{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.23,
  "eval_steps": 500,
  "global_step": 23000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001,
      "grad_norm": 12.033066749572754,
      "learning_rate": 9.998000000000002e-06,
      "loss": 5.0328,
      "step": 100
    },
    {
      "epoch": 0.002,
      "grad_norm": 12.140076637268066,
      "learning_rate": 9.996e-06,
      "loss": 4.8188,
      "step": 200
    },
    {
      "epoch": 0.003,
      "grad_norm": 11.839982032775879,
      "learning_rate": 9.994000000000001e-06,
      "loss": 4.6021,
      "step": 300
    },
    {
      "epoch": 0.004,
      "grad_norm": 13.72381591796875,
      "learning_rate": 9.992e-06,
      "loss": 4.4731,
      "step": 400
    },
    {
      "epoch": 0.005,
      "grad_norm": 11.587430000305176,
      "learning_rate": 9.990000000000001e-06,
      "loss": 4.4251,
      "step": 500
    },
    {
      "epoch": 0.006,
      "grad_norm": 12.022308349609375,
      "learning_rate": 9.988000000000002e-06,
      "loss": 4.4416,
      "step": 600
    },
    {
      "epoch": 0.007,
      "grad_norm": 12.038565635681152,
      "learning_rate": 9.986e-06,
      "loss": 4.3319,
      "step": 700
    },
    {
      "epoch": 0.008,
      "grad_norm": 11.564003944396973,
      "learning_rate": 9.984e-06,
      "loss": 4.3267,
      "step": 800
    },
    {
      "epoch": 0.009,
      "grad_norm": 11.518474578857422,
      "learning_rate": 9.982e-06,
      "loss": 4.2818,
      "step": 900
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.206191062927246,
      "learning_rate": 9.980000000000001e-06,
      "loss": 4.3167,
      "step": 1000
    },
    {
      "epoch": 0.011,
      "grad_norm": 11.940592765808105,
      "learning_rate": 9.978000000000002e-06,
      "loss": 4.3204,
      "step": 1100
    },
    {
      "epoch": 0.012,
      "grad_norm": 12.109590530395508,
      "learning_rate": 9.976e-06,
      "loss": 4.3286,
      "step": 1200
    },
    {
      "epoch": 0.013,
      "grad_norm": 12.218742370605469,
      "learning_rate": 9.974e-06,
      "loss": 4.4106,
      "step": 1300
    },
    {
      "epoch": 0.014,
      "grad_norm": 11.53864860534668,
      "learning_rate": 9.972e-06,
      "loss": 4.3091,
      "step": 1400
    },
    {
      "epoch": 0.015,
      "grad_norm": 13.800053596496582,
      "learning_rate": 9.970000000000001e-06,
      "loss": 4.2455,
      "step": 1500
    },
    {
      "epoch": 0.016,
      "grad_norm": 13.468424797058105,
      "learning_rate": 9.968000000000002e-06,
      "loss": 4.305,
      "step": 1600
    },
    {
      "epoch": 0.017,
      "grad_norm": 11.736213684082031,
      "learning_rate": 9.966e-06,
      "loss": 4.3232,
      "step": 1700
    },
    {
      "epoch": 0.018,
      "grad_norm": 10.893006324768066,
      "learning_rate": 9.964e-06,
      "loss": 4.3139,
      "step": 1800
    },
    {
      "epoch": 0.019,
      "grad_norm": 14.407844543457031,
      "learning_rate": 9.962e-06,
      "loss": 4.2437,
      "step": 1900
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.889874458312988,
      "learning_rate": 9.960000000000001e-06,
      "loss": 4.2543,
      "step": 2000
    },
    {
      "epoch": 0.021,
      "grad_norm": 11.396763801574707,
      "learning_rate": 9.958e-06,
      "loss": 4.1477,
      "step": 2100
    },
    {
      "epoch": 0.022,
      "grad_norm": 13.99250316619873,
      "learning_rate": 9.956000000000001e-06,
      "loss": 4.248,
      "step": 2200
    },
    {
      "epoch": 0.023,
      "grad_norm": 11.850446701049805,
      "learning_rate": 9.954e-06,
      "loss": 4.3014,
      "step": 2300
    },
    {
      "epoch": 0.024,
      "grad_norm": 13.699837684631348,
      "learning_rate": 9.952e-06,
      "loss": 4.1937,
      "step": 2400
    },
    {
      "epoch": 0.025,
      "grad_norm": 11.772696495056152,
      "learning_rate": 9.950000000000001e-06,
      "loss": 4.264,
      "step": 2500
    },
    {
      "epoch": 0.026,
      "grad_norm": 13.460022926330566,
      "learning_rate": 9.948e-06,
      "loss": 4.2278,
      "step": 2600
    },
    {
      "epoch": 0.027,
      "grad_norm": 11.80987548828125,
      "learning_rate": 9.946000000000001e-06,
      "loss": 4.2416,
      "step": 2700
    },
    {
      "epoch": 0.028,
      "grad_norm": 12.391595840454102,
      "learning_rate": 9.944e-06,
      "loss": 4.2183,
      "step": 2800
    },
    {
      "epoch": 0.029,
      "grad_norm": 12.336369514465332,
      "learning_rate": 9.942e-06,
      "loss": 4.225,
      "step": 2900
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.137269020080566,
      "learning_rate": 9.940000000000001e-06,
      "loss": 4.2045,
      "step": 3000
    },
    {
      "epoch": 0.031,
      "grad_norm": 12.397940635681152,
      "learning_rate": 9.938e-06,
      "loss": 4.1523,
      "step": 3100
    },
    {
      "epoch": 0.032,
      "grad_norm": 12.940911293029785,
      "learning_rate": 9.936000000000001e-06,
      "loss": 4.0193,
      "step": 3200
    },
    {
      "epoch": 0.033,
      "grad_norm": 16.68646812438965,
      "learning_rate": 9.934e-06,
      "loss": 4.1939,
      "step": 3300
    },
    {
      "epoch": 0.034,
      "grad_norm": 12.541526794433594,
      "learning_rate": 9.932e-06,
      "loss": 4.0829,
      "step": 3400
    },
    {
      "epoch": 0.035,
      "grad_norm": 11.975446701049805,
      "learning_rate": 9.930000000000001e-06,
      "loss": 4.2043,
      "step": 3500
    },
    {
      "epoch": 0.036,
      "grad_norm": 12.638479232788086,
      "learning_rate": 9.928e-06,
      "loss": 4.1315,
      "step": 3600
    },
    {
      "epoch": 0.037,
      "grad_norm": 13.302968978881836,
      "learning_rate": 9.926000000000001e-06,
      "loss": 4.0406,
      "step": 3700
    },
    {
      "epoch": 0.038,
      "grad_norm": 12.131388664245605,
      "learning_rate": 9.924e-06,
      "loss": 4.0571,
      "step": 3800
    },
    {
      "epoch": 0.039,
      "grad_norm": 13.895309448242188,
      "learning_rate": 9.922000000000001e-06,
      "loss": 4.0597,
      "step": 3900
    },
    {
      "epoch": 0.04,
      "grad_norm": 15.263091087341309,
      "learning_rate": 9.920000000000002e-06,
      "loss": 4.1587,
      "step": 4000
    },
    {
      "epoch": 0.041,
      "grad_norm": 12.314478874206543,
      "learning_rate": 9.918e-06,
      "loss": 4.0695,
      "step": 4100
    },
    {
      "epoch": 0.042,
      "grad_norm": 13.542490005493164,
      "learning_rate": 9.916000000000001e-06,
      "loss": 4.088,
      "step": 4200
    },
    {
      "epoch": 0.043,
      "grad_norm": 14.835192680358887,
      "learning_rate": 9.914e-06,
      "loss": 4.1067,
      "step": 4300
    },
    {
      "epoch": 0.044,
      "grad_norm": 13.269238471984863,
      "learning_rate": 9.912000000000001e-06,
      "loss": 4.0273,
      "step": 4400
    },
    {
      "epoch": 0.045,
      "grad_norm": 12.532042503356934,
      "learning_rate": 9.91e-06,
      "loss": 3.9738,
      "step": 4500
    },
    {
      "epoch": 0.046,
      "grad_norm": 14.506613731384277,
      "learning_rate": 9.908e-06,
      "loss": 4.0215,
      "step": 4600
    },
    {
      "epoch": 0.047,
      "grad_norm": 12.91763973236084,
      "learning_rate": 9.906000000000001e-06,
      "loss": 3.9976,
      "step": 4700
    },
    {
      "epoch": 0.048,
      "grad_norm": 12.20261001586914,
      "learning_rate": 9.904e-06,
      "loss": 3.9172,
      "step": 4800
    },
    {
      "epoch": 0.049,
      "grad_norm": 13.156211853027344,
      "learning_rate": 9.902000000000001e-06,
      "loss": 3.9871,
      "step": 4900
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.59281063079834,
      "learning_rate": 9.9e-06,
      "loss": 3.9724,
      "step": 5000
    },
    {
      "epoch": 0.051,
      "grad_norm": 13.202598571777344,
      "learning_rate": 9.898e-06,
      "loss": 3.8869,
      "step": 5100
    },
    {
      "epoch": 0.052,
      "grad_norm": 16.82631492614746,
      "learning_rate": 9.896000000000001e-06,
      "loss": 4.0354,
      "step": 5200
    },
    {
      "epoch": 0.053,
      "grad_norm": 14.205794334411621,
      "learning_rate": 9.894e-06,
      "loss": 4.0322,
      "step": 5300
    },
    {
      "epoch": 0.054,
      "grad_norm": 14.135138511657715,
      "learning_rate": 9.892000000000001e-06,
      "loss": 4.01,
      "step": 5400
    },
    {
      "epoch": 0.055,
      "grad_norm": 14.365863800048828,
      "learning_rate": 9.89e-06,
      "loss": 3.9914,
      "step": 5500
    },
    {
      "epoch": 0.056,
      "grad_norm": 14.389483451843262,
      "learning_rate": 9.888000000000001e-06,
      "loss": 3.8672,
      "step": 5600
    },
    {
      "epoch": 0.057,
      "grad_norm": 12.534111976623535,
      "learning_rate": 9.886000000000002e-06,
      "loss": 3.8888,
      "step": 5700
    },
    {
      "epoch": 0.058,
      "grad_norm": 12.352408409118652,
      "learning_rate": 9.884e-06,
      "loss": 3.9466,
      "step": 5800
    },
    {
      "epoch": 0.059,
      "grad_norm": 11.810559272766113,
      "learning_rate": 9.882000000000001e-06,
      "loss": 3.8322,
      "step": 5900
    },
    {
      "epoch": 0.06,
      "grad_norm": 15.00880241394043,
      "learning_rate": 9.88e-06,
      "loss": 3.9209,
      "step": 6000
    },
    {
      "epoch": 0.061,
      "grad_norm": 13.07725715637207,
      "learning_rate": 9.878000000000001e-06,
      "loss": 3.935,
      "step": 6100
    },
    {
      "epoch": 0.062,
      "grad_norm": 11.981454849243164,
      "learning_rate": 9.876000000000002e-06,
      "loss": 3.808,
      "step": 6200
    },
    {
      "epoch": 0.063,
      "grad_norm": 14.732353210449219,
      "learning_rate": 9.874e-06,
      "loss": 3.9113,
      "step": 6300
    },
    {
      "epoch": 0.064,
      "grad_norm": 14.65374755859375,
      "learning_rate": 9.872e-06,
      "loss": 3.9368,
      "step": 6400
    },
    {
      "epoch": 0.065,
      "grad_norm": 19.678359985351562,
      "learning_rate": 9.87e-06,
      "loss": 3.8806,
      "step": 6500
    },
    {
      "epoch": 0.066,
      "grad_norm": 13.10519790649414,
      "learning_rate": 9.868000000000001e-06,
      "loss": 3.7696,
      "step": 6600
    },
    {
      "epoch": 0.067,
      "grad_norm": 13.150310516357422,
      "learning_rate": 9.866000000000002e-06,
      "loss": 3.7716,
      "step": 6700
    },
    {
      "epoch": 0.068,
      "grad_norm": 12.890090942382812,
      "learning_rate": 9.864e-06,
      "loss": 3.8475,
      "step": 6800
    },
    {
      "epoch": 0.069,
      "grad_norm": 14.92232894897461,
      "learning_rate": 9.862e-06,
      "loss": 3.9018,
      "step": 6900
    },
    {
      "epoch": 0.07,
      "grad_norm": 17.389636993408203,
      "learning_rate": 9.86e-06,
      "loss": 3.8824,
      "step": 7000
    },
    {
      "epoch": 0.071,
      "grad_norm": 13.9534330368042,
      "learning_rate": 9.858000000000001e-06,
      "loss": 3.8621,
      "step": 7100
    },
    {
      "epoch": 0.072,
      "grad_norm": 13.286906242370605,
      "learning_rate": 9.856000000000002e-06,
      "loss": 3.9293,
      "step": 7200
    },
    {
      "epoch": 0.073,
      "grad_norm": 13.590744972229004,
      "learning_rate": 9.854000000000001e-06,
      "loss": 3.8883,
      "step": 7300
    },
    {
      "epoch": 0.074,
      "grad_norm": 10.841891288757324,
      "learning_rate": 9.852e-06,
      "loss": 3.8022,
      "step": 7400
    },
    {
      "epoch": 0.075,
      "grad_norm": 14.516188621520996,
      "learning_rate": 9.85e-06,
      "loss": 3.713,
      "step": 7500
    },
    {
      "epoch": 0.076,
      "grad_norm": 16.878511428833008,
      "learning_rate": 9.848000000000001e-06,
      "loss": 3.8758,
      "step": 7600
    },
    {
      "epoch": 0.077,
      "grad_norm": 16.681041717529297,
      "learning_rate": 9.846000000000002e-06,
      "loss": 3.7602,
      "step": 7700
    },
    {
      "epoch": 0.078,
      "grad_norm": 14.792035102844238,
      "learning_rate": 9.844000000000001e-06,
      "loss": 3.7145,
      "step": 7800
    },
    {
      "epoch": 0.079,
      "grad_norm": 15.74644660949707,
      "learning_rate": 9.842e-06,
      "loss": 3.7551,
      "step": 7900
    },
    {
      "epoch": 0.08,
      "grad_norm": 16.060367584228516,
      "learning_rate": 9.84e-06,
      "loss": 3.7701,
      "step": 8000
    },
    {
      "epoch": 0.081,
      "grad_norm": 13.487128257751465,
      "learning_rate": 9.838000000000001e-06,
      "loss": 3.6881,
      "step": 8100
    },
    {
      "epoch": 0.082,
      "grad_norm": 15.940110206604004,
      "learning_rate": 9.836e-06,
      "loss": 3.9418,
      "step": 8200
    },
    {
      "epoch": 0.083,
      "grad_norm": 13.603132247924805,
      "learning_rate": 9.834000000000001e-06,
      "loss": 3.7518,
      "step": 8300
    },
    {
      "epoch": 0.084,
      "grad_norm": 17.045244216918945,
      "learning_rate": 9.832e-06,
      "loss": 3.7167,
      "step": 8400
    },
    {
      "epoch": 0.085,
      "grad_norm": 12.39263916015625,
      "learning_rate": 9.83e-06,
      "loss": 3.6829,
      "step": 8500
    },
    {
      "epoch": 0.086,
      "grad_norm": 12.677363395690918,
      "learning_rate": 9.828000000000001e-06,
      "loss": 3.7458,
      "step": 8600
    },
    {
      "epoch": 0.087,
      "grad_norm": 15.024678230285645,
      "learning_rate": 9.826e-06,
      "loss": 3.7334,
      "step": 8700
    },
    {
      "epoch": 0.088,
      "grad_norm": 17.41254997253418,
      "learning_rate": 9.824000000000001e-06,
      "loss": 3.7704,
      "step": 8800
    },
    {
      "epoch": 0.089,
      "grad_norm": 19.782014846801758,
      "learning_rate": 9.822e-06,
      "loss": 3.6834,
      "step": 8900
    },
    {
      "epoch": 0.09,
      "grad_norm": 16.899019241333008,
      "learning_rate": 9.820000000000001e-06,
      "loss": 3.7304,
      "step": 9000
    },
    {
      "epoch": 0.091,
      "grad_norm": 14.481075286865234,
      "learning_rate": 9.818000000000002e-06,
      "loss": 3.7307,
      "step": 9100
    },
    {
      "epoch": 0.092,
      "grad_norm": 18.121864318847656,
      "learning_rate": 9.816e-06,
      "loss": 3.6793,
      "step": 9200
    },
    {
      "epoch": 0.093,
      "grad_norm": 15.916873931884766,
      "learning_rate": 9.814000000000001e-06,
      "loss": 3.7664,
      "step": 9300
    },
    {
      "epoch": 0.094,
      "grad_norm": 18.305234909057617,
      "learning_rate": 9.812e-06,
      "loss": 3.6887,
      "step": 9400
    },
    {
      "epoch": 0.095,
      "grad_norm": 18.262725830078125,
      "learning_rate": 9.810000000000001e-06,
      "loss": 3.6865,
      "step": 9500
    },
    {
      "epoch": 0.096,
      "grad_norm": 21.94981575012207,
      "learning_rate": 9.808000000000002e-06,
      "loss": 3.5916,
      "step": 9600
    },
    {
      "epoch": 0.097,
      "grad_norm": 15.031508445739746,
      "learning_rate": 9.806e-06,
      "loss": 3.5732,
      "step": 9700
    },
    {
      "epoch": 0.098,
      "grad_norm": 13.64002799987793,
      "learning_rate": 9.804000000000001e-06,
      "loss": 3.6104,
      "step": 9800
    },
    {
      "epoch": 0.099,
      "grad_norm": 22.877960205078125,
      "learning_rate": 9.802e-06,
      "loss": 3.622,
      "step": 9900
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.404180526733398,
      "learning_rate": 9.800000000000001e-06,
      "loss": 3.6693,
      "step": 10000
    },
    {
      "epoch": 0.101,
      "grad_norm": 14.348539352416992,
      "learning_rate": 9.798e-06,
      "loss": 3.4753,
      "step": 10100
    },
    {
      "epoch": 0.102,
      "grad_norm": 15.996590614318848,
      "learning_rate": 9.796e-06,
      "loss": 3.5059,
      "step": 10200
    },
    {
      "epoch": 0.103,
      "grad_norm": 16.6004638671875,
      "learning_rate": 9.794000000000001e-06,
      "loss": 3.7168,
      "step": 10300
    },
    {
      "epoch": 0.104,
      "grad_norm": 22.660940170288086,
      "learning_rate": 9.792e-06,
      "loss": 3.5521,
      "step": 10400
    },
    {
      "epoch": 0.105,
      "grad_norm": 16.634521484375,
      "learning_rate": 9.790000000000001e-06,
      "loss": 3.5894,
      "step": 10500
    },
    {
      "epoch": 0.106,
      "grad_norm": 17.89203643798828,
      "learning_rate": 9.788e-06,
      "loss": 3.5524,
      "step": 10600
    },
    {
      "epoch": 0.107,
      "grad_norm": 17.027833938598633,
      "learning_rate": 9.786e-06,
      "loss": 3.5953,
      "step": 10700
    },
    {
      "epoch": 0.108,
      "grad_norm": 19.78436279296875,
      "learning_rate": 9.784000000000002e-06,
      "loss": 3.6688,
      "step": 10800
    },
    {
      "epoch": 0.109,
      "grad_norm": 17.20643424987793,
      "learning_rate": 9.782e-06,
      "loss": 3.5023,
      "step": 10900
    },
    {
      "epoch": 0.11,
      "grad_norm": 14.809402465820312,
      "learning_rate": 9.780000000000001e-06,
      "loss": 3.4398,
      "step": 11000
    },
    {
      "epoch": 0.111,
      "grad_norm": 21.907175064086914,
      "learning_rate": 9.778e-06,
      "loss": 3.5254,
      "step": 11100
    },
    {
      "epoch": 0.112,
      "grad_norm": 23.719179153442383,
      "learning_rate": 9.776000000000001e-06,
      "loss": 3.4801,
      "step": 11200
    },
    {
      "epoch": 0.113,
      "grad_norm": 15.437023162841797,
      "learning_rate": 9.774000000000002e-06,
      "loss": 3.5724,
      "step": 11300
    },
    {
      "epoch": 0.114,
      "grad_norm": 21.82857322692871,
      "learning_rate": 9.772e-06,
      "loss": 3.5787,
      "step": 11400
    },
    {
      "epoch": 0.115,
      "grad_norm": 20.26848602294922,
      "learning_rate": 9.770000000000001e-06,
      "loss": 3.6333,
      "step": 11500
    },
    {
      "epoch": 0.116,
      "grad_norm": 17.0762882232666,
      "learning_rate": 9.768e-06,
      "loss": 3.3418,
      "step": 11600
    },
    {
      "epoch": 0.117,
      "grad_norm": 20.440383911132812,
      "learning_rate": 9.766000000000001e-06,
      "loss": 3.5072,
      "step": 11700
    },
    {
      "epoch": 0.118,
      "grad_norm": 17.19301986694336,
      "learning_rate": 9.764000000000002e-06,
      "loss": 3.491,
      "step": 11800
    },
    {
      "epoch": 0.119,
      "grad_norm": 20.88847541809082,
      "learning_rate": 9.762e-06,
      "loss": 3.5482,
      "step": 11900
    },
    {
      "epoch": 0.12,
      "grad_norm": 17.677921295166016,
      "learning_rate": 9.760000000000001e-06,
      "loss": 3.5495,
      "step": 12000
    },
    {
      "epoch": 0.121,
      "grad_norm": 19.91204833984375,
      "learning_rate": 9.758e-06,
      "loss": 3.652,
      "step": 12100
    },
    {
      "epoch": 0.122,
      "grad_norm": 39.53171920776367,
      "learning_rate": 9.756000000000001e-06,
      "loss": 3.5966,
      "step": 12200
    },
    {
      "epoch": 0.123,
      "grad_norm": 15.958291053771973,
      "learning_rate": 9.754000000000002e-06,
      "loss": 3.5364,
      "step": 12300
    },
    {
      "epoch": 0.124,
      "grad_norm": 15.132060050964355,
      "learning_rate": 9.752e-06,
      "loss": 3.468,
      "step": 12400
    },
    {
      "epoch": 0.125,
      "grad_norm": 14.320738792419434,
      "learning_rate": 9.75e-06,
      "loss": 3.3429,
      "step": 12500
    },
    {
      "epoch": 0.126,
      "grad_norm": 20.6189022064209,
      "learning_rate": 9.748e-06,
      "loss": 3.3507,
      "step": 12600
    },
    {
      "epoch": 0.127,
      "grad_norm": 14.898783683776855,
      "learning_rate": 9.746000000000001e-06,
      "loss": 3.3785,
      "step": 12700
    },
    {
      "epoch": 0.128,
      "grad_norm": 22.72285270690918,
      "learning_rate": 9.744000000000002e-06,
      "loss": 3.5126,
      "step": 12800
    },
    {
      "epoch": 0.129,
      "grad_norm": 22.329116821289062,
      "learning_rate": 9.742000000000001e-06,
      "loss": 3.4503,
      "step": 12900
    },
    {
      "epoch": 0.13,
      "grad_norm": 18.049467086791992,
      "learning_rate": 9.74e-06,
      "loss": 3.4891,
      "step": 13000
    },
    {
      "epoch": 0.131,
      "grad_norm": 14.28784465789795,
      "learning_rate": 9.738e-06,
      "loss": 3.33,
      "step": 13100
    },
    {
      "epoch": 0.132,
      "grad_norm": 19.659822463989258,
      "learning_rate": 9.736000000000001e-06,
      "loss": 3.4724,
      "step": 13200
    },
    {
      "epoch": 0.133,
      "grad_norm": 19.972923278808594,
      "learning_rate": 9.734000000000002e-06,
      "loss": 3.3965,
      "step": 13300
    },
    {
      "epoch": 0.134,
      "grad_norm": 21.733108520507812,
      "learning_rate": 9.732000000000001e-06,
      "loss": 3.584,
      "step": 13400
    },
    {
      "epoch": 0.135,
      "grad_norm": 13.769856452941895,
      "learning_rate": 9.73e-06,
      "loss": 3.4789,
      "step": 13500
    },
    {
      "epoch": 0.136,
      "grad_norm": 15.672243118286133,
      "learning_rate": 9.728e-06,
      "loss": 3.5081,
      "step": 13600
    },
    {
      "epoch": 0.137,
      "grad_norm": 17.671894073486328,
      "learning_rate": 9.726000000000001e-06,
      "loss": 3.3696,
      "step": 13700
    },
    {
      "epoch": 0.138,
      "grad_norm": 19.69550323486328,
      "learning_rate": 9.724e-06,
      "loss": 3.3707,
      "step": 13800
    },
    {
      "epoch": 0.139,
      "grad_norm": 14.621719360351562,
      "learning_rate": 9.722000000000001e-06,
      "loss": 3.4709,
      "step": 13900
    },
    {
      "epoch": 0.14,
      "grad_norm": 17.52949333190918,
      "learning_rate": 9.72e-06,
      "loss": 3.4979,
      "step": 14000
    },
    {
      "epoch": 0.141,
      "grad_norm": 15.679729461669922,
      "learning_rate": 9.718e-06,
      "loss": 3.4584,
      "step": 14100
    },
    {
      "epoch": 0.142,
      "grad_norm": 17.527435302734375,
      "learning_rate": 9.716000000000002e-06,
      "loss": 3.3812,
      "step": 14200
    },
    {
      "epoch": 0.143,
      "grad_norm": 24.084278106689453,
      "learning_rate": 9.714e-06,
      "loss": 3.4052,
      "step": 14300
    },
    {
      "epoch": 0.144,
      "grad_norm": 20.039127349853516,
      "learning_rate": 9.712e-06,
      "loss": 3.3659,
      "step": 14400
    },
    {
      "epoch": 0.145,
      "grad_norm": 18.518939971923828,
      "learning_rate": 9.71e-06,
      "loss": 3.4249,
      "step": 14500
    },
    {
      "epoch": 0.146,
      "grad_norm": 19.596946716308594,
      "learning_rate": 9.708000000000001e-06,
      "loss": 3.3909,
      "step": 14600
    },
    {
      "epoch": 0.147,
      "grad_norm": 14.774336814880371,
      "learning_rate": 9.706000000000002e-06,
      "loss": 3.3216,
      "step": 14700
    },
    {
      "epoch": 0.148,
      "grad_norm": 27.980627059936523,
      "learning_rate": 9.704e-06,
      "loss": 3.3794,
      "step": 14800
    },
    {
      "epoch": 0.149,
      "grad_norm": 16.481491088867188,
      "learning_rate": 9.702e-06,
      "loss": 3.4138,
      "step": 14900
    },
    {
      "epoch": 0.15,
      "grad_norm": 18.48386573791504,
      "learning_rate": 9.7e-06,
      "loss": 3.3635,
      "step": 15000
    },
    {
      "epoch": 0.151,
      "grad_norm": 14.089752197265625,
      "learning_rate": 9.698000000000001e-06,
      "loss": 3.363,
      "step": 15100
    },
    {
      "epoch": 0.152,
      "grad_norm": 15.205988883972168,
      "learning_rate": 9.696000000000002e-06,
      "loss": 3.3038,
      "step": 15200
    },
    {
      "epoch": 0.153,
      "grad_norm": 18.17800521850586,
      "learning_rate": 9.694e-06,
      "loss": 3.2748,
      "step": 15300
    },
    {
      "epoch": 0.154,
      "grad_norm": 15.958276748657227,
      "learning_rate": 9.692e-06,
      "loss": 3.2809,
      "step": 15400
    },
    {
      "epoch": 0.155,
      "grad_norm": 20.20997428894043,
      "learning_rate": 9.69e-06,
      "loss": 3.3366,
      "step": 15500
    },
    {
      "epoch": 0.156,
      "grad_norm": 13.844518661499023,
      "learning_rate": 9.688000000000001e-06,
      "loss": 3.424,
      "step": 15600
    },
    {
      "epoch": 0.157,
      "grad_norm": 28.56269645690918,
      "learning_rate": 9.686000000000002e-06,
      "loss": 3.4567,
      "step": 15700
    },
    {
      "epoch": 0.158,
      "grad_norm": 14.436336517333984,
      "learning_rate": 9.684e-06,
      "loss": 3.2089,
      "step": 15800
    },
    {
      "epoch": 0.159,
      "grad_norm": 28.26078987121582,
      "learning_rate": 9.682e-06,
      "loss": 3.439,
      "step": 15900
    },
    {
      "epoch": 0.16,
      "grad_norm": 19.394569396972656,
      "learning_rate": 9.68e-06,
      "loss": 3.3376,
      "step": 16000
    },
    {
      "epoch": 0.161,
      "grad_norm": 18.12739372253418,
      "learning_rate": 9.678000000000001e-06,
      "loss": 3.2295,
      "step": 16100
    },
    {
      "epoch": 0.162,
      "grad_norm": 20.420162200927734,
      "learning_rate": 9.676e-06,
      "loss": 3.307,
      "step": 16200
    },
    {
      "epoch": 0.163,
      "grad_norm": 15.34536361694336,
      "learning_rate": 9.674000000000001e-06,
      "loss": 3.3014,
      "step": 16300
    },
    {
      "epoch": 0.164,
      "grad_norm": 20.239303588867188,
      "learning_rate": 9.672e-06,
      "loss": 3.3339,
      "step": 16400
    },
    {
      "epoch": 0.165,
      "grad_norm": 15.123329162597656,
      "learning_rate": 9.67e-06,
      "loss": 3.3073,
      "step": 16500
    },
    {
      "epoch": 0.166,
      "grad_norm": 21.282299041748047,
      "learning_rate": 9.668000000000001e-06,
      "loss": 3.2809,
      "step": 16600
    },
    {
      "epoch": 0.167,
      "grad_norm": 18.039140701293945,
      "learning_rate": 9.666e-06,
      "loss": 3.3368,
      "step": 16700
    },
    {
      "epoch": 0.168,
      "grad_norm": 17.391878128051758,
      "learning_rate": 9.664000000000001e-06,
      "loss": 3.2135,
      "step": 16800
    },
    {
      "epoch": 0.169,
      "grad_norm": 29.216005325317383,
      "learning_rate": 9.662e-06,
      "loss": 3.2238,
      "step": 16900
    },
    {
      "epoch": 0.17,
      "grad_norm": 24.716182708740234,
      "learning_rate": 9.66e-06,
      "loss": 3.4032,
      "step": 17000
    },
    {
      "epoch": 0.171,
      "grad_norm": 19.68560791015625,
      "learning_rate": 9.658000000000001e-06,
      "loss": 3.2144,
      "step": 17100
    },
    {
      "epoch": 0.172,
      "grad_norm": 20.55443572998047,
      "learning_rate": 9.656e-06,
      "loss": 3.347,
      "step": 17200
    },
    {
      "epoch": 0.173,
      "grad_norm": 23.09670639038086,
      "learning_rate": 9.654000000000001e-06,
      "loss": 3.2204,
      "step": 17300
    },
    {
      "epoch": 0.174,
      "grad_norm": 21.916152954101562,
      "learning_rate": 9.652e-06,
      "loss": 3.1898,
      "step": 17400
    },
    {
      "epoch": 0.175,
      "grad_norm": 15.10058879852295,
      "learning_rate": 9.65e-06,
      "loss": 3.3174,
      "step": 17500
    },
    {
      "epoch": 0.176,
      "grad_norm": 18.47793197631836,
      "learning_rate": 9.648000000000001e-06,
      "loss": 3.152,
      "step": 17600
    },
    {
      "epoch": 0.177,
      "grad_norm": 20.482669830322266,
      "learning_rate": 9.646e-06,
      "loss": 3.2341,
      "step": 17700
    },
    {
      "epoch": 0.178,
      "grad_norm": 17.341407775878906,
      "learning_rate": 9.644000000000001e-06,
      "loss": 3.2945,
      "step": 17800
    },
    {
      "epoch": 0.179,
      "grad_norm": 25.537378311157227,
      "learning_rate": 9.642e-06,
      "loss": 3.3279,
      "step": 17900
    },
    {
      "epoch": 0.18,
      "grad_norm": 25.134294509887695,
      "learning_rate": 9.640000000000001e-06,
      "loss": 3.4005,
      "step": 18000
    },
    {
      "epoch": 0.181,
      "grad_norm": 14.844265937805176,
      "learning_rate": 9.638e-06,
      "loss": 3.2125,
      "step": 18100
    },
    {
      "epoch": 0.182,
      "grad_norm": 12.517401695251465,
      "learning_rate": 9.636e-06,
      "loss": 3.4222,
      "step": 18200
    },
    {
      "epoch": 0.183,
      "grad_norm": 13.91508674621582,
      "learning_rate": 9.634000000000001e-06,
      "loss": 3.2464,
      "step": 18300
    },
    {
      "epoch": 0.184,
      "grad_norm": 20.34067726135254,
      "learning_rate": 9.632e-06,
      "loss": 3.1909,
      "step": 18400
    },
    {
      "epoch": 0.185,
      "grad_norm": 20.126605987548828,
      "learning_rate": 9.630000000000001e-06,
      "loss": 3.1106,
      "step": 18500
    },
    {
      "epoch": 0.186,
      "grad_norm": 20.69412612915039,
      "learning_rate": 9.628e-06,
      "loss": 3.2615,
      "step": 18600
    },
    {
      "epoch": 0.187,
      "grad_norm": 29.957561492919922,
      "learning_rate": 9.626e-06,
      "loss": 3.2019,
      "step": 18700
    },
    {
      "epoch": 0.188,
      "grad_norm": 12.36296558380127,
      "learning_rate": 9.624000000000001e-06,
      "loss": 3.2337,
      "step": 18800
    },
    {
      "epoch": 0.189,
      "grad_norm": 13.685921669006348,
      "learning_rate": 9.622000000000002e-06,
      "loss": 3.282,
      "step": 18900
    },
    {
      "epoch": 0.19,
      "grad_norm": 20.060331344604492,
      "learning_rate": 9.620000000000001e-06,
      "loss": 3.1947,
      "step": 19000
    },
    {
      "epoch": 0.191,
      "grad_norm": 13.936528205871582,
      "learning_rate": 9.618e-06,
      "loss": 3.174,
      "step": 19100
    },
    {
      "epoch": 0.192,
      "grad_norm": 21.87002944946289,
      "learning_rate": 9.616e-06,
      "loss": 3.2355,
      "step": 19200
    },
    {
      "epoch": 0.193,
      "grad_norm": 24.834264755249023,
      "learning_rate": 9.614000000000001e-06,
      "loss": 3.1823,
      "step": 19300
    },
    {
      "epoch": 0.194,
      "grad_norm": 30.83481216430664,
      "learning_rate": 9.612000000000002e-06,
      "loss": 3.1645,
      "step": 19400
    },
    {
      "epoch": 0.195,
      "grad_norm": 18.5225830078125,
      "learning_rate": 9.610000000000001e-06,
      "loss": 3.2636,
      "step": 19500
    },
    {
      "epoch": 0.196,
      "grad_norm": 27.33846664428711,
      "learning_rate": 9.608e-06,
      "loss": 3.1772,
      "step": 19600
    },
    {
      "epoch": 0.197,
      "grad_norm": 21.9919490814209,
      "learning_rate": 9.606000000000001e-06,
      "loss": 3.229,
      "step": 19700
    },
    {
      "epoch": 0.198,
      "grad_norm": 19.65387725830078,
      "learning_rate": 9.604000000000002e-06,
      "loss": 3.1524,
      "step": 19800
    },
    {
      "epoch": 0.199,
      "grad_norm": 18.683229446411133,
      "learning_rate": 9.602e-06,
      "loss": 3.4325,
      "step": 19900
    },
    {
      "epoch": 0.2,
      "grad_norm": 16.26070785522461,
      "learning_rate": 9.600000000000001e-06,
      "loss": 3.1583,
      "step": 20000
    },
    {
      "epoch": 0.201,
      "grad_norm": 17.30815887451172,
      "learning_rate": 9.598e-06,
      "loss": 3.1904,
      "step": 20100
    },
    {
      "epoch": 0.202,
      "grad_norm": 28.912694931030273,
      "learning_rate": 9.596000000000001e-06,
      "loss": 3.0794,
      "step": 20200
    },
    {
      "epoch": 0.203,
      "grad_norm": 20.792774200439453,
      "learning_rate": 9.594000000000002e-06,
      "loss": 3.2403,
      "step": 20300
    },
    {
      "epoch": 0.204,
      "grad_norm": 22.178218841552734,
      "learning_rate": 9.592e-06,
      "loss": 3.1408,
      "step": 20400
    },
    {
      "epoch": 0.205,
      "grad_norm": 15.090167045593262,
      "learning_rate": 9.59e-06,
      "loss": 3.1513,
      "step": 20500
    },
    {
      "epoch": 0.206,
      "grad_norm": 19.66379737854004,
      "learning_rate": 9.588e-06,
      "loss": 3.1574,
      "step": 20600
    },
    {
      "epoch": 0.207,
      "grad_norm": 20.961610794067383,
      "learning_rate": 9.586000000000001e-06,
      "loss": 3.284,
      "step": 20700
    },
    {
      "epoch": 0.208,
      "grad_norm": 19.434553146362305,
      "learning_rate": 9.584000000000002e-06,
      "loss": 3.0802,
      "step": 20800
    },
    {
      "epoch": 0.209,
      "grad_norm": 30.214740753173828,
      "learning_rate": 9.582e-06,
      "loss": 3.2555,
      "step": 20900
    },
    {
      "epoch": 0.21,
      "grad_norm": 18.16490364074707,
      "learning_rate": 9.58e-06,
      "loss": 3.2179,
      "step": 21000
    },
    {
      "epoch": 0.211,
      "grad_norm": 22.568527221679688,
      "learning_rate": 9.578e-06,
      "loss": 3.2268,
      "step": 21100
    },
    {
      "epoch": 0.212,
      "grad_norm": 20.349346160888672,
      "learning_rate": 9.576000000000001e-06,
      "loss": 3.159,
      "step": 21200
    },
    {
      "epoch": 0.213,
      "grad_norm": 23.45667266845703,
      "learning_rate": 9.574000000000002e-06,
      "loss": 3.1728,
      "step": 21300
    },
    {
      "epoch": 0.214,
      "grad_norm": 20.883718490600586,
      "learning_rate": 9.572000000000001e-06,
      "loss": 3.1258,
      "step": 21400
    },
    {
      "epoch": 0.215,
      "grad_norm": 25.16787338256836,
      "learning_rate": 9.57e-06,
      "loss": 3.0726,
      "step": 21500
    },
    {
      "epoch": 0.216,
      "grad_norm": 21.36046028137207,
      "learning_rate": 9.568e-06,
      "loss": 3.1267,
      "step": 21600
    },
    {
      "epoch": 0.217,
      "grad_norm": 26.431421279907227,
      "learning_rate": 9.566000000000001e-06,
      "loss": 3.1588,
      "step": 21700
    },
    {
      "epoch": 0.218,
      "grad_norm": 27.33740997314453,
      "learning_rate": 9.564e-06,
      "loss": 3.0679,
      "step": 21800
    },
    {
      "epoch": 0.219,
      "grad_norm": 17.818220138549805,
      "learning_rate": 9.562000000000001e-06,
      "loss": 3.1104,
      "step": 21900
    },
    {
      "epoch": 0.22,
      "grad_norm": 25.339937210083008,
      "learning_rate": 9.56e-06,
      "loss": 3.2342,
      "step": 22000
    },
    {
      "epoch": 0.221,
      "grad_norm": 19.325305938720703,
      "learning_rate": 9.558e-06,
      "loss": 3.1085,
      "step": 22100
    },
    {
      "epoch": 0.222,
      "grad_norm": 19.849441528320312,
      "learning_rate": 9.556000000000001e-06,
      "loss": 3.2078,
      "step": 22200
    },
    {
      "epoch": 0.223,
      "grad_norm": 22.334917068481445,
      "learning_rate": 9.554e-06,
      "loss": 3.1536,
      "step": 22300
    },
    {
      "epoch": 0.224,
      "grad_norm": 16.38900375366211,
      "learning_rate": 9.552000000000001e-06,
      "loss": 3.106,
      "step": 22400
    },
    {
      "epoch": 0.225,
      "grad_norm": 24.00871467590332,
      "learning_rate": 9.55e-06,
      "loss": 3.1776,
      "step": 22500
    },
    {
      "epoch": 0.226,
      "grad_norm": 19.4804744720459,
      "learning_rate": 9.548e-06,
      "loss": 3.164,
      "step": 22600
    },
    {
      "epoch": 0.227,
      "grad_norm": 17.325008392333984,
      "learning_rate": 9.546000000000001e-06,
      "loss": 3.2499,
      "step": 22700
    },
    {
      "epoch": 0.228,
      "grad_norm": 19.2254695892334,
      "learning_rate": 9.544e-06,
      "loss": 3.1912,
      "step": 22800
    },
    {
      "epoch": 0.229,
      "grad_norm": 19.877927780151367,
      "learning_rate": 9.542000000000001e-06,
      "loss": 3.1645,
      "step": 22900
    },
    {
      "epoch": 0.23,
      "grad_norm": 21.79277992248535,
      "learning_rate": 9.54e-06,
      "loss": 3.0794,
      "step": 23000
    }
  ],
  "logging_steps": 100,
  "max_steps": 500000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4239469785088e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}