{
  "best_metric": 0.8431416153907776,
  "best_model_checkpoint": "miner_id_24/checkpoint-50",
  "epoch": 2.6315789473684212,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02631578947368421,
      "grad_norm": 35.830352783203125,
      "learning_rate": 1e-05,
      "loss": 5.5742,
      "step": 1
    },
    {
      "epoch": 0.02631578947368421,
      "eval_loss": 1.6754724979400635,
      "eval_runtime": 4.6369,
      "eval_samples_per_second": 13.802,
      "eval_steps_per_second": 3.451,
      "step": 1
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 45.00202560424805,
      "learning_rate": 2e-05,
      "loss": 6.0369,
      "step": 2
    },
    {
      "epoch": 0.07894736842105263,
      "grad_norm": 46.17910385131836,
      "learning_rate": 3e-05,
      "loss": 6.0642,
      "step": 3
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 49.760963439941406,
      "learning_rate": 4e-05,
      "loss": 5.5496,
      "step": 4
    },
    {
      "epoch": 0.13157894736842105,
      "grad_norm": 39.75947189331055,
      "learning_rate": 5e-05,
      "loss": 5.168,
      "step": 5
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 41.92656707763672,
      "learning_rate": 6e-05,
      "loss": 4.7943,
      "step": 6
    },
    {
      "epoch": 0.18421052631578946,
      "grad_norm": 27.3580322265625,
      "learning_rate": 7e-05,
      "loss": 3.9595,
      "step": 7
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 34.04036331176758,
      "learning_rate": 8e-05,
      "loss": 4.0193,
      "step": 8
    },
    {
      "epoch": 0.23684210526315788,
      "grad_norm": 39.38542556762695,
      "learning_rate": 9e-05,
      "loss": 4.7816,
      "step": 9
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 42.04862976074219,
      "learning_rate": 0.0001,
      "loss": 4.7192,
      "step": 10
    },
    {
      "epoch": 0.2894736842105263,
      "grad_norm": 16.514020919799805,
      "learning_rate": 9.997718922447667e-05,
      "loss": 3.6934,
      "step": 11
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 14.977937698364258,
      "learning_rate": 9.990877771116589e-05,
      "loss": 4.0051,
      "step": 12
    },
    {
      "epoch": 0.34210526315789475,
      "grad_norm": 16.822614669799805,
      "learning_rate": 9.979482788085454e-05,
      "loss": 3.3175,
      "step": 13
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 17.509784698486328,
      "learning_rate": 9.96354437049027e-05,
      "loss": 3.4816,
      "step": 14
    },
    {
      "epoch": 0.39473684210526316,
      "grad_norm": 15.930502891540527,
      "learning_rate": 9.943077061037671e-05,
      "loss": 3.2727,
      "step": 15
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 20.140819549560547,
      "learning_rate": 9.918099534735718e-05,
      "loss": 3.531,
      "step": 16
    },
    {
      "epoch": 0.4473684210526316,
      "grad_norm": 21.51377296447754,
      "learning_rate": 9.888634581854234e-05,
      "loss": 3.2109,
      "step": 17
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 21.92951202392578,
      "learning_rate": 9.85470908713026e-05,
      "loss": 3.6691,
      "step": 18
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.309015274047852,
      "learning_rate": 9.816354005237583e-05,
      "loss": 3.3325,
      "step": 19
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 14.019783973693848,
      "learning_rate": 9.773604332542729e-05,
      "loss": 3.8225,
      "step": 20
    },
    {
      "epoch": 0.5526315789473685,
      "grad_norm": 12.386563301086426,
      "learning_rate": 9.726499075173201e-05,
      "loss": 3.4366,
      "step": 21
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 13.775650978088379,
      "learning_rate": 9.675081213427076e-05,
      "loss": 3.6102,
      "step": 22
    },
    {
      "epoch": 0.6052631578947368,
      "grad_norm": 13.588586807250977,
      "learning_rate": 9.619397662556435e-05,
      "loss": 3.7363,
      "step": 23
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 13.898175239562988,
      "learning_rate": 9.559499229960451e-05,
      "loss": 3.2825,
      "step": 24
    },
    {
      "epoch": 0.6578947368421053,
      "grad_norm": 15.590497970581055,
      "learning_rate": 9.495440568827129e-05,
      "loss": 3.2594,
      "step": 25
    },
    {
      "epoch": 0.6842105263157895,
      "grad_norm": 18.052392959594727,
      "learning_rate": 9.42728012826605e-05,
      "loss": 3.4811,
      "step": 26
    },
    {
      "epoch": 0.7105263157894737,
      "grad_norm": 20.789844512939453,
      "learning_rate": 9.355080099977578e-05,
      "loss": 4.004,
      "step": 27
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 11.545269012451172,
      "learning_rate": 9.278906361507238e-05,
      "loss": 3.407,
      "step": 28
    },
    {
      "epoch": 0.7631578947368421,
      "grad_norm": 11.368681907653809,
      "learning_rate": 9.19882841613699e-05,
      "loss": 3.212,
      "step": 29
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 12.940420150756836,
      "learning_rate": 9.114919329468282e-05,
      "loss": 3.4642,
      "step": 30
    },
    {
      "epoch": 0.8157894736842105,
      "grad_norm": 11.59272575378418,
      "learning_rate": 9.02725566275473e-05,
      "loss": 3.3124,
      "step": 31
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 17.267396926879883,
      "learning_rate": 8.935917403045251e-05,
      "loss": 3.5721,
      "step": 32
    },
    {
      "epoch": 0.868421052631579,
      "grad_norm": 14.882741928100586,
      "learning_rate": 8.840987890201403e-05,
      "loss": 3.7928,
      "step": 33
    },
    {
      "epoch": 0.8947368421052632,
      "grad_norm": 14.032687187194824,
      "learning_rate": 8.742553740855506e-05,
      "loss": 3.2582,
      "step": 34
    },
    {
      "epoch": 0.9210526315789473,
      "grad_norm": 13.775826454162598,
      "learning_rate": 8.640704769378942e-05,
      "loss": 3.4732,
      "step": 35
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 17.22196388244629,
      "learning_rate": 8.535533905932738e-05,
      "loss": 3.6359,
      "step": 36
    },
    {
      "epoch": 0.9736842105263158,
      "grad_norm": 14.44973373413086,
      "learning_rate": 8.427137111675199e-05,
      "loss": 3.5035,
      "step": 37
    },
    {
      "epoch": 1.0,
      "grad_norm": 17.80911636352539,
      "learning_rate": 8.315613291203976e-05,
      "loss": 3.3937,
      "step": 38
    },
    {
      "epoch": 1.0263157894736843,
      "grad_norm": 8.416217803955078,
      "learning_rate": 8.201064202312441e-05,
      "loss": 2.8219,
      "step": 39
    },
    {
      "epoch": 1.0526315789473684,
      "grad_norm": 8.224820137023926,
      "learning_rate": 8.083594363142717e-05,
      "loss": 1.9419,
      "step": 40
    },
    {
      "epoch": 1.0789473684210527,
      "grad_norm": 8.183276176452637,
      "learning_rate": 7.963310956820085e-05,
      "loss": 1.795,
      "step": 41
    },
    {
      "epoch": 1.1052631578947367,
      "grad_norm": 8.23655891418457,
      "learning_rate": 7.840323733655778e-05,
      "loss": 1.6889,
      "step": 42
    },
    {
      "epoch": 1.131578947368421,
      "grad_norm": 11.236390113830566,
      "learning_rate": 7.714744911007394e-05,
      "loss": 1.6847,
      "step": 43
    },
    {
      "epoch": 1.1578947368421053,
      "grad_norm": 10.796011924743652,
      "learning_rate": 7.586689070888284e-05,
      "loss": 1.4455,
      "step": 44
    },
    {
      "epoch": 1.1842105263157894,
      "grad_norm": 10.806239128112793,
      "learning_rate": 7.456273055419388e-05,
      "loss": 1.1321,
      "step": 45
    },
    {
      "epoch": 1.2105263157894737,
      "grad_norm": 11.67683219909668,
      "learning_rate": 7.323615860218843e-05,
      "loss": 1.0502,
      "step": 46
    },
    {
      "epoch": 1.236842105263158,
      "grad_norm": 15.044631958007812,
      "learning_rate": 7.188838525826702e-05,
      "loss": 1.189,
      "step": 47
    },
    {
      "epoch": 1.263157894736842,
      "grad_norm": 13.156881332397461,
      "learning_rate": 7.052064027263786e-05,
      "loss": 2.1428,
      "step": 48
    },
    {
      "epoch": 1.2894736842105263,
      "grad_norm": 11.645526885986328,
      "learning_rate": 6.91341716182545e-05,
      "loss": 1.6295,
      "step": 49
    },
    {
      "epoch": 1.3157894736842106,
      "grad_norm": 10.376519203186035,
      "learning_rate": 6.773024435212678e-05,
      "loss": 1.609,
      "step": 50
    },
    {
      "epoch": 1.3157894736842106,
      "eval_loss": 0.8431416153907776,
      "eval_runtime": 4.7683,
      "eval_samples_per_second": 13.422,
      "eval_steps_per_second": 3.355,
      "step": 50
    },
    {
      "epoch": 1.3421052631578947,
      "grad_norm": 9.306763648986816,
      "learning_rate": 6.631013946104347e-05,
      "loss": 1.7307,
      "step": 51
    },
    {
      "epoch": 1.368421052631579,
      "grad_norm": 9.575918197631836,
      "learning_rate": 6.487515269276016e-05,
      "loss": 1.4591,
      "step": 52
    },
    {
      "epoch": 1.3947368421052633,
      "grad_norm": 11.862869262695312,
      "learning_rate": 6.342659337371885e-05,
      "loss": 1.1487,
      "step": 53
    },
    {
      "epoch": 1.4210526315789473,
      "grad_norm": 10.206483840942383,
      "learning_rate": 6.19657832143779e-05,
      "loss": 1.3008,
      "step": 54
    },
    {
      "epoch": 1.4473684210526316,
      "grad_norm": 12.067391395568848,
      "learning_rate": 6.049405510324238e-05,
      "loss": 1.1472,
      "step": 55
    },
    {
      "epoch": 1.4736842105263157,
      "grad_norm": 15.269783020019531,
      "learning_rate": 5.90127518906953e-05,
      "loss": 1.6159,
      "step": 56
    },
    {
      "epoch": 1.5,
      "grad_norm": 10.770502090454102,
      "learning_rate": 5.752322516373916e-05,
      "loss": 2.367,
      "step": 57
    },
    {
      "epoch": 1.526315789473684,
      "grad_norm": 8.79799747467041,
      "learning_rate": 5.602683401276615e-05,
      "loss": 1.6744,
      "step": 58
    },
    {
      "epoch": 1.5526315789473686,
      "grad_norm": 10.059900283813477,
      "learning_rate": 5.45249437914819e-05,
      "loss": 1.6454,
      "step": 59
    },
    {
      "epoch": 1.5789473684210527,
      "grad_norm": 8.759090423583984,
      "learning_rate": 5.3018924871114305e-05,
      "loss": 1.47,
      "step": 60
    },
    {
      "epoch": 1.6052631578947367,
      "grad_norm": 9.53599739074707,
      "learning_rate": 5.151015139004445e-05,
      "loss": 1.2789,
      "step": 61
    },
    {
      "epoch": 1.631578947368421,
      "grad_norm": 10.73863697052002,
      "learning_rate": 5e-05,
      "loss": 1.3115,
      "step": 62
    },
    {
      "epoch": 1.6578947368421053,
      "grad_norm": 14.568950653076172,
      "learning_rate": 4.848984860995557e-05,
      "loss": 1.2271,
      "step": 63
    },
    {
      "epoch": 1.6842105263157894,
      "grad_norm": 10.92098617553711,
      "learning_rate": 4.6981075128885693e-05,
      "loss": 1.2193,
      "step": 64
    },
    {
      "epoch": 1.7105263157894737,
      "grad_norm": 15.96349811553955,
      "learning_rate": 4.547505620851811e-05,
      "loss": 1.0586,
      "step": 65
    },
    {
      "epoch": 1.736842105263158,
      "grad_norm": 10.191640853881836,
      "learning_rate": 4.397316598723385e-05,
      "loss": 2.1627,
      "step": 66
    },
    {
      "epoch": 1.763157894736842,
      "grad_norm": 8.593461036682129,
      "learning_rate": 4.2476774836260845e-05,
      "loss": 1.5354,
      "step": 67
    },
    {
      "epoch": 1.7894736842105263,
      "grad_norm": 8.911328315734863,
      "learning_rate": 4.0987248109304714e-05,
      "loss": 1.5173,
      "step": 68
    },
    {
      "epoch": 1.8157894736842106,
      "grad_norm": 9.834583282470703,
      "learning_rate": 3.950594489675763e-05,
      "loss": 1.2573,
      "step": 69
    },
    {
      "epoch": 1.8421052631578947,
      "grad_norm": 14.315779685974121,
      "learning_rate": 3.803421678562213e-05,
      "loss": 1.5104,
      "step": 70
    },
    {
      "epoch": 1.868421052631579,
      "grad_norm": 10.957352638244629,
      "learning_rate": 3.657340662628116e-05,
      "loss": 1.2571,
      "step": 71
    },
    {
      "epoch": 1.8947368421052633,
      "grad_norm": 8.86029052734375,
      "learning_rate": 3.512484730723986e-05,
      "loss": 1.1201,
      "step": 72
    },
    {
      "epoch": 1.9210526315789473,
      "grad_norm": 19.02982521057129,
      "learning_rate": 3.368986053895655e-05,
      "loss": 1.3237,
      "step": 73
    },
    {
      "epoch": 1.9473684210526314,
      "grad_norm": 14.128595352172852,
      "learning_rate": 3.226975564787322e-05,
      "loss": 1.2664,
      "step": 74
    },
    {
      "epoch": 1.973684210526316,
      "grad_norm": 9.162415504455566,
      "learning_rate": 3.086582838174551e-05,
      "loss": 1.7131,
      "step": 75
    },
    {
      "epoch": 2.0,
      "grad_norm": 17.592689514160156,
      "learning_rate": 2.9479359727362173e-05,
      "loss": 1.3134,
      "step": 76
    },
    {
      "epoch": 2.026315789473684,
      "grad_norm": 5.179409027099609,
      "learning_rate": 2.811161474173297e-05,
      "loss": 1.1209,
      "step": 77
    },
    {
      "epoch": 2.0526315789473686,
      "grad_norm": 5.210384845733643,
      "learning_rate": 2.6763841397811573e-05,
      "loss": 0.6815,
      "step": 78
    },
    {
      "epoch": 2.0789473684210527,
      "grad_norm": 4.825469970703125,
      "learning_rate": 2.5437269445806145e-05,
      "loss": 0.5254,
      "step": 79
    },
    {
      "epoch": 2.1052631578947367,
      "grad_norm": 6.13608980178833,
      "learning_rate": 2.4133109291117156e-05,
      "loss": 0.4893,
      "step": 80
    },
    {
      "epoch": 2.1315789473684212,
      "grad_norm": 5.591810703277588,
      "learning_rate": 2.2852550889926067e-05,
      "loss": 0.4407,
      "step": 81
    },
    {
      "epoch": 2.1578947368421053,
      "grad_norm": 7.048465728759766,
      "learning_rate": 2.1596762663442218e-05,
      "loss": 0.3898,
      "step": 82
    },
    {
      "epoch": 2.1842105263157894,
      "grad_norm": 5.151188850402832,
      "learning_rate": 2.0366890431799167e-05,
      "loss": 0.2723,
      "step": 83
    },
    {
      "epoch": 2.2105263157894735,
      "grad_norm": 10.724916458129883,
      "learning_rate": 1.9164056368572846e-05,
      "loss": 0.3257,
      "step": 84
    },
    {
      "epoch": 2.236842105263158,
      "grad_norm": 8.111312866210938,
      "learning_rate": 1.7989357976875603e-05,
      "loss": 0.3198,
      "step": 85
    },
    {
      "epoch": 2.263157894736842,
      "grad_norm": 9.9498291015625,
      "learning_rate": 1.684386708796025e-05,
      "loss": 1.3307,
      "step": 86
    },
    {
      "epoch": 2.2894736842105265,
      "grad_norm": 9.693532943725586,
      "learning_rate": 1.5728628883248007e-05,
      "loss": 0.5567,
      "step": 87
    },
    {
      "epoch": 2.3157894736842106,
      "grad_norm": 9.346628189086914,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.5224,
      "step": 88
    },
    {
      "epoch": 2.3421052631578947,
      "grad_norm": 7.563309192657471,
      "learning_rate": 1.3592952306210588e-05,
      "loss": 0.3015,
      "step": 89
    },
    {
      "epoch": 2.3684210526315788,
      "grad_norm": 14.845423698425293,
      "learning_rate": 1.257446259144494e-05,
      "loss": 0.4559,
      "step": 90
    },
    {
      "epoch": 2.3947368421052633,
      "grad_norm": 7.790347099304199,
      "learning_rate": 1.159012109798598e-05,
      "loss": 0.2831,
      "step": 91
    },
    {
      "epoch": 2.4210526315789473,
      "grad_norm": 6.179875373840332,
      "learning_rate": 1.0640825969547496e-05,
      "loss": 0.2589,
      "step": 92
    },
    {
      "epoch": 2.4473684210526314,
      "grad_norm": 11.59399700164795,
      "learning_rate": 9.7274433724527e-06,
      "loss": 0.3414,
      "step": 93
    },
    {
      "epoch": 2.473684210526316,
      "grad_norm": 11.449027061462402,
      "learning_rate": 8.850806705317183e-06,
      "loss": 0.3312,
      "step": 94
    },
    {
      "epoch": 2.5,
      "grad_norm": 8.000958442687988,
      "learning_rate": 8.011715838630107e-06,
      "loss": 0.8506,
      "step": 95
    },
    {
      "epoch": 2.526315789473684,
      "grad_norm": 7.2397871017456055,
      "learning_rate": 7.21093638492763e-06,
      "loss": 0.5091,
      "step": 96
    },
    {
      "epoch": 2.5526315789473686,
      "grad_norm": 10.7366304397583,
      "learning_rate": 6.449199000224221e-06,
      "loss": 0.502,
      "step": 97
    },
    {
      "epoch": 2.5789473684210527,
      "grad_norm": 9.00999641418457,
      "learning_rate": 5.727198717339511e-06,
      "loss": 0.4939,
      "step": 98
    },
    {
      "epoch": 2.6052631578947367,
      "grad_norm": 8.740212440490723,
      "learning_rate": 5.045594311728707e-06,
      "loss": 0.4444,
      "step": 99
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 9.263764381408691,
      "learning_rate": 4.405007700395497e-06,
      "loss": 0.3217,
      "step": 100
    },
    {
      "epoch": 2.6315789473684212,
      "eval_loss": 0.9828011393547058,
      "eval_runtime": 4.7664,
      "eval_samples_per_second": 13.427,
      "eval_steps_per_second": 3.357,
      "step": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 114,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 1
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.4247389962646323e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}