{
  "best_global_step": null,
  "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": 1.2840936183929443,
      "learning_rate": 1e-05,
      "loss": 2.592,
      "step": 1
    },
    {
      "epoch": 0.009259259259259259,
      "grad_norm": 1.2497975826263428,
      "learning_rate": 9.953703703703704e-06,
      "loss": 2.5955,
      "step": 2
    },
    {
      "epoch": 0.013888888888888888,
      "grad_norm": 1.1559158563613892,
      "learning_rate": 9.907407407407408e-06,
      "loss": 2.5476,
      "step": 3
    },
    {
      "epoch": 0.018518518518518517,
      "grad_norm": 1.2224756479263306,
      "learning_rate": 9.861111111111112e-06,
      "loss": 2.5497,
      "step": 4
    },
    {
      "epoch": 0.023148148148148147,
      "grad_norm": 1.1082508563995361,
      "learning_rate": 9.814814814814815e-06,
      "loss": 2.5263,
      "step": 5
    },
    {
      "epoch": 0.027777777777777776,
      "grad_norm": 1.0577224493026733,
      "learning_rate": 9.768518518518519e-06,
      "loss": 2.4677,
      "step": 6
    },
    {
      "epoch": 0.032407407407407406,
      "grad_norm": 1.0209864377975464,
      "learning_rate": 9.722222222222223e-06,
      "loss": 2.4699,
      "step": 7
    },
    {
      "epoch": 0.037037037037037035,
      "grad_norm": 1.0040022134780884,
      "learning_rate": 9.675925925925926e-06,
      "loss": 2.4408,
      "step": 8
    },
    {
      "epoch": 0.041666666666666664,
      "grad_norm": 1.003764271736145,
      "learning_rate": 9.62962962962963e-06,
      "loss": 2.4429,
      "step": 9
    },
    {
      "epoch": 0.046296296296296294,
      "grad_norm": 0.9816301465034485,
      "learning_rate": 9.583333333333335e-06,
      "loss": 2.4405,
      "step": 10
    },
    {
      "epoch": 0.05092592592592592,
      "grad_norm": 0.9731383323669434,
      "learning_rate": 9.537037037037037e-06,
      "loss": 2.3876,
      "step": 11
    },
    {
      "epoch": 0.05555555555555555,
      "grad_norm": 0.9626306891441345,
      "learning_rate": 9.490740740740741e-06,
      "loss": 2.458,
      "step": 12
    },
    {
      "epoch": 0.06018518518518518,
      "grad_norm": 0.8909545540809631,
      "learning_rate": 9.444444444444445e-06,
      "loss": 2.3516,
      "step": 13
    },
    {
      "epoch": 0.06481481481481481,
      "grad_norm": 0.863825261592865,
      "learning_rate": 9.398148148148148e-06,
      "loss": 2.3272,
      "step": 14
    },
    {
      "epoch": 0.06944444444444445,
      "grad_norm": 0.8534576892852783,
      "learning_rate": 9.351851851851854e-06,
      "loss": 2.3088,
      "step": 15
    },
    {
      "epoch": 0.07407407407407407,
      "grad_norm": 0.8410207033157349,
      "learning_rate": 9.305555555555557e-06,
      "loss": 2.3867,
      "step": 16
    },
    {
      "epoch": 0.0787037037037037,
      "grad_norm": 0.8216672539710999,
      "learning_rate": 9.25925925925926e-06,
      "loss": 2.2946,
      "step": 17
    },
    {
      "epoch": 0.08333333333333333,
      "grad_norm": 0.7917430400848389,
      "learning_rate": 9.212962962962963e-06,
      "loss": 2.3206,
      "step": 18
    },
    {
      "epoch": 0.08796296296296297,
      "grad_norm": 0.7604931592941284,
      "learning_rate": 9.166666666666666e-06,
      "loss": 2.2694,
      "step": 19
    },
    {
      "epoch": 0.09259259259259259,
      "grad_norm": 0.7862806916236877,
      "learning_rate": 9.120370370370372e-06,
      "loss": 2.3059,
      "step": 20
    },
    {
      "epoch": 0.09722222222222222,
      "grad_norm": 0.7812303304672241,
      "learning_rate": 9.074074074074075e-06,
      "loss": 2.2944,
      "step": 21
    },
    {
      "epoch": 0.10185185185185185,
      "grad_norm": 0.6887988448143005,
      "learning_rate": 9.027777777777779e-06,
      "loss": 2.1687,
      "step": 22
    },
    {
      "epoch": 0.10648148148148148,
      "grad_norm": 0.7094734907150269,
      "learning_rate": 8.981481481481483e-06,
      "loss": 2.25,
      "step": 23
    },
    {
      "epoch": 0.1111111111111111,
      "grad_norm": 0.735929548740387,
      "learning_rate": 8.935185185185186e-06,
      "loss": 2.2761,
      "step": 24
    },
    {
      "epoch": 0.11574074074074074,
      "grad_norm": 0.7055741548538208,
      "learning_rate": 8.888888888888888e-06,
      "loss": 2.22,
      "step": 25
    },
    {
      "epoch": 0.12037037037037036,
      "grad_norm": 0.6612609028816223,
      "learning_rate": 8.842592592592594e-06,
      "loss": 2.1327,
      "step": 26
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.6664066910743713,
      "learning_rate": 8.796296296296297e-06,
      "loss": 2.2184,
      "step": 27
    },
    {
      "epoch": 0.12962962962962962,
      "grad_norm": 0.6584565043449402,
      "learning_rate": 8.750000000000001e-06,
      "loss": 2.2032,
      "step": 28
    },
    {
      "epoch": 0.13425925925925927,
      "grad_norm": 0.653407096862793,
      "learning_rate": 8.703703703703705e-06,
      "loss": 2.1957,
      "step": 29
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 0.6251302361488342,
      "learning_rate": 8.657407407407408e-06,
      "loss": 2.1389,
      "step": 30
    },
    {
      "epoch": 0.14351851851851852,
      "grad_norm": 0.6295129656791687,
      "learning_rate": 8.611111111111112e-06,
      "loss": 2.1333,
      "step": 31
    },
    {
      "epoch": 0.14814814814814814,
      "grad_norm": 0.6065976619720459,
      "learning_rate": 8.564814814814816e-06,
      "loss": 2.1016,
      "step": 32
    },
    {
      "epoch": 0.1527777777777778,
      "grad_norm": 0.5833759307861328,
      "learning_rate": 8.518518518518519e-06,
      "loss": 2.1216,
      "step": 33
    },
    {
      "epoch": 0.1574074074074074,
      "grad_norm": 0.580999493598938,
      "learning_rate": 8.472222222222223e-06,
      "loss": 2.1216,
      "step": 34
    },
    {
      "epoch": 0.16203703703703703,
      "grad_norm": 0.5744687914848328,
      "learning_rate": 8.425925925925926e-06,
      "loss": 2.0704,
      "step": 35
    },
    {
      "epoch": 0.16666666666666666,
      "grad_norm": 0.5671817660331726,
      "learning_rate": 8.37962962962963e-06,
      "loss": 2.1015,
      "step": 36
    },
    {
      "epoch": 0.1712962962962963,
      "grad_norm": 0.5558776259422302,
      "learning_rate": 8.333333333333334e-06,
      "loss": 2.107,
      "step": 37
    },
    {
      "epoch": 0.17592592592592593,
      "grad_norm": 0.5685739517211914,
      "learning_rate": 8.287037037037037e-06,
      "loss": 2.1195,
      "step": 38
    },
    {
      "epoch": 0.18055555555555555,
      "grad_norm": 0.5277007818222046,
      "learning_rate": 8.240740740740741e-06,
      "loss": 2.0058,
      "step": 39
    },
    {
      "epoch": 0.18518518518518517,
      "grad_norm": 0.5531312227249146,
      "learning_rate": 8.194444444444445e-06,
      "loss": 2.0461,
      "step": 40
    },
    {
      "epoch": 0.18981481481481483,
      "grad_norm": 0.5435283780097961,
      "learning_rate": 8.148148148148148e-06,
      "loss": 2.0629,
      "step": 41
    },
    {
      "epoch": 0.19444444444444445,
      "grad_norm": 0.5239248275756836,
      "learning_rate": 8.101851851851854e-06,
      "loss": 2.0484,
      "step": 42
    },
    {
      "epoch": 0.19907407407407407,
      "grad_norm": 0.5030505061149597,
      "learning_rate": 8.055555555555557e-06,
      "loss": 2.0164,
      "step": 43
    },
    {
      "epoch": 0.2037037037037037,
      "grad_norm": 0.5091701149940491,
      "learning_rate": 8.00925925925926e-06,
      "loss": 2.028,
      "step": 44
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 0.5030836462974548,
      "learning_rate": 7.962962962962963e-06,
      "loss": 2.0184,
      "step": 45
    },
    {
      "epoch": 0.21296296296296297,
      "grad_norm": 0.47289565205574036,
      "learning_rate": 7.916666666666667e-06,
      "loss": 2.009,
      "step": 46
    },
    {
      "epoch": 0.2175925925925926,
      "grad_norm": 0.5161042809486389,
      "learning_rate": 7.870370370370372e-06,
      "loss": 2.0182,
      "step": 47
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.48933666944503784,
      "learning_rate": 7.824074074074076e-06,
      "loss": 1.9673,
      "step": 48
    },
    {
      "epoch": 0.22685185185185186,
      "grad_norm": 0.4742085933685303,
      "learning_rate": 7.77777777777778e-06,
      "loss": 2.0158,
      "step": 49
    },
    {
      "epoch": 0.23148148148148148,
      "grad_norm": 0.47005364298820496,
      "learning_rate": 7.731481481481483e-06,
      "loss": 1.9924,
      "step": 50
    },
    {
      "epoch": 0.2361111111111111,
      "grad_norm": 0.4893855154514313,
      "learning_rate": 7.685185185185185e-06,
      "loss": 1.9469,
      "step": 51
    },
    {
      "epoch": 0.24074074074074073,
      "grad_norm": 0.4696884751319885,
      "learning_rate": 7.638888888888888e-06,
      "loss": 1.9769,
      "step": 52
    },
    {
      "epoch": 0.24537037037037038,
      "grad_norm": 0.46378329396247864,
      "learning_rate": 7.592592592592594e-06,
      "loss": 1.9459,
      "step": 53
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4593982994556427,
      "learning_rate": 7.546296296296297e-06,
      "loss": 1.999,
      "step": 54
    },
    {
      "epoch": 0.25462962962962965,
      "grad_norm": 0.4580429494380951,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.9357,
      "step": 55
    },
    {
      "epoch": 0.25925925925925924,
      "grad_norm": 0.46820005774497986,
      "learning_rate": 7.453703703703704e-06,
      "loss": 1.9138,
      "step": 56
    },
    {
      "epoch": 0.2638888888888889,
      "grad_norm": 0.45933252573013306,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 1.936,
      "step": 57
    },
    {
      "epoch": 0.26851851851851855,
      "grad_norm": 0.5902265310287476,
      "learning_rate": 7.361111111111112e-06,
      "loss": 1.969,
      "step": 58
    },
    {
      "epoch": 0.27314814814814814,
      "grad_norm": 0.46466007828712463,
      "learning_rate": 7.314814814814816e-06,
      "loss": 1.9416,
      "step": 59
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.45180487632751465,
      "learning_rate": 7.268518518518519e-06,
      "loss": 1.944,
      "step": 60
    },
    {
      "epoch": 0.2824074074074074,
      "grad_norm": 0.46131211519241333,
      "learning_rate": 7.222222222222223e-06,
      "loss": 1.9654,
      "step": 61
    },
    {
      "epoch": 0.28703703703703703,
      "grad_norm": 0.45331230759620667,
      "learning_rate": 7.1759259259259266e-06,
      "loss": 1.9138,
      "step": 62
    },
    {
      "epoch": 0.2916666666666667,
      "grad_norm": 0.43383753299713135,
      "learning_rate": 7.129629629629629e-06,
      "loss": 1.882,
      "step": 63
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 0.4365609288215637,
      "learning_rate": 7.083333333333335e-06,
      "loss": 1.8858,
      "step": 64
    },
    {
      "epoch": 0.30092592592592593,
      "grad_norm": 0.4502002000808716,
      "learning_rate": 7.0370370370370375e-06,
      "loss": 1.8772,
      "step": 65
    },
    {
      "epoch": 0.3055555555555556,
      "grad_norm": 0.465324729681015,
      "learning_rate": 6.990740740740741e-06,
      "loss": 1.9398,
      "step": 66
    },
    {
      "epoch": 0.3101851851851852,
      "grad_norm": 0.4249098598957062,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.8745,
      "step": 67
    },
    {
      "epoch": 0.3148148148148148,
      "grad_norm": 0.436286062002182,
      "learning_rate": 6.898148148148148e-06,
      "loss": 1.9138,
      "step": 68
    },
    {
      "epoch": 0.3194444444444444,
      "grad_norm": 0.43040260672569275,
      "learning_rate": 6.851851851851853e-06,
      "loss": 1.8652,
      "step": 69
    },
    {
      "epoch": 0.32407407407407407,
      "grad_norm": 0.44443291425704956,
      "learning_rate": 6.8055555555555566e-06,
      "loss": 1.883,
      "step": 70
    },
    {
      "epoch": 0.3287037037037037,
      "grad_norm": 0.44603466987609863,
      "learning_rate": 6.75925925925926e-06,
      "loss": 1.8576,
      "step": 71
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.42672744393348694,
      "learning_rate": 6.712962962962963e-06,
      "loss": 1.8487,
      "step": 72
    },
    {
      "epoch": 0.33796296296296297,
      "grad_norm": 0.42773088812828064,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.8638,
      "step": 73
    },
    {
      "epoch": 0.3425925925925926,
      "grad_norm": 0.43208903074264526,
      "learning_rate": 6.620370370370371e-06,
      "loss": 1.8965,
      "step": 74
    },
    {
      "epoch": 0.3472222222222222,
      "grad_norm": 0.454003244638443,
      "learning_rate": 6.574074074074075e-06,
      "loss": 1.9215,
      "step": 75
    },
    {
      "epoch": 0.35185185185185186,
      "grad_norm": 0.43666955828666687,
      "learning_rate": 6.5277777777777784e-06,
      "loss": 1.8701,
      "step": 76
    },
    {
      "epoch": 0.35648148148148145,
      "grad_norm": 0.4259486198425293,
      "learning_rate": 6.481481481481482e-06,
      "loss": 1.8115,
      "step": 77
    },
    {
      "epoch": 0.3611111111111111,
      "grad_norm": 0.421562135219574,
      "learning_rate": 6.435185185185186e-06,
      "loss": 1.8425,
      "step": 78
    },
    {
      "epoch": 0.36574074074074076,
      "grad_norm": 0.4166297912597656,
      "learning_rate": 6.3888888888888885e-06,
      "loss": 1.8147,
      "step": 79
    },
    {
      "epoch": 0.37037037037037035,
      "grad_norm": 0.44431668519973755,
      "learning_rate": 6.342592592592594e-06,
      "loss": 1.8654,
      "step": 80
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.4198393225669861,
      "learning_rate": 6.296296296296297e-06,
      "loss": 1.884,
      "step": 81
    },
    {
      "epoch": 0.37962962962962965,
      "grad_norm": 0.42111140489578247,
      "learning_rate": 6.25e-06,
      "loss": 1.8586,
      "step": 82
    },
    {
      "epoch": 0.38425925925925924,
      "grad_norm": 0.4411139190196991,
      "learning_rate": 6.203703703703704e-06,
      "loss": 1.7706,
      "step": 83
    },
    {
      "epoch": 0.3888888888888889,
      "grad_norm": 0.41565534472465515,
      "learning_rate": 6.157407407407408e-06,
      "loss": 1.8398,
      "step": 84
    },
    {
      "epoch": 0.39351851851851855,
      "grad_norm": 0.43128618597984314,
      "learning_rate": 6.111111111111112e-06,
      "loss": 1.8371,
      "step": 85
    },
    {
      "epoch": 0.39814814814814814,
      "grad_norm": 0.4438124895095825,
      "learning_rate": 6.064814814814816e-06,
      "loss": 1.868,
      "step": 86
    },
    {
      "epoch": 0.4027777777777778,
      "grad_norm": 0.42422446608543396,
      "learning_rate": 6.018518518518519e-06,
      "loss": 1.834,
      "step": 87
    },
    {
      "epoch": 0.4074074074074074,
      "grad_norm": 0.41801929473876953,
      "learning_rate": 5.972222222222222e-06,
      "loss": 1.8652,
      "step": 88
    },
    {
      "epoch": 0.41203703703703703,
      "grad_norm": 0.42849692702293396,
      "learning_rate": 5.925925925925926e-06,
      "loss": 1.8603,
      "step": 89
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.42593345046043396,
      "learning_rate": 5.8796296296296295e-06,
      "loss": 1.8334,
      "step": 90
    },
    {
      "epoch": 0.4212962962962963,
      "grad_norm": 0.42619192600250244,
      "learning_rate": 5.833333333333334e-06,
      "loss": 1.7856,
      "step": 91
    },
    {
      "epoch": 0.42592592592592593,
      "grad_norm": 0.4271165430545807,
      "learning_rate": 5.787037037037038e-06,
      "loss": 1.7967,
      "step": 92
    },
    {
      "epoch": 0.4305555555555556,
      "grad_norm": 0.44702404737472534,
      "learning_rate": 5.740740740740741e-06,
      "loss": 1.8364,
      "step": 93
    },
    {
      "epoch": 0.4351851851851852,
      "grad_norm": 0.4314418137073517,
      "learning_rate": 5.694444444444445e-06,
      "loss": 1.7939,
      "step": 94
    },
    {
      "epoch": 0.4398148148148148,
      "grad_norm": 0.4346453547477722,
      "learning_rate": 5.6481481481481485e-06,
      "loss": 1.7767,
      "step": 95
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 0.4198530614376068,
      "learning_rate": 5.601851851851853e-06,
      "loss": 1.7897,
      "step": 96
    },
    {
      "epoch": 0.44907407407407407,
      "grad_norm": 0.449927419424057,
      "learning_rate": 5.555555555555557e-06,
      "loss": 1.798,
      "step": 97
    },
    {
      "epoch": 0.4537037037037037,
      "grad_norm": 0.43222421407699585,
      "learning_rate": 5.5092592592592595e-06,
      "loss": 1.7874,
      "step": 98
    },
    {
      "epoch": 0.4583333333333333,
      "grad_norm": 0.43431252241134644,
      "learning_rate": 5.462962962962963e-06,
      "loss": 1.78,
      "step": 99
    },
    {
      "epoch": 0.46296296296296297,
      "grad_norm": 0.45050230622291565,
      "learning_rate": 5.416666666666667e-06,
      "loss": 1.768,
      "step": 100
    },
    {
      "epoch": 0.4675925925925926,
      "grad_norm": 0.42318132519721985,
      "learning_rate": 5.370370370370371e-06,
      "loss": 1.7958,
      "step": 101
    },
    {
      "epoch": 0.4722222222222222,
      "grad_norm": 0.4494766592979431,
      "learning_rate": 5.324074074074075e-06,
      "loss": 1.7956,
      "step": 102
    },
    {
      "epoch": 0.47685185185185186,
      "grad_norm": 0.43541356921195984,
      "learning_rate": 5.2777777777777785e-06,
      "loss": 1.7959,
      "step": 103
    },
    {
      "epoch": 0.48148148148148145,
      "grad_norm": 0.4341510832309723,
      "learning_rate": 5.231481481481482e-06,
      "loss": 1.8226,
      "step": 104
    },
    {
      "epoch": 0.4861111111111111,
      "grad_norm": 0.43825647234916687,
      "learning_rate": 5.185185185185185e-06,
      "loss": 1.8148,
      "step": 105
    },
    {
      "epoch": 0.49074074074074076,
      "grad_norm": 0.42040881514549255,
      "learning_rate": 5.138888888888889e-06,
      "loss": 1.7834,
      "step": 106
    },
    {
      "epoch": 0.49537037037037035,
      "grad_norm": 0.4249044954776764,
      "learning_rate": 5.092592592592593e-06,
      "loss": 1.7924,
      "step": 107
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.4380439817905426,
      "learning_rate": 5.046296296296297e-06,
      "loss": 1.7732,
      "step": 108
    },
    {
      "epoch": 0.5046296296296297,
      "grad_norm": 0.42866915464401245,
      "learning_rate": 5e-06,
      "loss": 1.7667,
      "step": 109
    },
    {
      "epoch": 0.5092592592592593,
      "grad_norm": 0.46283015608787537,
      "learning_rate": 4.953703703703704e-06,
      "loss": 1.8107,
      "step": 110
    },
    {
      "epoch": 0.5138888888888888,
      "grad_norm": 0.44372284412384033,
      "learning_rate": 4.907407407407408e-06,
      "loss": 1.7735,
      "step": 111
    },
    {
      "epoch": 0.5185185185185185,
      "grad_norm": 0.4327738285064697,
      "learning_rate": 4.861111111111111e-06,
      "loss": 1.7344,
      "step": 112
    },
    {
      "epoch": 0.5231481481481481,
      "grad_norm": 0.4242445230484009,
      "learning_rate": 4.814814814814815e-06,
      "loss": 1.7834,
      "step": 113
    },
    {
      "epoch": 0.5277777777777778,
      "grad_norm": 0.4237624406814575,
      "learning_rate": 4.768518518518519e-06,
      "loss": 1.7874,
      "step": 114
    },
    {
      "epoch": 0.5324074074074074,
      "grad_norm": 0.44031742215156555,
      "learning_rate": 4.722222222222222e-06,
      "loss": 1.754,
      "step": 115
    },
    {
      "epoch": 0.5370370370370371,
      "grad_norm": 0.44006913900375366,
      "learning_rate": 4.675925925925927e-06,
      "loss": 1.8411,
      "step": 116
    },
    {
      "epoch": 0.5416666666666666,
      "grad_norm": 0.42018765211105347,
      "learning_rate": 4.62962962962963e-06,
      "loss": 1.7611,
      "step": 117
    },
    {
      "epoch": 0.5462962962962963,
      "grad_norm": 0.4189184308052063,
      "learning_rate": 4.583333333333333e-06,
      "loss": 1.7608,
      "step": 118
    },
    {
      "epoch": 0.5509259259259259,
      "grad_norm": 0.4487444758415222,
      "learning_rate": 4.537037037037038e-06,
      "loss": 1.7308,
      "step": 119
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.43615803122520447,
      "learning_rate": 4.490740740740741e-06,
      "loss": 1.7991,
      "step": 120
    },
    {
      "epoch": 0.5601851851851852,
      "grad_norm": 0.44524088501930237,
      "learning_rate": 4.444444444444444e-06,
      "loss": 1.6982,
      "step": 121
    },
    {
      "epoch": 0.5648148148148148,
      "grad_norm": 0.41494986414909363,
      "learning_rate": 4.398148148148149e-06,
      "loss": 1.7386,
      "step": 122
    },
    {
      "epoch": 0.5694444444444444,
      "grad_norm": 0.42478981614112854,
      "learning_rate": 4.351851851851852e-06,
      "loss": 1.7154,
      "step": 123
    },
    {
      "epoch": 0.5740740740740741,
      "grad_norm": 0.4177737832069397,
      "learning_rate": 4.305555555555556e-06,
      "loss": 1.7567,
      "step": 124
    },
    {
      "epoch": 0.5787037037037037,
      "grad_norm": 0.41471073031425476,
      "learning_rate": 4.2592592592592596e-06,
      "loss": 1.757,
      "step": 125
    },
    {
      "epoch": 0.5833333333333334,
      "grad_norm": 0.3962550461292267,
      "learning_rate": 4.212962962962963e-06,
      "loss": 1.7559,
      "step": 126
    },
    {
      "epoch": 0.5879629629629629,
      "grad_norm": 0.4219653904438019,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.7109,
      "step": 127
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 0.4559638798236847,
      "learning_rate": 4.1203703703703705e-06,
      "loss": 1.7746,
      "step": 128
    },
    {
      "epoch": 0.5972222222222222,
      "grad_norm": 0.46450746059417725,
      "learning_rate": 4.074074074074074e-06,
      "loss": 1.7442,
      "step": 129
    },
    {
      "epoch": 0.6018518518518519,
      "grad_norm": 0.4612331688404083,
      "learning_rate": 4.027777777777779e-06,
      "loss": 1.7993,
      "step": 130
    },
    {
      "epoch": 0.6064814814814815,
      "grad_norm": 0.40595942735671997,
      "learning_rate": 3.9814814814814814e-06,
      "loss": 1.7207,
      "step": 131
    },
    {
      "epoch": 0.6111111111111112,
      "grad_norm": 0.444697767496109,
      "learning_rate": 3.935185185185186e-06,
      "loss": 1.7626,
      "step": 132
    },
    {
      "epoch": 0.6157407407407407,
      "grad_norm": 0.4202960133552551,
      "learning_rate": 3.88888888888889e-06,
      "loss": 1.7015,
      "step": 133
    },
    {
      "epoch": 0.6203703703703703,
      "grad_norm": 0.425169974565506,
      "learning_rate": 3.842592592592592e-06,
      "loss": 1.7348,
      "step": 134
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.4239872694015503,
      "learning_rate": 3.796296296296297e-06,
      "loss": 1.7044,
      "step": 135
    },
    {
      "epoch": 0.6296296296296297,
      "grad_norm": 0.4289877116680145,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 1.7418,
      "step": 136
    },
    {
      "epoch": 0.6342592592592593,
      "grad_norm": 0.4335390627384186,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 1.7311,
      "step": 137
    },
    {
      "epoch": 0.6388888888888888,
      "grad_norm": 0.44100385904312134,
      "learning_rate": 3.657407407407408e-06,
      "loss": 1.6752,
      "step": 138
    },
    {
      "epoch": 0.6435185185185185,
      "grad_norm": 0.4602786600589752,
      "learning_rate": 3.6111111111111115e-06,
      "loss": 1.7786,
      "step": 139
    },
    {
      "epoch": 0.6481481481481481,
      "grad_norm": 0.43911105394363403,
      "learning_rate": 3.5648148148148147e-06,
      "loss": 1.7343,
      "step": 140
    },
    {
      "epoch": 0.6527777777777778,
      "grad_norm": 0.4818420708179474,
      "learning_rate": 3.5185185185185187e-06,
      "loss": 1.6351,
      "step": 141
    },
    {
      "epoch": 0.6574074074074074,
      "grad_norm": 0.4308430552482605,
      "learning_rate": 3.4722222222222224e-06,
      "loss": 1.7021,
      "step": 142
    },
    {
      "epoch": 0.6620370370370371,
      "grad_norm": 0.4618017077445984,
      "learning_rate": 3.4259259259259265e-06,
      "loss": 1.7931,
      "step": 143
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.418113648891449,
      "learning_rate": 3.37962962962963e-06,
      "loss": 1.6651,
      "step": 144
    },
    {
      "epoch": 0.6712962962962963,
      "grad_norm": 0.43931904435157776,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.7363,
      "step": 145
    },
    {
      "epoch": 0.6759259259259259,
      "grad_norm": 0.42020654678344727,
      "learning_rate": 3.2870370370370374e-06,
      "loss": 1.7501,
      "step": 146
    },
    {
      "epoch": 0.6805555555555556,
      "grad_norm": 0.41723671555519104,
      "learning_rate": 3.240740740740741e-06,
      "loss": 1.7354,
      "step": 147
    },
    {
      "epoch": 0.6851851851851852,
      "grad_norm": 0.44364774227142334,
      "learning_rate": 3.1944444444444443e-06,
      "loss": 1.6479,
      "step": 148
    },
    {
      "epoch": 0.6898148148148148,
      "grad_norm": 0.41947969794273376,
      "learning_rate": 3.1481481481481483e-06,
      "loss": 1.7081,
      "step": 149
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.42689067125320435,
      "learning_rate": 3.101851851851852e-06,
      "loss": 1.6716,
      "step": 150
    },
    {
      "epoch": 0.6990740740740741,
      "grad_norm": 0.4190499484539032,
      "learning_rate": 3.055555555555556e-06,
      "loss": 1.6715,
      "step": 151
    },
    {
      "epoch": 0.7037037037037037,
      "grad_norm": 0.4444749355316162,
      "learning_rate": 3.0092592592592597e-06,
      "loss": 1.7352,
      "step": 152
    },
    {
      "epoch": 0.7083333333333334,
      "grad_norm": 0.4346008002758026,
      "learning_rate": 2.962962962962963e-06,
      "loss": 1.6889,
      "step": 153
    },
    {
      "epoch": 0.7129629629629629,
      "grad_norm": 0.4162415862083435,
      "learning_rate": 2.916666666666667e-06,
      "loss": 1.6959,
      "step": 154
    },
    {
      "epoch": 0.7175925925925926,
      "grad_norm": 0.41898995637893677,
      "learning_rate": 2.8703703703703706e-06,
      "loss": 1.6694,
      "step": 155
    },
    {
      "epoch": 0.7222222222222222,
      "grad_norm": 0.41438576579093933,
      "learning_rate": 2.8240740740740743e-06,
      "loss": 1.6653,
      "step": 156
    },
    {
      "epoch": 0.7268518518518519,
      "grad_norm": 0.5035732388496399,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 1.7819,
      "step": 157
    },
    {
      "epoch": 0.7314814814814815,
      "grad_norm": 0.4908410310745239,
      "learning_rate": 2.7314814814814816e-06,
      "loss": 1.7884,
      "step": 158
    },
    {
      "epoch": 0.7361111111111112,
      "grad_norm": 0.4252499043941498,
      "learning_rate": 2.6851851851851856e-06,
      "loss": 1.6733,
      "step": 159
    },
    {
      "epoch": 0.7407407407407407,
      "grad_norm": 0.42570921778678894,
      "learning_rate": 2.6388888888888893e-06,
      "loss": 1.706,
      "step": 160
    },
    {
      "epoch": 0.7453703703703703,
      "grad_norm": 0.4818575978279114,
      "learning_rate": 2.5925925925925925e-06,
      "loss": 1.6386,
      "step": 161
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.4613746106624603,
      "learning_rate": 2.5462962962962966e-06,
      "loss": 1.7099,
      "step": 162
    },
    {
      "epoch": 0.7546296296296297,
      "grad_norm": 0.4468544125556946,
      "learning_rate": 2.5e-06,
      "loss": 1.6778,
      "step": 163
    },
    {
      "epoch": 0.7592592592592593,
      "grad_norm": 0.43980488181114197,
      "learning_rate": 2.453703703703704e-06,
      "loss": 1.6743,
      "step": 164
    },
    {
      "epoch": 0.7638888888888888,
      "grad_norm": 0.4213511645793915,
      "learning_rate": 2.4074074074074075e-06,
      "loss": 1.6769,
      "step": 165
    },
    {
      "epoch": 0.7685185185185185,
      "grad_norm": 0.43999916315078735,
      "learning_rate": 2.361111111111111e-06,
      "loss": 1.6725,
      "step": 166
    },
    {
      "epoch": 0.7731481481481481,
      "grad_norm": 0.44871917366981506,
      "learning_rate": 2.314814814814815e-06,
      "loss": 1.713,
      "step": 167
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 0.4404929578304291,
      "learning_rate": 2.268518518518519e-06,
      "loss": 1.6882,
      "step": 168
    },
    {
      "epoch": 0.7824074074074074,
      "grad_norm": 0.4368396997451782,
      "learning_rate": 2.222222222222222e-06,
      "loss": 1.6467,
      "step": 169
    },
    {
      "epoch": 0.7870370370370371,
      "grad_norm": 0.4457816183567047,
      "learning_rate": 2.175925925925926e-06,
      "loss": 1.6818,
      "step": 170
    },
    {
      "epoch": 0.7916666666666666,
      "grad_norm": 0.43753132224082947,
      "learning_rate": 2.1296296296296298e-06,
      "loss": 1.7128,
      "step": 171
    },
    {
      "epoch": 0.7962962962962963,
      "grad_norm": 0.44547879695892334,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.694,
      "step": 172
    },
    {
      "epoch": 0.8009259259259259,
      "grad_norm": 0.41959348320961,
      "learning_rate": 2.037037037037037e-06,
      "loss": 1.6486,
      "step": 173
    },
    {
      "epoch": 0.8055555555555556,
      "grad_norm": 0.46483269333839417,
      "learning_rate": 1.9907407407407407e-06,
      "loss": 1.7278,
      "step": 174
    },
    {
      "epoch": 0.8101851851851852,
      "grad_norm": 0.4479513168334961,
      "learning_rate": 1.944444444444445e-06,
      "loss": 1.7069,
      "step": 175
    },
    {
      "epoch": 0.8148148148148148,
      "grad_norm": 0.4265024960041046,
      "learning_rate": 1.8981481481481484e-06,
      "loss": 1.718,
      "step": 176
    },
    {
      "epoch": 0.8194444444444444,
      "grad_norm": 0.4072514474391937,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 1.6924,
      "step": 177
    },
    {
      "epoch": 0.8240740740740741,
      "grad_norm": 0.43383556604385376,
      "learning_rate": 1.8055555555555557e-06,
      "loss": 1.6466,
      "step": 178
    },
    {
      "epoch": 0.8287037037037037,
      "grad_norm": 0.41848164796829224,
      "learning_rate": 1.7592592592592594e-06,
      "loss": 1.7023,
      "step": 179
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.45098960399627686,
      "learning_rate": 1.7129629629629632e-06,
      "loss": 1.7352,
      "step": 180
    },
    {
      "epoch": 0.8379629629629629,
      "grad_norm": 0.4191969335079193,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 1.6721,
      "step": 181
    },
    {
      "epoch": 0.8425925925925926,
      "grad_norm": 0.4636867344379425,
      "learning_rate": 1.6203703703703705e-06,
      "loss": 1.6469,
      "step": 182
    },
    {
      "epoch": 0.8472222222222222,
      "grad_norm": 0.42679235339164734,
      "learning_rate": 1.5740740740740742e-06,
      "loss": 1.7365,
      "step": 183
    },
    {
      "epoch": 0.8518518518518519,
      "grad_norm": 0.416758269071579,
      "learning_rate": 1.527777777777778e-06,
      "loss": 1.6835,
      "step": 184
    },
    {
      "epoch": 0.8564814814814815,
      "grad_norm": 0.4596608281135559,
      "learning_rate": 1.4814814814814815e-06,
      "loss": 1.705,
      "step": 185
    },
    {
      "epoch": 0.8611111111111112,
      "grad_norm": 0.4271976947784424,
      "learning_rate": 1.4351851851851853e-06,
      "loss": 1.6272,
      "step": 186
    },
    {
      "epoch": 0.8657407407407407,
      "grad_norm": 0.42482131719589233,
      "learning_rate": 1.3888888888888892e-06,
      "loss": 1.6447,
      "step": 187
    },
    {
      "epoch": 0.8703703703703703,
      "grad_norm": 0.40397384762763977,
      "learning_rate": 1.3425925925925928e-06,
      "loss": 1.6812,
      "step": 188
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.6595266461372375,
      "learning_rate": 1.2962962962962962e-06,
      "loss": 1.7041,
      "step": 189
    },
    {
      "epoch": 0.8796296296296297,
      "grad_norm": 0.49033576250076294,
      "learning_rate": 1.25e-06,
      "loss": 1.7392,
      "step": 190
    },
    {
      "epoch": 0.8842592592592593,
      "grad_norm": 0.4406794011592865,
      "learning_rate": 1.2037037037037037e-06,
      "loss": 1.7136,
      "step": 191
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.4479171335697174,
      "learning_rate": 1.1574074074074076e-06,
      "loss": 1.7089,
      "step": 192
    },
    {
      "epoch": 0.8935185185185185,
      "grad_norm": 0.4189581871032715,
      "learning_rate": 1.111111111111111e-06,
      "loss": 1.6785,
      "step": 193
    },
    {
      "epoch": 0.8981481481481481,
      "grad_norm": 0.4524936079978943,
      "learning_rate": 1.0648148148148149e-06,
      "loss": 1.7061,
      "step": 194
    },
    {
      "epoch": 0.9027777777777778,
      "grad_norm": 0.42082199454307556,
      "learning_rate": 1.0185185185185185e-06,
      "loss": 1.691,
      "step": 195
    },
    {
      "epoch": 0.9074074074074074,
      "grad_norm": 0.4415839910507202,
      "learning_rate": 9.722222222222224e-07,
      "loss": 1.7295,
      "step": 196
    },
    {
      "epoch": 0.9120370370370371,
      "grad_norm": 0.41390758752822876,
      "learning_rate": 9.259259259259259e-07,
      "loss": 1.7035,
      "step": 197
    },
    {
      "epoch": 0.9166666666666666,
      "grad_norm": 0.42637038230895996,
      "learning_rate": 8.796296296296297e-07,
      "loss": 1.6505,
      "step": 198
    },
    {
      "epoch": 0.9212962962962963,
      "grad_norm": 0.4274970591068268,
      "learning_rate": 8.333333333333333e-07,
      "loss": 1.6936,
      "step": 199
    },
    {
      "epoch": 0.9259259259259259,
      "grad_norm": 0.42553502321243286,
      "learning_rate": 7.870370370370371e-07,
      "loss": 1.6519,
      "step": 200
    },
    {
      "epoch": 0.9305555555555556,
      "grad_norm": 0.44404858350753784,
      "learning_rate": 7.407407407407407e-07,
      "loss": 1.6605,
      "step": 201
    },
    {
      "epoch": 0.9351851851851852,
      "grad_norm": 0.4228724539279938,
      "learning_rate": 6.944444444444446e-07,
      "loss": 1.6861,
      "step": 202
    },
    {
      "epoch": 0.9398148148148148,
      "grad_norm": 0.4268285036087036,
      "learning_rate": 6.481481481481481e-07,
      "loss": 1.6696,
      "step": 203
    },
    {
      "epoch": 0.9444444444444444,
      "grad_norm": 0.6097678542137146,
      "learning_rate": 6.018518518518519e-07,
      "loss": 1.6768,
      "step": 204
    },
    {
      "epoch": 0.9490740740740741,
      "grad_norm": 0.434126079082489,
      "learning_rate": 5.555555555555555e-07,
      "loss": 1.6355,
      "step": 205
    },
    {
      "epoch": 0.9537037037037037,
      "grad_norm": 0.4083269238471985,
      "learning_rate": 5.092592592592593e-07,
      "loss": 1.6595,
      "step": 206
    },
    {
      "epoch": 0.9583333333333334,
      "grad_norm": 0.4251357316970825,
      "learning_rate": 4.6296296296296297e-07,
      "loss": 1.7208,
      "step": 207
    },
    {
      "epoch": 0.9629629629629629,
      "grad_norm": 0.4255342185497284,
      "learning_rate": 4.1666666666666667e-07,
      "loss": 1.7005,
      "step": 208
    },
    {
      "epoch": 0.9675925925925926,
      "grad_norm": 0.4373617172241211,
      "learning_rate": 3.7037037037037036e-07,
      "loss": 1.7133,
      "step": 209
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.4113195836544037,
      "learning_rate": 3.2407407407407406e-07,
      "loss": 1.6353,
      "step": 210
    },
    {
      "epoch": 0.9768518518518519,
      "grad_norm": 0.4072490930557251,
      "learning_rate": 2.7777777777777776e-07,
      "loss": 1.6936,
      "step": 211
    },
    {
      "epoch": 0.9814814814814815,
      "grad_norm": 0.44550320506095886,
      "learning_rate": 2.3148148148148148e-07,
      "loss": 1.6657,
      "step": 212
    },
    {
      "epoch": 0.9861111111111112,
      "grad_norm": 0.4239332675933838,
      "learning_rate": 1.8518518518518518e-07,
      "loss": 1.6619,
      "step": 213
    },
    {
      "epoch": 0.9907407407407407,
      "grad_norm": 0.4457761347293854,
      "learning_rate": 1.3888888888888888e-07,
      "loss": 1.6724,
      "step": 214
    },
    {
      "epoch": 0.9953703703703703,
      "grad_norm": 0.44246068596839905,
      "learning_rate": 9.259259259259259e-08,
      "loss": 1.6661,
      "step": 215
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.43305960297584534,
      "learning_rate": 4.6296296296296295e-08,
      "loss": 1.7103,
      "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": 1.0138967789797376e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}