{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7111111111111111,
  "eval_steps": 100,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0071111111111111115,
      "grad_norm": 34.9204216003418,
      "learning_rate": 0.0,
      "loss": 6.0755,
      "step": 1
    },
    {
      "epoch": 0.014222222222222223,
      "grad_norm": 36.789642333984375,
      "learning_rate": 4.000000000000001e-06,
      "loss": 6.0835,
      "step": 2
    },
    {
      "epoch": 0.021333333333333333,
      "grad_norm": 31.312091827392578,
      "learning_rate": 8.000000000000001e-06,
      "loss": 6.0388,
      "step": 3
    },
    {
      "epoch": 0.028444444444444446,
      "grad_norm": 39.18890380859375,
      "learning_rate": 1.2e-05,
      "loss": 5.9818,
      "step": 4
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 44.72117614746094,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 6.0128,
      "step": 5
    },
    {
      "epoch": 0.042666666666666665,
      "grad_norm": 35.42960739135742,
      "learning_rate": 2e-05,
      "loss": 5.9398,
      "step": 6
    },
    {
      "epoch": 0.049777777777777775,
      "grad_norm": 42.25581359863281,
      "learning_rate": 1.985294117647059e-05,
      "loss": 6.0472,
      "step": 7
    },
    {
      "epoch": 0.05688888888888889,
      "grad_norm": 41.63883972167969,
      "learning_rate": 1.9705882352941178e-05,
      "loss": 5.9591,
      "step": 8
    },
    {
      "epoch": 0.064,
      "grad_norm": 49.907203674316406,
      "learning_rate": 1.9558823529411766e-05,
      "loss": 5.945,
      "step": 9
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 54.05959701538086,
      "learning_rate": 1.9411764705882355e-05,
      "loss": 6.207,
      "step": 10
    },
    {
      "epoch": 0.07822222222222222,
      "grad_norm": 56.534915924072266,
      "learning_rate": 1.9264705882352943e-05,
      "loss": 5.7977,
      "step": 11
    },
    {
      "epoch": 0.08533333333333333,
      "grad_norm": 61.910316467285156,
      "learning_rate": 1.911764705882353e-05,
      "loss": 5.8917,
      "step": 12
    },
    {
      "epoch": 0.09244444444444444,
      "grad_norm": 58.80147933959961,
      "learning_rate": 1.897058823529412e-05,
      "loss": 5.759,
      "step": 13
    },
    {
      "epoch": 0.09955555555555555,
      "grad_norm": 72.54361724853516,
      "learning_rate": 1.8823529411764708e-05,
      "loss": 5.7343,
      "step": 14
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 63.22592544555664,
      "learning_rate": 1.8676470588235296e-05,
      "loss": 5.5566,
      "step": 15
    },
    {
      "epoch": 0.11377777777777778,
      "grad_norm": 69.95305633544922,
      "learning_rate": 1.8529411764705884e-05,
      "loss": 5.4865,
      "step": 16
    },
    {
      "epoch": 0.12088888888888889,
      "grad_norm": 65.0767822265625,
      "learning_rate": 1.8382352941176472e-05,
      "loss": 5.5866,
      "step": 17
    },
    {
      "epoch": 0.128,
      "grad_norm": 85.87500762939453,
      "learning_rate": 1.823529411764706e-05,
      "loss": 5.792,
      "step": 18
    },
    {
      "epoch": 0.1351111111111111,
      "grad_norm": 98.18934631347656,
      "learning_rate": 1.808823529411765e-05,
      "loss": 5.585,
      "step": 19
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 98.60446166992188,
      "learning_rate": 1.7941176470588237e-05,
      "loss": 5.3223,
      "step": 20
    },
    {
      "epoch": 0.14933333333333335,
      "grad_norm": 94.6161880493164,
      "learning_rate": 1.7794117647058825e-05,
      "loss": 5.3288,
      "step": 21
    },
    {
      "epoch": 0.15644444444444444,
      "grad_norm": 94.35704803466797,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 5.0231,
      "step": 22
    },
    {
      "epoch": 0.16355555555555557,
      "grad_norm": 94.95283508300781,
      "learning_rate": 1.7500000000000002e-05,
      "loss": 5.0508,
      "step": 23
    },
    {
      "epoch": 0.17066666666666666,
      "grad_norm": 103.9607162475586,
      "learning_rate": 1.735294117647059e-05,
      "loss": 5.1962,
      "step": 24
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 80.75688934326172,
      "learning_rate": 1.720588235294118e-05,
      "loss": 5.1615,
      "step": 25
    },
    {
      "epoch": 0.18488888888888888,
      "grad_norm": 105.40040588378906,
      "learning_rate": 1.7058823529411767e-05,
      "loss": 4.9877,
      "step": 26
    },
    {
      "epoch": 0.192,
      "grad_norm": 127.15079498291016,
      "learning_rate": 1.6911764705882355e-05,
      "loss": 4.5645,
      "step": 27
    },
    {
      "epoch": 0.1991111111111111,
      "grad_norm": 115.33025360107422,
      "learning_rate": 1.6764705882352943e-05,
      "loss": 4.6548,
      "step": 28
    },
    {
      "epoch": 0.20622222222222222,
      "grad_norm": 118.01459503173828,
      "learning_rate": 1.661764705882353e-05,
      "loss": 4.3385,
      "step": 29
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 115.82317352294922,
      "learning_rate": 1.647058823529412e-05,
      "loss": 4.3594,
      "step": 30
    },
    {
      "epoch": 0.22044444444444444,
      "grad_norm": 127.07386779785156,
      "learning_rate": 1.6323529411764708e-05,
      "loss": 4.3937,
      "step": 31
    },
    {
      "epoch": 0.22755555555555557,
      "grad_norm": 121.3897705078125,
      "learning_rate": 1.6176470588235296e-05,
      "loss": 4.2158,
      "step": 32
    },
    {
      "epoch": 0.23466666666666666,
      "grad_norm": 148.39422607421875,
      "learning_rate": 1.6029411764705884e-05,
      "loss": 3.9157,
      "step": 33
    },
    {
      "epoch": 0.24177777777777779,
      "grad_norm": 131.06150817871094,
      "learning_rate": 1.5882352941176473e-05,
      "loss": 4.0095,
      "step": 34
    },
    {
      "epoch": 0.24888888888888888,
      "grad_norm": 141.60255432128906,
      "learning_rate": 1.573529411764706e-05,
      "loss": 3.6485,
      "step": 35
    },
    {
      "epoch": 0.256,
      "grad_norm": 157.71084594726562,
      "learning_rate": 1.558823529411765e-05,
      "loss": 3.4432,
      "step": 36
    },
    {
      "epoch": 0.26311111111111113,
      "grad_norm": 167.8360137939453,
      "learning_rate": 1.5441176470588237e-05,
      "loss": 3.278,
      "step": 37
    },
    {
      "epoch": 0.2702222222222222,
      "grad_norm": 165.1108856201172,
      "learning_rate": 1.5294117647058822e-05,
      "loss": 3.4334,
      "step": 38
    },
    {
      "epoch": 0.2773333333333333,
      "grad_norm": 162.79019165039062,
      "learning_rate": 1.5147058823529412e-05,
      "loss": 3.5797,
      "step": 39
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 160.90052795410156,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 2.8089,
      "step": 40
    },
    {
      "epoch": 0.29155555555555557,
      "grad_norm": 179.30435180664062,
      "learning_rate": 1.485294117647059e-05,
      "loss": 3.0078,
      "step": 41
    },
    {
      "epoch": 0.2986666666666667,
      "grad_norm": 148.90118408203125,
      "learning_rate": 1.4705882352941179e-05,
      "loss": 2.5788,
      "step": 42
    },
    {
      "epoch": 0.30577777777777776,
      "grad_norm": 155.11196899414062,
      "learning_rate": 1.4558823529411765e-05,
      "loss": 2.7402,
      "step": 43
    },
    {
      "epoch": 0.3128888888888889,
      "grad_norm": 154.757080078125,
      "learning_rate": 1.4411764705882353e-05,
      "loss": 2.3056,
      "step": 44
    },
    {
      "epoch": 0.32,
      "grad_norm": 142.23069763183594,
      "learning_rate": 1.4264705882352943e-05,
      "loss": 2.6879,
      "step": 45
    },
    {
      "epoch": 0.32711111111111113,
      "grad_norm": 134.41738891601562,
      "learning_rate": 1.4117647058823532e-05,
      "loss": 2.5199,
      "step": 46
    },
    {
      "epoch": 0.3342222222222222,
      "grad_norm": 114.61293029785156,
      "learning_rate": 1.3970588235294118e-05,
      "loss": 2.4227,
      "step": 47
    },
    {
      "epoch": 0.3413333333333333,
      "grad_norm": 135.13818359375,
      "learning_rate": 1.3823529411764706e-05,
      "loss": 2.5127,
      "step": 48
    },
    {
      "epoch": 0.34844444444444445,
      "grad_norm": 123.55288696289062,
      "learning_rate": 1.3676470588235296e-05,
      "loss": 2.0785,
      "step": 49
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 111.29792022705078,
      "learning_rate": 1.3529411764705885e-05,
      "loss": 2.0248,
      "step": 50
    },
    {
      "epoch": 0.3626666666666667,
      "grad_norm": 119.89157104492188,
      "learning_rate": 1.3382352941176471e-05,
      "loss": 2.0584,
      "step": 51
    },
    {
      "epoch": 0.36977777777777776,
      "grad_norm": 115.83722686767578,
      "learning_rate": 1.323529411764706e-05,
      "loss": 1.9607,
      "step": 52
    },
    {
      "epoch": 0.3768888888888889,
      "grad_norm": 107.39794921875,
      "learning_rate": 1.3088235294117648e-05,
      "loss": 1.6879,
      "step": 53
    },
    {
      "epoch": 0.384,
      "grad_norm": 100.39895629882812,
      "learning_rate": 1.2941176470588238e-05,
      "loss": 1.9351,
      "step": 54
    },
    {
      "epoch": 0.39111111111111113,
      "grad_norm": 95.0646743774414,
      "learning_rate": 1.2794117647058824e-05,
      "loss": 1.9057,
      "step": 55
    },
    {
      "epoch": 0.3982222222222222,
      "grad_norm": 95.53547668457031,
      "learning_rate": 1.2647058823529412e-05,
      "loss": 1.709,
      "step": 56
    },
    {
      "epoch": 0.4053333333333333,
      "grad_norm": 98.19358825683594,
      "learning_rate": 1.25e-05,
      "loss": 1.7018,
      "step": 57
    },
    {
      "epoch": 0.41244444444444445,
      "grad_norm": 99.6186752319336,
      "learning_rate": 1.235294117647059e-05,
      "loss": 1.4786,
      "step": 58
    },
    {
      "epoch": 0.41955555555555557,
      "grad_norm": 93.17119598388672,
      "learning_rate": 1.2205882352941179e-05,
      "loss": 1.471,
      "step": 59
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 96.7269515991211,
      "learning_rate": 1.2058823529411765e-05,
      "loss": 1.4655,
      "step": 60
    },
    {
      "epoch": 0.43377777777777776,
      "grad_norm": 95.324462890625,
      "learning_rate": 1.1911764705882354e-05,
      "loss": 1.4218,
      "step": 61
    },
    {
      "epoch": 0.4408888888888889,
      "grad_norm": 97.12982940673828,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 1.5315,
      "step": 62
    },
    {
      "epoch": 0.448,
      "grad_norm": 96.19091796875,
      "learning_rate": 1.1617647058823532e-05,
      "loss": 1.475,
      "step": 63
    },
    {
      "epoch": 0.45511111111111113,
      "grad_norm": 96.00982666015625,
      "learning_rate": 1.1470588235294118e-05,
      "loss": 1.2604,
      "step": 64
    },
    {
      "epoch": 0.4622222222222222,
      "grad_norm": 89.28826904296875,
      "learning_rate": 1.1323529411764707e-05,
      "loss": 1.4599,
      "step": 65
    },
    {
      "epoch": 0.4693333333333333,
      "grad_norm": 100.41348266601562,
      "learning_rate": 1.1176470588235295e-05,
      "loss": 1.0656,
      "step": 66
    },
    {
      "epoch": 0.47644444444444445,
      "grad_norm": 96.15679168701172,
      "learning_rate": 1.1029411764705885e-05,
      "loss": 1.1693,
      "step": 67
    },
    {
      "epoch": 0.48355555555555557,
      "grad_norm": 91.05823516845703,
      "learning_rate": 1.0882352941176471e-05,
      "loss": 1.1941,
      "step": 68
    },
    {
      "epoch": 0.49066666666666664,
      "grad_norm": 89.54459381103516,
      "learning_rate": 1.073529411764706e-05,
      "loss": 1.1337,
      "step": 69
    },
    {
      "epoch": 0.49777777777777776,
      "grad_norm": 86.42699432373047,
      "learning_rate": 1.0588235294117648e-05,
      "loss": 1.0407,
      "step": 70
    },
    {
      "epoch": 0.5048888888888889,
      "grad_norm": 87.24800109863281,
      "learning_rate": 1.0441176470588236e-05,
      "loss": 1.0783,
      "step": 71
    },
    {
      "epoch": 0.512,
      "grad_norm": 76.11488342285156,
      "learning_rate": 1.0294117647058823e-05,
      "loss": 0.9683,
      "step": 72
    },
    {
      "epoch": 0.5191111111111111,
      "grad_norm": 81.84366607666016,
      "learning_rate": 1.0147058823529413e-05,
      "loss": 0.923,
      "step": 73
    },
    {
      "epoch": 0.5262222222222223,
      "grad_norm": 85.62808990478516,
      "learning_rate": 1e-05,
      "loss": 0.7631,
      "step": 74
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 85.5390853881836,
      "learning_rate": 9.852941176470589e-06,
      "loss": 1.118,
      "step": 75
    },
    {
      "epoch": 0.5404444444444444,
      "grad_norm": 73.89140319824219,
      "learning_rate": 9.705882352941177e-06,
      "loss": 0.9038,
      "step": 76
    },
    {
      "epoch": 0.5475555555555556,
      "grad_norm": 85.44682312011719,
      "learning_rate": 9.558823529411766e-06,
      "loss": 0.8516,
      "step": 77
    },
    {
      "epoch": 0.5546666666666666,
      "grad_norm": 78.56574249267578,
      "learning_rate": 9.411764705882354e-06,
      "loss": 0.7713,
      "step": 78
    },
    {
      "epoch": 0.5617777777777778,
      "grad_norm": 79.07495880126953,
      "learning_rate": 9.264705882352942e-06,
      "loss": 0.6441,
      "step": 79
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 68.79619598388672,
      "learning_rate": 9.11764705882353e-06,
      "loss": 0.6827,
      "step": 80
    },
    {
      "epoch": 0.576,
      "grad_norm": 66.87134552001953,
      "learning_rate": 8.970588235294119e-06,
      "loss": 0.8977,
      "step": 81
    },
    {
      "epoch": 0.5831111111111111,
      "grad_norm": 67.361083984375,
      "learning_rate": 8.823529411764707e-06,
      "loss": 0.9475,
      "step": 82
    },
    {
      "epoch": 0.5902222222222222,
      "grad_norm": 55.0753173828125,
      "learning_rate": 8.676470588235295e-06,
      "loss": 0.9048,
      "step": 83
    },
    {
      "epoch": 0.5973333333333334,
      "grad_norm": 57.5294075012207,
      "learning_rate": 8.529411764705883e-06,
      "loss": 0.6339,
      "step": 84
    },
    {
      "epoch": 0.6044444444444445,
      "grad_norm": 53.539459228515625,
      "learning_rate": 8.382352941176472e-06,
      "loss": 0.4992,
      "step": 85
    },
    {
      "epoch": 0.6115555555555555,
      "grad_norm": 53.93544006347656,
      "learning_rate": 8.23529411764706e-06,
      "loss": 0.51,
      "step": 86
    },
    {
      "epoch": 0.6186666666666667,
      "grad_norm": 47.16350173950195,
      "learning_rate": 8.088235294117648e-06,
      "loss": 0.563,
      "step": 87
    },
    {
      "epoch": 0.6257777777777778,
      "grad_norm": 55.971187591552734,
      "learning_rate": 7.941176470588236e-06,
      "loss": 0.6333,
      "step": 88
    },
    {
      "epoch": 0.6328888888888888,
      "grad_norm": 56.874691009521484,
      "learning_rate": 7.794117647058825e-06,
      "loss": 0.6348,
      "step": 89
    },
    {
      "epoch": 0.64,
      "grad_norm": 41.867008209228516,
      "learning_rate": 7.647058823529411e-06,
      "loss": 0.6136,
      "step": 90
    },
    {
      "epoch": 0.6471111111111111,
      "grad_norm": 44.80308151245117,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3709,
      "step": 91
    },
    {
      "epoch": 0.6542222222222223,
      "grad_norm": 40.61701965332031,
      "learning_rate": 7.352941176470589e-06,
      "loss": 0.4615,
      "step": 92
    },
    {
      "epoch": 0.6613333333333333,
      "grad_norm": 38.063255310058594,
      "learning_rate": 7.205882352941177e-06,
      "loss": 0.3217,
      "step": 93
    },
    {
      "epoch": 0.6684444444444444,
      "grad_norm": 46.94659423828125,
      "learning_rate": 7.058823529411766e-06,
      "loss": 0.4784,
      "step": 94
    },
    {
      "epoch": 0.6755555555555556,
      "grad_norm": 37.998130798339844,
      "learning_rate": 6.911764705882353e-06,
      "loss": 0.368,
      "step": 95
    },
    {
      "epoch": 0.6826666666666666,
      "grad_norm": 50.27403259277344,
      "learning_rate": 6.764705882352942e-06,
      "loss": 0.562,
      "step": 96
    },
    {
      "epoch": 0.6897777777777778,
      "grad_norm": 54.42430114746094,
      "learning_rate": 6.61764705882353e-06,
      "loss": 0.7482,
      "step": 97
    },
    {
      "epoch": 0.6968888888888889,
      "grad_norm": 45.63814926147461,
      "learning_rate": 6.470588235294119e-06,
      "loss": 0.4871,
      "step": 98
    },
    {
      "epoch": 0.704,
      "grad_norm": 54.874820709228516,
      "learning_rate": 6.323529411764706e-06,
      "loss": 0.5447,
      "step": 99
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 39.86157989501953,
      "learning_rate": 6.176470588235295e-06,
      "loss": 0.4678,
      "step": 100
    },
    {
      "epoch": 0.7111111111111111,
      "eval_loss": 0.4020882248878479,
      "eval_runtime": 6.103,
      "eval_samples_per_second": 40.964,
      "eval_steps_per_second": 20.482,
      "step": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 141,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2155093667758080.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}