norm": 0.1640625,
      "learning_rate": 7.507507507507507e-06,
      "loss": 1.2138,
      "step": 83
    },
    {
      "epoch": 0.25225225225225223,
      "grad_norm": 0.1904296875,
      "learning_rate": 7.477477477477479e-06,
      "loss": 1.1995,
      "step": 84
    },
    {
      "epoch": 0.2552552552552553,
      "grad_norm": 0.1572265625,
      "learning_rate": 7.447447447447448e-06,
      "loss": 1.2269,
      "step": 85
    },
    {
      "epoch": 0.25825825825825827,
      "grad_norm": 0.15625,
      "learning_rate": 7.417417417417418e-06,
      "loss": 1.2196,
      "step": 86
    },
    {
      "epoch": 0.26126126126126126,
      "grad_norm": 0.4453125,
      "learning_rate": 7.387387387387388e-06,
      "loss": 1.2454,
      "step": 87
    },
    {
      "epoch": 0.26426426426426425,
      "grad_norm": 0.228515625,
      "learning_rate": 7.3573573573573575e-06,
      "loss": 1.1846,
      "step": 88
    },
    {
      "epoch": 0.2672672672672673,
      "grad_norm": 0.16015625,
      "learning_rate": 7.327327327327328e-06,
      "loss": 1.1562,
      "step": 89
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 0.177734375,
      "learning_rate": 7.297297297297298e-06,
      "loss": 1.1801,
      "step": 90
    },
    {
      "epoch": 0.2732732732732733,
      "grad_norm": 0.1611328125,
      "learning_rate": 7.267267267267268e-06,
      "loss": 1.1426,
      "step": 91
    },
    {
      "epoch": 0.27627627627627627,
      "grad_norm": 0.177734375,
      "learning_rate": 7.237237237237238e-06,
      "loss": 1.1826,
      "step": 92
    },
    {
      "epoch": 0.27927927927927926,
      "grad_norm": 0.1875,
      "learning_rate": 7.207207207207208e-06,
      "loss": 1.1896,
      "step": 93
    },
    {
      "epoch": 0.2822822822822823,
      "grad_norm": 0.1611328125,
      "learning_rate": 7.177177177177178e-06,
      "loss": 1.1674,
      "step": 94
    },
    {
      "epoch": 0.2852852852852853,
      "grad_norm": 0.1904296875,
      "learning_rate": 7.147147147147148e-06,
      "loss": 1.2365,
      "step": 95
    },
    {
      "epoch": 0.2882882882882883,
      "grad_norm": 0.185546875,
      "learning_rate": 7.117117117117117e-06,
      "loss": 1.1366,
      "step": 96
    },
    {
      "epoch": 0.2912912912912913,
      "grad_norm": 0.294921875,
      "learning_rate": 7.087087087087087e-06,
      "loss": 1.2386,
      "step": 97
    },
    {
      "epoch": 0.29429429429429427,
      "grad_norm": 0.1630859375,
      "learning_rate": 7.057057057057057e-06,
      "loss": 1.1802,
      "step": 98
    },
    {
      "epoch": 0.2972972972972973,
      "grad_norm": 0.1591796875,
      "learning_rate": 7.027027027027028e-06,
      "loss": 1.1794,
      "step": 99
    },
    {
      "epoch": 0.3003003003003003,
      "grad_norm": 0.17578125,
      "learning_rate": 6.996996996996997e-06,
      "loss": 1.18,
      "step": 100
    },
    {
      "epoch": 0.3033033033033033,
      "grad_norm": 0.1689453125,
      "learning_rate": 6.966966966966967e-06,
      "loss": 1.1714,
      "step": 101
    },
    {
      "epoch": 0.3063063063063063,
      "grad_norm": 0.1767578125,
      "learning_rate": 6.936936936936938e-06,
      "loss": 1.2206,
      "step": 102
    },
    {
      "epoch": 0.30930930930930933,
      "grad_norm": 0.173828125,
      "learning_rate": 6.906906906906907e-06,
      "loss": 1.136,
      "step": 103
    },
    {
      "epoch": 0.3123123123123123,
      "grad_norm": 0.16796875,
      "learning_rate": 6.876876876876878e-06,
      "loss": 1.2245,
      "step": 104
    },
    {
      "epoch": 0.3153153153153153,
      "grad_norm": 0.15234375,
      "learning_rate": 6.846846846846848e-06,
      "loss": 1.1216,
      "step": 105
    },
    {
      "epoch": 0.3183183183183183,
      "grad_norm": 0.166015625,
      "learning_rate": 6.816816816816817e-06,
      "loss": 1.268,
      "step": 106
    },
    {
      "epoch": 0.3213213213213213,
      "grad_norm": 0.236328125,
      "learning_rate": 6.786786786786788e-06,
      "loss": 1.1769,
      "step": 107
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 0.181640625,
      "learning_rate": 6.7567567567567575e-06,
      "loss": 1.1444,
      "step": 108
    },
    {
      "epoch": 0.32732732732732733,
      "grad_norm": 0.16015625,
      "learning_rate": 6.726726726726728e-06,
      "loss": 1.0906,
      "step": 109
    },
    {
      "epoch": 0.3303303303303303,
      "grad_norm": 0.158203125,
      "learning_rate": 6.696696696696697e-06,
      "loss": 1.2011,
      "step": 110
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.162109375,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.189,
      "step": 111
    },
    {
      "epoch": 0.33633633633633636,
      "grad_norm": 0.302734375,
      "learning_rate": 6.636636636636637e-06,
      "loss": 1.1255,
      "step": 112
    },
    {
      "epoch": 0.33933933933933935,
      "grad_norm": 0.162109375,
      "learning_rate": 6.606606606606607e-06,
      "loss": 1.1752,
      "step": 113
    },
    {
      "epoch": 0.34234234234234234,
      "grad_norm": 0.18359375,
      "learning_rate": 6.5765765765765775e-06,
      "loss": 1.2018,
      "step": 114
    },
    {
      "epoch": 0.34534534534534533,
      "grad_norm": 0.15625,
      "learning_rate": 6.546546546546547e-06,
      "loss": 1.2111,
      "step": 115
    },
    {
      "epoch": 0.3483483483483483,
      "grad_norm": 0.171875,
      "learning_rate": 6.516516516516517e-06,
      "loss": 1.1466,
      "step": 116
    },
    {
      "epoch": 0.35135135135135137,
      "grad_norm": 0.1640625,
      "learning_rate": 6.486486486486487e-06,
      "loss": 1.1442,
      "step": 117
    },
    {
      "epoch": 0.35435435435435436,
      "grad_norm": 0.1748046875,
      "learning_rate": 6.456456456456457e-06,
      "loss": 1.1633,
      "step": 118
    },
    {
      "epoch": 0.35735735735735735,
      "grad_norm": 0.1884765625,
      "learning_rate": 6.426426426426427e-06,
      "loss": 1.1273,
      "step": 119
    },
    {
      "epoch": 0.36036036036036034,
      "grad_norm": 0.162109375,
      "learning_rate": 6.396396396396397e-06,
      "loss": 1.1763,
      "step": 120
    },
    {
      "epoch": 0.3633633633633634,
      "grad_norm": 0.16015625,
      "learning_rate": 6.366366366366366e-06,
      "loss": 1.1546,
      "step": 121
    },
    {
      "epoch": 0.3663663663663664,
      "grad_norm": 0.1611328125,
      "learning_rate": 6.336336336336338e-06,
      "loss": 1.1653,
      "step": 122
    },
    {
      "epoch": 0.36936936936936937,
      "grad_norm": 0.1591796875,
      "learning_rate": 6.3063063063063065e-06,
      "loss": 1.1174,
      "step": 123
    },
    {
      "epoch": 0.37237237237237236,
      "grad_norm": 0.1650390625,
      "learning_rate": 6.276276276276276e-06,
      "loss": 1.1281,
      "step": 124
    },
    {
      "epoch": 0.37537537537537535,
      "grad_norm": 0.1650390625,
      "learning_rate": 6.246246246246247e-06,
      "loss": 1.1666,
      "step": 125
    },
    {
      "epoch": 0.3783783783783784,
      "grad_norm": 0.1689453125,
      "learning_rate": 6.2162162162162164e-06,
      "loss": 1.1531,
      "step": 126
    },
    {
      "epoch": 0.3813813813813814,
      "grad_norm": 0.158203125,
      "learning_rate": 6.186186186186187e-06,
      "loss": 1.1346,
      "step": 127
    },
    {
      "epoch": 0.3843843843843844,
      "grad_norm": 0.22265625,
      "learning_rate": 6.156156156156157e-06,
      "loss": 1.1847,
      "step": 128
    },
    {
      "epoch": 0.38738738738738737,
      "grad_norm": 0.16796875,
      "learning_rate": 6.126126126126126e-06,
      "loss": 1.2132,
      "step": 129
    },
    {
      "epoch": 0.39039039039039036,
      "grad_norm": 0.1826171875,
      "learning_rate": 6.096096096096097e-06,
      "loss": 1.1165,
      "step": 130
    },
    {
      "epoch": 0.3933933933933934,
      "grad_norm": 0.15625,
      "learning_rate": 6.066066066066067e-06,
      "loss": 1.1176,
      "step": 131
    },
    {
      "epoch": 0.3963963963963964,
      "grad_norm": 0.1650390625,
      "learning_rate": 6.036036036036037e-06,
      "loss": 1.1516,
      "step": 132
    },
    {
      "epoch": 0.3993993993993994,
      "grad_norm": 0.189453125,
      "learning_rate": 6.006006006006007e-06,
      "loss": 1.1323,
      "step": 133
    },
    {
      "epoch": 0.4024024024024024,
      "grad_norm": 0.1865234375,
      "learning_rate": 5.975975975975976e-06,
      "loss": 1.1841,
      "step": 134
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 0.275390625,
      "learning_rate": 5.945945945945947e-06,
      "loss": 1.1832,
      "step": 135
    },
    {
      "epoch": 0.4084084084084084,
      "grad_norm": 0.1826171875,
      "learning_rate": 5.915915915915916e-06,
      "loss": 1.146,
      "step": 136
    },
    {
      "epoch": 0.4114114114114114,
      "grad_norm": 0.177734375,
      "learning_rate": 5.885885885885886e-06,
      "loss": 1.1122,
      "step": 137
    },
    {
      "epoch": 0.4144144144144144,
      "grad_norm": 0.1845703125,
      "learning_rate": 5.855855855855856e-06,
      "loss": 1.113,
      "step": 138
    },
    {
      "epoch": 0.4174174174174174,
      "grad_norm": 0.171875,
      "learning_rate": 5.825825825825826e-06,
      "loss": 1.1337,
      "step": 139
    },
    {
      "epoch": 0.42042042042042044,
      "grad_norm": 0.1630859375,
      "learning_rate": 5.7957957957957965e-06,
      "loss": 1.1391,
      "step": 140
    },
    {
      "epoch": 0.42342342342342343,
      "grad_norm": 0.1826171875,
      "learning_rate": 5.765765765765766e-06,
      "loss": 1.1301,
      "step": 141
    },
    {
      "epoch": 0.4264264264264264,
      "grad_norm": 0.197265625,
      "learning_rate": 5.735735735735736e-06,
      "loss": 1.1652,
      "step": 142
    },
    {
      "epoch": 0.4294294294294294,
      "grad_norm": 0.1630859375,
      "learning_rate": 5.7057057057057065e-06,
      "loss": 1.1437,
      "step": 143
    },
    {
      "epoch": 0.43243243243243246,
      "grad_norm": 0.2119140625,
      "learning_rate": 5.675675675675676e-06,
      "loss": 1.1069,
      "step": 144
    },
    {
      "epoch": 0.43543543543543545,
      "grad_norm": 0.1640625,
      "learning_rate": 5.645645645645647e-06,
      "loss": 1.195,
      "step": 145
    },
    {
      "epoch": 0.43843843843843844,
      "grad_norm": 0.158203125,
      "learning_rate": 5.615615615615616e-06,
      "loss": 1.1339,
      "step": 146
    },
    {
      "epoch": 0.44144144144144143,
      "grad_norm": 0.166015625,
      "learning_rate": 5.585585585585585e-06,
      "loss": 1.1131,
      "step": 147
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.1572265625,
      "learning_rate": 5.555555555555557e-06,
      "loss": 1.1069,
      "step": 148
    },
    {
      "epoch": 0.44744744744744747,
      "grad_norm": 0.2373046875,
      "learning_rate": 5.5255255255255255e-06,
      "loss": 1.1584,
      "step": 149
    },
    {
      "epoch": 0.45045045045045046,
      "grad_norm": 0.185546875,
      "learning_rate": 5.495495495495496e-06,
      "loss": 1.1105,
      "step": 150
    },
    {
      "epoch": 0.45345345345345345,
      "grad_norm": 0.1650390625,
      "learning_rate": 5.465465465465466e-06,
      "loss": 1.1319,
      "step": 151
    },
    {
      "epoch": 0.45645645645645644,
      "grad_norm": 0.251953125,
      "learning_rate": 5.4354354354354355e-06,
      "loss": 1.1589,
      "step": 152
    },
    {
      "epoch": 0.4594594594594595,
      "grad_norm": 0.154296875,
      "learning_rate": 5.405405405405406e-06,
      "loss": 1.0879,
      "step": 153
    },
    {
      "epoch": 0.4624624624624625,
      "grad_norm": 0.1806640625,
      "learning_rate": 5.375375375375376e-06,
      "loss": 1.1298,
      "step": 154
    },
    {
      "epoch": 0.46546546546546547,
      "grad_norm": 0.181640625,
      "learning_rate": 5.345345345345346e-06,
      "loss": 1.1458,
      "step": 155
    },
    {
      "epoch": 0.46846846846846846,
      "grad_norm": 0.26171875,
      "learning_rate": 5.315315315315316e-06,
      "loss": 1.1429,
      "step": 156
    },
    {
      "epoch": 0.47147147147147145,
      "grad_norm": 0.181640625,
      "learning_rate": 5.285285285285286e-06,
      "loss": 1.0869,
      "step": 157
    },
    {
      "epoch": 0.4744744744744745,
      "grad_norm": 0.177734375,
      "learning_rate": 5.255255255255256e-06,
      "loss": 1.1667,
      "step": 158
    },
    {
      "epoch": 0.4774774774774775,
      "grad_norm": 0.181640625,
      "learning_rate": 5.225225225225226e-06,
      "loss": 1.1541,
      "step": 159
    },
    {
      "epoch": 0.4804804804804805,
      "grad_norm": 0.1728515625,
      "learning_rate": 5.195195195195195e-06,
      "loss": 1.1825,
      "step": 160
    },
    {
      "epoch": 0.48348348348348347,
      "grad_norm": 0.1669921875,
      "learning_rate": 5.165165165165165e-06,
      "loss": 1.153,
      "step": 161
    },
    {
      "epoch": 0.4864864864864865,
      "grad_norm": 0.1611328125,
      "learning_rate": 5.135135135135135e-06,
      "loss": 1.1311,
      "step": 162
    },
    {
      "epoch": 0.4894894894894895,
      "grad_norm": 0.162109375,
      "learning_rate": 5.105105105105106e-06,
      "loss": 1.0976,
      "step": 163
    },
    {
      "epoch": 0.4924924924924925,
      "grad_norm": 0.16015625,
      "learning_rate": 5.075075075075075e-06,
      "loss": 1.1386,
      "step": 164
    },
    {
      "epoch": 0.4954954954954955,
      "grad_norm": 0.1572265625,
      "learning_rate": 5.045045045045045e-06,
      "loss": 1.0932,
      "step": 165
    },
    {
      "epoch": 0.4984984984984985,
      "grad_norm": 0.1708984375,
      "learning_rate": 5.0150150150150156e-06,
      "loss": 1.1707,
      "step": 166
    },
    {
      "epoch": 0.5015015015015015,
      "grad_norm": 0.1640625,
      "learning_rate": 4.984984984984985e-06,
      "loss": 1.1248,
      "step": 167
    },
    {
      "epoch": 0.5045045045045045,
      "grad_norm": 0.1611328125,
      "learning_rate": 4.954954954954955e-06,
      "loss": 1.1096,
      "step": 168
    },
    {
      "epoch": 0.5075075075075075,
      "grad_norm": 0.1689453125,
      "learning_rate": 4.9249249249249255e-06,
      "loss": 1.1042,
      "step": 169
    },
    {
      "epoch": 0.5105105105105106,
      "grad_norm": 0.1728515625,
      "learning_rate": 4.894894894894895e-06,
      "loss": 1.1279,
      "step": 170
    },
    {
      "epoch": 0.5135135135135135,
      "grad_norm": 0.203125,
      "learning_rate": 4.864864864864866e-06,
      "loss": 1.1772,
      "step": 171
    },
    {
      "epoch": 0.5165165165165165,
      "grad_norm": 0.19140625,
      "learning_rate": 4.8348348348348355e-06,
      "loss": 1.1029,
      "step": 172
    },
    {
      "epoch": 0.5195195195195195,
      "grad_norm": 0.1728515625,
      "learning_rate": 4.804804804804805e-06,
      "loss": 1.0913,
      "step": 173
    },
    {
      "epoch": 0.5225225225225225,
      "grad_norm": 0.1748046875,
      "learning_rate": 4.774774774774775e-06,
      "loss": 1.1033,
      "step": 174
    },
    {
      "epoch": 0.5255255255255256,
      "grad_norm": 0.173828125,
      "learning_rate": 4.7447447447447454e-06,
      "loss": 1.195,
      "step": 175
    },
    {
      "epoch": 0.5285285285285285,
      "grad_norm": 0.1708984375,
      "learning_rate": 4.714714714714715e-06,
      "loss": 1.0949,
      "step": 176
    },
    {
      "epoch": 0.5315315315315315,
      "grad_norm": 0.1689453125,
      "learning_rate": 4.684684684684685e-06,
      "loss": 1.1073,
      "step": 177
    },
    {
      "epoch": 0.5345345345345346,
      "grad_norm": 0.1982421875,
      "learning_rate": 4.654654654654655e-06,
      "loss": 1.1048,
      "step": 178
    },
    {
      "epoch": 0.5375375375375375,
      "grad_norm": 0.26953125,
      "learning_rate": 4.624624624624625e-06,
      "loss": 1.0907,
      "step": 179
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.1640625,
      "learning_rate": 4.594594594594596e-06,
      "loss": 1.0883,
      "step": 180
    },
    {
      "epoch": 0.5435435435435435,
      "grad_norm": 0.16796875,
      "learning_rate": 4.5645645645645645e-06,
      "loss": 1.0493,
      "step": 181
    },
    {
      "epoch": 0.5465465465465466,
      "grad_norm": 0.1591796875,
      "learning_rate": 4.534534534534535e-06,
      "loss": 1.1485,
      "step": 182
    },
    {
      "epoch": 0.5495495495495496,
      "grad_norm": 0.1787109375,
      "learning_rate": 4.504504504504505e-06,
      "loss": 1.0686,
      "step": 183
    },
    {
      "epoch": 0.5525525525525525,
      "grad_norm": 0.1572265625,
      "learning_rate": 4.474474474474475e-06,
      "loss": 1.0874,
      "step": 184
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.1728515625,
      "learning_rate": 4.444444444444444e-06,
      "loss": 1.1569,
      "step": 185
    },
    {
      "epoch": 0.5585585585585585,
      "grad_norm": 0.1767578125,
      "learning_rate": 4.414414414414415e-06,
      "loss": 1.1681,
      "step": 186
    },
    {
      "epoch": 0.5615615615615616,
      "grad_norm": 0.189453125,
      "learning_rate": 4.384384384384384e-06,
      "loss": 1.0983,
      "step": 187
    },
    {
      "epoch": 0.5645645645645646,
      "grad_norm": 0.2099609375,
      "learning_rate": 4.354354354354355e-06,
      "loss": 1.0566,
      "step": 188
    },
    {
      "epoch": 0.5675675675675675,
      "grad_norm": 0.1728515625,
      "learning_rate": 4.324324324324325e-06,
      "loss": 1.1551,
      "step": 189
    },
    {
      "epoch": 0.5705705705705706,
      "grad_norm": 0.162109375,
      "learning_rate": 4.294294294294294e-06,
      "loss": 1.09,
      "step": 190
    },
    {
      "epoch": 0.5735735735735735,
      "grad_norm": 0.19140625,
      "learning_rate": 4.264264264264265e-06,
      "loss": 1.0999,
      "step": 191
    },
    {
      "epoch": 0.5765765765765766,
      "grad_norm": 0.171875,
      "learning_rate": 4.234234234234235e-06,
      "loss": 1.1073,
      "step": 192
    },
    {
      "epoch": 0.5795795795795796,
      "grad_norm": 0.1591796875,
      "learning_rate": 4.204204204204204e-06,
      "loss": 1.0777,
      "step": 193
    },
    {
      "epoch": 0.5825825825825826,
      "grad_norm": 0.1669921875,
      "learning_rate": 4.174174174174174e-06,
      "loss": 1.0979,
      "step": 194
    },
    {
      "epoch": 0.5855855855855856,
      "grad_norm": 0.1884765625,
      "learning_rate": 4.1441441441441446e-06,
      "loss": 1.0813,
      "step": 195
    },
    {
      "epoch": 0.5885885885885885,
      "grad_norm": 0.1669921875,
      "learning_rate": 4.114114114114114e-06,
      "loss": 1.1012,
      "step": 196
    },
    {
      "epoch": 0.5915915915915916,
      "grad_norm": 0.1767578125,
      "learning_rate": 4.084084084084085e-06,
      "loss": 1.1045,
      "step": 197
    },
    {
      "epoch": 0.5945945945945946,
      "grad_norm": 0.171875,
      "learning_rate": 4.0540540540540545e-06,
      "loss": 1.114,
      "step": 198
    },
    {
      "epoch": 0.5975975975975976,
      "grad_norm": 0.1796875,
      "learning_rate": 4.024024024024024e-06,
      "loss": 1.1106,
      "step": 199
    },
    {
      "epoch": 0.6006006006006006,
      "grad_norm": 0.1552734375,
      "learning_rate": 3.993993993993994e-06,
      "loss": 1.1128,
      "step": 200
    },
    {
      "epoch": 0.6036036036036037,
      "grad_norm": 0.1611328125,
      "learning_rate": 3.9639639639639645e-06,
      "loss": 1.0868,
      "step": 201
    },
    {
      "epoch": 0.6066066066066066,
      "grad_norm": 0.169921875,
      "learning_rate": 3.933933933933934e-06,
      "loss": 1.0821,
      "step": 202
    },
    {
      "epoch": 0.6096096096096096,
      "grad_norm": 0.1865234375,
      "learning_rate": 3.903903903903904e-06,
      "loss": 1.1271,
      "step": 203
    },
    {
      "epoch": 0.6126126126126126,
      "grad_norm": 0.1611328125,
      "learning_rate": 3.8738738738738744e-06,
      "loss": 1.0876,
      "step": 204
    },
    {
      "epoch": 0.6156156156156156,
      "grad_norm": 0.1552734375,
      "learning_rate": 3.843843843843844e-06,
      "loss": 1.0997,
      "step": 205
    },
    {
      "epoch": 0.6186186186186187,
      "grad_norm": 0.154296875,
      "learning_rate": 3.8138138138138143e-06,
      "loss": 1.037,
      "step": 206
    },
    {
      "epoch": 0.6216216216216216,
      "grad_norm": 0.1826171875,
      "learning_rate": 3.7837837837837844e-06,
      "loss": 1.0881,
      "step": 207
    },
    {
      "epoch": 0.6246246246246246,
      "grad_norm": 0.16015625,
      "learning_rate": 3.7537537537537537e-06,
      "loss": 1.0631,
      "step": 208
    },
    {
      "epoch": 0.6276276276276276,
      "grad_norm": 0.1669921875,
      "learning_rate": 3.723723723723724e-06,
      "loss": 1.1442,
      "step": 209
    },
    {
      "epoch": 0.6306306306306306,
      "grad_norm": 0.173828125,
      "learning_rate": 3.693693693693694e-06,
      "loss": 1.0803,
      "step": 210
    },
    {
      "epoch": 0.6336336336336337,
      "grad_norm": 0.1728515625,
      "learning_rate": 3.663663663663664e-06,
      "loss": 1.1381,
      "step": 211
    },
    {
      "epoch": 0.6366366366366366,
      "grad_norm": 0.1650390625,
      "learning_rate": 3.633633633633634e-06,
      "loss": 1.0691,
      "step": 212
    },
    {
      "epoch": 0.6396396396396397,
      "grad_norm": 0.1640625,
      "learning_rate": 3.603603603603604e-06,
      "loss": 1.0561,
      "step": 213
    },
    {
      "epoch": 0.6426426426426426,
      "grad_norm": 0.1904296875,
      "learning_rate": 3.573573573573574e-06,
      "loss": 1.1164,
      "step": 214
    },
    {
      "epoch": 0.6456456456456456,
      "grad_norm": 0.162109375,
      "learning_rate": 3.5435435435435437e-06,
      "loss": 1.0622,
      "step": 215
    },
    {
      "epoch": 0.6486486486486487,
      "grad_norm": 0.1630859375,
      "learning_rate": 3.513513513513514e-06,
      "loss": 1.0498,
      "step": 216
    },
    {
      "epoch": 0.6516516516516516,
      "grad_norm": 0.1806640625,
      "learning_rate": 3.4834834834834835e-06,
      "loss": 1.1679,
      "step": 217
    },
    {
      "epoch": 0.6546546546546547,
      "grad_norm": 0.1650390625,
      "learning_rate": 3.4534534534534537e-06,
      "loss": 1.1076,
      "step": 218
    },
    {
      "epoch": 0.6576576576576577,
      "grad_norm": 0.1826171875,
      "learning_rate": 3.423423423423424e-06,
      "loss": 1.1107,
      "step": 219
    },
    {
      "epoch": 0.6606606606606606,
      "grad_norm": 0.16796875,
      "learning_rate": 3.393393393393394e-06,
      "loss": 1.0676,
      "step": 220
    },
    {
      "epoch": 0.6636636636636637,
      "grad_norm": 0.1904296875,
      "learning_rate": 3.363363363363364e-06,
      "loss": 1.0815,
      "step": 221
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.2080078125,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.0982,
      "step": 222
    },
    {
      "epoch": 0.6696696696696697,
      "grad_norm": 0.16796875,
      "learning_rate": 3.3033033033033035e-06,
      "loss": 1.0767,
      "step": 223
    },
    {
      "epoch": 0.6726726726726727,
      "grad_norm": 0.2373046875,
      "learning_rate": 3.2732732732732736e-06,
      "loss": 1.1405,
      "step": 224
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 0.1669921875,
      "learning_rate": 3.2432432432432437e-06,
      "loss": 1.039,
      "step": 225
    },
    {
      "epoch": 0.6786786786786787,
      "grad_norm": 0.2197265625,
      "learning_rate": 3.2132132132132134e-06,
      "loss": 1.0743,
      "step": 226
    },
    {
      "epoch": 0.6816816816816816,
      "grad_norm": 0.1689453125,
      "learning_rate": 3.183183183183183e-06,
      "loss": 1.0779,
      "step": 227
    },
    {
      "epoch": 0.6846846846846847,
      "grad_norm": 0.15625,
      "learning_rate": 3.1531531531531532e-06,
      "loss": 1.1009,
      "step": 228
    },
    {
      "epoch": 0.6876876876876877,
      "grad_norm": 0.1689453125,
      "learning_rate": 3.1231231231231234e-06,
      "loss": 1.0662,
      "step": 229
    },
    {
      "epoch": 0.6906906906906907,
      "grad_norm": 0.1982421875,
      "learning_rate": 3.0930930930930935e-06,
      "loss": 1.1135,
      "step": 230
    },
    {
      "epoch": 0.6936936936936937,
      "grad_norm": 0.1650390625,
      "learning_rate": 3.063063063063063e-06,
      "loss": 1.0928,
      "step": 231
    },
    {
      "epoch": 0.6966966966966966,
      "grad_norm": 0.1875,
      "learning_rate": 3.0330330330330333e-06,
      "loss": 1.0988,
      "step": 232
    },
    {
      "epoch": 0.6996996996996997,
      "grad_norm": 0.1640625,
      "learning_rate": 3.0030030030030034e-06,
      "loss": 1.0743,
      "step": 233
    },
    {
      "epoch": 0.7027027027027027,
      "grad_norm": 0.228515625,
      "learning_rate": 2.9729729729729736e-06,
      "loss": 1.1108,
      "step": 234
    },
    {
      "epoch": 0.7057057057057057,
      "grad_norm": 0.1953125,
      "learning_rate": 2.942942942942943e-06,
      "loss": 1.1121,
      "step": 235
    },
    {
      "epoch": 0.7087087087087087,
      "grad_norm": 0.1669921875,
      "learning_rate": 2.912912912912913e-06,
      "loss": 1.1025,
      "step": 236
    },
    {
      "epoch": 0.7117117117117117,
      "grad_norm": 0.1669921875,
      "learning_rate": 2.882882882882883e-06,
      "loss": 1.1022,
      "step": 237
    },
    {
      "epoch": 0.7147147147147147,
      "grad_norm": 0.1826171875,
      "learning_rate": 2.8528528528528532e-06,
      "loss": 1.148,
      "step": 238
    },
    {
      "epoch": 0.7177177177177178,
      "grad_norm": 0.1650390625,
      "learning_rate": 2.8228228228228234e-06,
      "loss": 1.1191,
      "step": 239
    },
    {
      "epoch": 0.7207207207207207,
      "grad_norm": 0.2373046875,
      "learning_rate": 2.7927927927927926e-06,
      "loss": 1.0489,
      "step": 240
    },
    {
      "epoch": 0.7237237237237237,
      "grad_norm": 0.197265625,
      "learning_rate": 2.7627627627627628e-06,
      "loss": 1.1088,
      "step": 241
    },
    {
      "epoch": 0.7267267267267268,
      "grad_norm": 0.1865234375,
      "learning_rate": 2.732732732732733e-06,
      "loss": 1.16,
      "step": 242
    },
    {
      "epoch": 0.7297297297297297,
      "grad_norm": 0.16796875,
      "learning_rate": 2.702702702702703e-06,
      "loss": 1.0668,
      "step": 243
    },
    {
      "epoch": 0.7327327327327328,
      "grad_norm": 0.32421875,
      "learning_rate": 2.672672672672673e-06,
      "loss": 1.1103,
      "step": 244
    },
    {
      "epoch": 0.7357357357357357,
      "grad_norm": 0.16796875,
      "learning_rate": 2.642642642642643e-06,
      "loss": 1.0715,
      "step": 245
    },
    {
      "epoch": 0.7387387387387387,
      "grad_norm": 0.16796875,
      "learning_rate": 2.612612612612613e-06,
      "loss": 1.068,
      "step": 246
    },
    {
      "epoch": 0.7417417417417418,
      "grad_norm": 0.1787109375,
      "learning_rate": 2.5825825825825827e-06,
      "loss": 1.0419,
      "step": 247
    },
    {
      "epoch": 0.7447447447447447,
      "grad_norm": 0.16796875,
      "learning_rate": 2.552552552552553e-06,
      "loss": 1.1577,
      "step": 248
    },
    {
      "epoch": 0.7477477477477478,
      "grad_norm": 0.16796875,
      "learning_rate": 2.5225225225225225e-06,
      "loss": 1.0416,
      "step": 249
    },
    {
      "epoch": 0.7507507507507507,
      "grad_norm": 0.1689453125,
      "learning_rate": 2.4924924924924926e-06,
      "loss": 1.0559,
      "step": 250
    },
    {
      "epoch": 0.7537537537537538,
      "grad_norm": 0.185546875,
      "learning_rate": 2.4624624624624628e-06,
      "loss": 1.1237,
      "step": 251
    },
    {
      "epoch": 0.7567567567567568,
      "grad_norm": 0.166015625,
      "learning_rate": 2.432432432432433e-06,
      "loss": 1.084,
      "step": 252
    },
    {
      "epoch": 0.7597597597597597,
      "grad_norm": 0.17578125,
      "learning_rate": 2.4024024024024026e-06,
      "loss": 1.1116,
      "step": 253
    },
    {
      "epoch": 0.7627627627627628,
      "grad_norm": 0.1689453125,
      "learning_rate": 2.3723723723723727e-06,
      "loss": 1.0977,
      "step": 254
    },
    {
      "epoch": 0.7657657657657657,
      "grad_norm": 0.2333984375,
      "learning_rate": 2.3423423423423424e-06,
      "loss": 1.0613,
      "step": 255
    },
    {
      "epoch": 0.7687687687687688,
      "grad_norm": 0.1669921875,
      "learning_rate": 2.3123123123123125e-06,
      "loss": 1.0908,
      "step": 256
    },
    {
      "epoch": 0.7717717717717718,
      "grad_norm": 0.2138671875,
      "learning_rate": 2.2822822822822822e-06,
      "loss": 1.1243,
      "step": 257
    },
    {
      "epoch": 0.7747747747747747,
      "grad_norm": 0.353515625,
      "learning_rate": 2.2522522522522524e-06,
      "loss": 1.1294,
      "step": 258
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.203125,
      "learning_rate": 2.222222222222222e-06,
      "loss": 1.1158,
      "step": 259
    },
    {
      "epoch": 0.7807807807807807,
      "grad_norm": 0.166015625,
      "learning_rate": 2.192192192192192e-06,
      "loss": 1.0986,
      "step": 260
    },
    {
      "epoch": 0.7837837837837838,
      "grad_norm": 0.1669921875,
      "learning_rate": 2.1621621621621623e-06,
      "loss": 1.0765,
      "step": 261
    },
    {
      "epoch": 0.7867867867867868,
      "grad_norm": 0.1845703125,
      "learning_rate": 2.1321321321321325e-06,
      "loss": 1.103,
      "step": 262
    },
    {
      "epoch": 0.7897897897897898,
      "grad_norm": 0.1845703125,
      "learning_rate": 2.102102102102102e-06,
      "loss": 1.0601,
      "step": 263
    },
    {
      "epoch": 0.7927927927927928,
      "grad_norm": 0.1630859375,
      "learning_rate": 2.0720720720720723e-06,
      "loss": 1.0695,
      "step": 264
    },
    {
      "epoch": 0.7957957957957958,
      "grad_norm": 0.251953125,
      "learning_rate": 2.0420420420420424e-06,
      "loss": 1.124,
      "step": 265
    },
    {
      "epoch": 0.7987987987987988,
      "grad_norm": 0.173828125,
      "learning_rate": 2.012012012012012e-06,
      "loss": 1.1085,
      "step": 266
    },
    {
      "epoch": 0.8018018018018018,
      "grad_norm": 0.212890625,
      "learning_rate": 1.9819819819819822e-06,
      "loss": 1.0957,
      "step": 267
    },
    {
      "epoch": 0.8048048048048048,
      "grad_norm": 0.1884765625,
      "learning_rate": 1.951951951951952e-06,
      "loss": 1.1813,
      "step": 268
    },
    {
      "epoch": 0.8078078078078078,
      "grad_norm": 0.181640625,
      "learning_rate": 1.921921921921922e-06,
      "loss": 1.1231,
      "step": 269
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 0.169921875,
      "learning_rate": 1.8918918918918922e-06,
      "loss": 1.0982,
      "step": 270
    },
    {
      "epoch": 0.8138138138138138,
      "grad_norm": 0.470703125,
      "learning_rate": 1.861861861861862e-06,
      "loss": 1.0698,
      "step": 271
    },
    {
      "epoch": 0.8168168168168168,
      "grad_norm": 0.1708984375,
      "learning_rate": 1.831831831831832e-06,
      "loss": 1.045,
      "step": 272
    },
    {
      "epoch": 0.8198198198198198,
      "grad_norm": 0.298828125,
      "learning_rate": 1.801801801801802e-06,
      "loss": 1.1209,
      "step": 273
    },
    {
      "epoch": 0.8228228228228228,
      "grad_norm": 0.169921875,
      "learning_rate": 1.7717717717717719e-06,
      "loss": 1.1016,
      "step": 274
    },
    {
      "epoch": 0.8258258258258259,
      "grad_norm": 0.287109375,
      "learning_rate": 1.7417417417417418e-06,
      "loss": 1.0462,
      "step": 275
    },
    {
      "epoch": 0.8288288288288288,
      "grad_norm": 0.1748046875,
      "learning_rate": 1.711711711711712e-06,
      "loss": 1.1063,
      "step": 276
    },
    {
      "epoch": 0.8318318318318318,
      "grad_norm": 0.1787109375,
      "learning_rate": 1.681681681681682e-06,
      "loss": 1.0933,
      "step": 277
    },
    {
      "epoch": 0.8348348348348348,
      "grad_norm": 0.17578125,
      "learning_rate": 1.6516516516516517e-06,
      "loss": 1.0975,
      "step": 278
    },
    {
      "epoch": 0.8378378378378378,
      "grad_norm": 0.1923828125,
      "learning_rate": 1.6216216216216219e-06,
      "loss": 1.0735,
      "step": 279
    },
    {
      "epoch": 0.8408408408408409,
      "grad_norm": 0.1708984375,
      "learning_rate": 1.5915915915915916e-06,
      "loss": 1.0195,
      "step": 280
    },
    {
      "epoch": 0.8438438438438438,
      "grad_norm": 0.2060546875,
      "learning_rate": 1.5615615615615617e-06,
      "loss": 1.0686,
      "step": 281
    },
    {
      "epoch": 0.8468468468468469,
      "grad_norm": 0.1708984375,
      "learning_rate": 1.5315315315315316e-06,
      "loss": 1.1048,
      "step": 282
    },
    {
      "epoch": 0.8498498498498499,
      "grad_norm": 0.171875,
      "learning_rate": 1.5015015015015017e-06,
      "loss": 1.0842,
      "step": 283
    },
    {
      "epoch": 0.8528528528528528,
      "grad_norm": 0.1884765625,
      "learning_rate": 1.4714714714714714e-06,
      "loss": 1.0679,
      "step": 284
    },
    {
      "epoch": 0.8558558558558559,
      "grad_norm": 0.1787109375,
      "learning_rate": 1.4414414414414416e-06,
      "loss": 1.1843,
      "step": 285
    },
    {
      "epoch": 0.8588588588588588,
      "grad_norm": 0.17578125,
      "learning_rate": 1.4114114114114117e-06,
      "loss": 1.1292,
      "step": 286
    },
    {
      "epoch": 0.8618618618618619,
      "grad_norm": 0.224609375,
      "learning_rate": 1.3813813813813814e-06,
      "loss": 1.0118,
      "step": 287
    },
    {
      "epoch": 0.8648648648648649,
      "grad_norm": 0.2236328125,
      "learning_rate": 1.3513513513513515e-06,
      "loss": 1.1594,
      "step": 288
    },
    {
      "epoch": 0.8678678678678678,
      "grad_norm": 0.171875,
      "learning_rate": 1.3213213213213214e-06,
      "loss": 1.0357,
      "step": 289
    },
    {
      "epoch": 0.8708708708708709,
      "grad_norm": 0.1630859375,
      "learning_rate": 1.2912912912912913e-06,
      "loss": 1.0619,
      "step": 290
    },
    {
      "epoch": 0.8738738738738738,
      "grad_norm": 0.310546875,
      "learning_rate": 1.2612612612612613e-06,
      "loss": 1.1613,
      "step": 291
    },
    {
      "epoch": 0.8768768768768769,
      "grad_norm": 0.306640625,
      "learning_rate": 1.2312312312312314e-06,
      "loss": 1.0938,
      "step": 292
    },
    {
      "epoch": 0.8798798798798799,
      "grad_norm": 0.17578125,
      "learning_rate": 1.2012012012012013e-06,
      "loss": 1.1483,
      "step": 293
    },
    {
      "epoch": 0.8828828828828829,
      "grad_norm": 0.197265625,
      "learning_rate": 1.1711711711711712e-06,
      "loss": 1.0938,
      "step": 294
    },
    {
      "epoch": 0.8858858858858859,
      "grad_norm": 0.1943359375,
      "learning_rate": 1.1411411411411411e-06,
      "loss": 1.0665,
      "step": 295
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.1630859375,
      "learning_rate": 1.111111111111111e-06,
      "loss": 1.0748,
      "step": 296
    },
    {
      "epoch": 0.8918918918918919,
      "grad_norm": 0.34765625,
      "learning_rate": 1.0810810810810812e-06,
      "loss": 1.1238,
      "step": 297
    },
    {
      "epoch": 0.8948948948948949,
      "grad_norm": 0.1796875,
      "learning_rate": 1.051051051051051e-06,
      "loss": 1.0461,
      "step": 298
    },
    {
      "epoch": 0.8978978978978979,
      "grad_norm": 0.2138671875,
      "learning_rate": 1.0210210210210212e-06,
      "loss": 1.0559,
      "step": 299
    },
    {
      "epoch": 0.9009009009009009,
      "grad_norm": 0.171875,
      "learning_rate": 9.909909909909911e-07,
      "loss": 1.0833,
      "step": 300
    },
    {
      "epoch": 0.9039039039039038,
      "grad_norm": 0.181640625,
      "learning_rate": 9.60960960960961e-07,
      "loss": 1.1111,
      "step": 301
    },
    {
      "epoch": 0.9069069069069069,
      "grad_norm": 0.197265625,
      "learning_rate": 9.30930930930931e-07,
      "loss": 1.1452,
      "step": 302
    },
    {
      "epoch": 0.9099099099099099,
      "grad_norm": 0.1640625,
      "learning_rate": 9.00900900900901e-07,
      "loss": 1.0994,
      "step": 303
    },
    {
      "epoch": 0.9129129129129129,
      "grad_norm": 0.1708984375,
      "learning_rate": 8.708708708708709e-07,
      "loss": 1.1092,
      "step": 304
    },
    {
      "epoch": 0.9159159159159159,
      "grad_norm": 0.1748046875,
      "learning_rate": 8.40840840840841e-07,
      "loss": 1.0825,
      "step": 305
    },
    {
      "epoch": 0.918918918918919,
      "grad_norm": 0.18359375,
      "learning_rate": 8.108108108108109e-07,
      "loss": 1.0963,
      "step": 306
    },
    {
      "epoch": 0.9219219219219219,
      "grad_norm": 0.162109375,
      "learning_rate": 7.807807807807808e-07,
      "loss": 1.0443,
      "step": 307
    },
    {
      "epoch": 0.924924924924925,
      "grad_norm": 0.16796875,
      "learning_rate": 7.507507507507509e-07,
      "loss": 1.0719,
      "step": 308
    },
    {
      "epoch": 0.9279279279279279,
      "grad_norm": 0.1640625,
      "learning_rate": 7.207207207207208e-07,
      "loss": 1.0749,
      "step": 309
    },
    {
      "epoch": 0.9309309309309309,
      "grad_norm": 0.1943359375,
      "learning_rate": 6.906906906906907e-07,
      "loss": 1.1254,
      "step": 310
    },
    {
      "epoch": 0.933933933933934,
      "grad_norm": 0.17578125,
      "learning_rate": 6.606606606606607e-07,
      "loss": 1.094,
      "step": 311
    },
    {
      "epoch": 0.9369369369369369,
      "grad_norm": 0.1591796875,
      "learning_rate": 6.306306306306306e-07,
      "loss": 1.0437,
      "step": 312
    },
    {
      "epoch": 0.93993993993994,
      "grad_norm": 0.1669921875,
      "learning_rate": 6.006006006006006e-07,
      "loss": 1.0943,
      "step": 313
    },
    {
      "epoch": 0.9429429429429429,
      "grad_norm": 0.1806640625,
      "learning_rate": 5.705705705705706e-07,
      "loss": 1.0969,
      "step": 314
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 0.171875,
      "learning_rate": 5.405405405405406e-07,
      "loss": 1.0641,
      "step": 315
    },
    {
      "epoch": 0.948948948948949,
      "grad_norm": 0.1611328125,
      "learning_rate": 5.105105105105106e-07,
      "loss": 1.07,
      "step": 316
    },
    {
      "epoch": 0.9519519519519519,
      "grad_norm": 0.1708984375,
      "learning_rate": 4.804804804804805e-07,
      "loss": 1.1072,
      "step": 317
    },
    {
      "epoch": 0.954954954954955,
      "grad_norm": 0.1796875,
      "learning_rate": 4.504504504504505e-07,
      "loss": 1.1152,
      "step": 318
    },
    {
      "epoch": 0.9579579579579579,
      "grad_norm": 0.259765625,
      "learning_rate": 4.204204204204205e-07,
      "loss": 1.1443,
      "step": 319
    },
    {
      "epoch": 0.960960960960961,
      "grad_norm": 0.2578125,
      "learning_rate": 3.903903903903904e-07,
      "loss": 1.1815,
      "step": 320
    },
    {
      "epoch": 0.963963963963964,
      "grad_norm": 0.1640625,
      "learning_rate": 3.603603603603604e-07,
      "loss": 1.0854,
      "step": 321
    },
    {
      "epoch": 0.9669669669669669,
      "grad_norm": 0.2109375,
      "learning_rate": 3.3033033033033036e-07,
      "loss": 1.1102,
      "step": 322
    },
    {
      "epoch": 0.96996996996997,
      "grad_norm": 0.1767578125,
      "learning_rate": 3.003003003003003e-07,
      "loss": 1.0807,
      "step": 323
    },
    {
      "epoch": 0.972972972972973,
      "grad_norm": 0.1953125,
      "learning_rate": 2.702702702702703e-07,
      "loss": 1.0706,
      "step": 324
    },
    {
      "epoch": 0.975975975975976,
      "grad_norm": 0.2353515625,
      "learning_rate": 2.4024024024024026e-07,
      "loss": 1.0874,
      "step": 325
    },
    {
      "epoch": 0.978978978978979,
      "grad_norm": 0.171875,
      "learning_rate": 2.1021021021021025e-07,
      "loss": 1.1157,
      "step": 326
    },
    {
      "epoch": 0.9819819819819819,
      "grad_norm": 0.1669921875,
      "learning_rate": 1.801801801801802e-07,
      "loss": 1.0717,
      "step": 327
    },
    {
      "epoch": 0.984984984984985,
      "grad_norm": 0.326171875,
      "learning_rate": 1.5015015015015016e-07,
      "loss": 1.0993,
      "step": 328
    },
    {
      "epoch": 0.987987987987988,
      "grad_norm": 0.1728515625,
      "learning_rate": 1.2012012012012013e-07,
      "loss": 1.0882,
      "step": 329
    },
    {
      "epoch": 0.990990990990991,
      "grad_norm": 0.1708984375,
      "learning_rate": 9.00900900900901e-08,
      "loss": 1.081,
      "step": 330
    },
    {
      "epoch": 0.993993993993994,
      "grad_norm": 0.162109375,
      "learning_rate": 6.006006006006006e-08,
      "loss": 1.083,
      "step": 331
    },
    {
      "epoch": 0.996996996996997,
      "grad_norm": 0.1845703125,
      "learning_rate": 3.003003003003003e-08,
      "loss": 1.1135,
      "step": 332
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0,
      "loss": 1.1455,
      "step": 333
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 333,
  "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": 1.0524472989712384e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}