{
  "best_metric": 0.9307088255882263,
  "best_model_checkpoint": "miner_id_24/checkpoint-100",
  "epoch": 2.797202797202797,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.027972027972027972,
      "grad_norm": 37.08354187011719,
      "learning_rate": 5.000000000000001e-07,
      "loss": 9.5799,
      "step": 1
    },
    {
      "epoch": 0.027972027972027972,
      "eval_loss": 1.4193955659866333,
      "eval_runtime": 5.1305,
      "eval_samples_per_second": 11.89,
      "eval_steps_per_second": 3.119,
      "step": 1
    },
    {
      "epoch": 0.055944055944055944,
      "grad_norm": 42.282569885253906,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 10.3182,
      "step": 2
    },
    {
      "epoch": 0.08391608391608392,
      "grad_norm": 42.52220153808594,
      "learning_rate": 1.5e-06,
      "loss": 10.4365,
      "step": 3
    },
    {
      "epoch": 0.11188811188811189,
      "grad_norm": 48.22676467895508,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 10.7285,
      "step": 4
    },
    {
      "epoch": 0.13986013986013987,
      "grad_norm": 50.60873031616211,
      "learning_rate": 2.5e-06,
      "loss": 11.4862,
      "step": 5
    },
    {
      "epoch": 0.16783216783216784,
      "grad_norm": 55.39051055908203,
      "learning_rate": 3e-06,
      "loss": 12.3056,
      "step": 6
    },
    {
      "epoch": 0.1958041958041958,
      "grad_norm": 63.52677536010742,
      "learning_rate": 3.5e-06,
      "loss": 12.4326,
      "step": 7
    },
    {
      "epoch": 0.22377622377622378,
      "grad_norm": 69.61125946044922,
      "learning_rate": 4.000000000000001e-06,
      "loss": 13.8581,
      "step": 8
    },
    {
      "epoch": 0.2517482517482518,
      "grad_norm": 35.220550537109375,
      "learning_rate": 4.5e-06,
      "loss": 9.856,
      "step": 9
    },
    {
      "epoch": 0.27972027972027974,
      "grad_norm": 37.11204528808594,
      "learning_rate": 5e-06,
      "loss": 9.8554,
      "step": 10
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 41.45880126953125,
      "learning_rate": 4.99847706754774e-06,
      "loss": 10.418,
      "step": 11
    },
    {
      "epoch": 0.3356643356643357,
      "grad_norm": 40.01206970214844,
      "learning_rate": 4.993910125649561e-06,
      "loss": 10.3614,
      "step": 12
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 41.16290283203125,
      "learning_rate": 4.986304738420684e-06,
      "loss": 10.5427,
      "step": 13
    },
    {
      "epoch": 0.3916083916083916,
      "grad_norm": 44.843379974365234,
      "learning_rate": 4.975670171853926e-06,
      "loss": 10.5171,
      "step": 14
    },
    {
      "epoch": 0.4195804195804196,
      "grad_norm": 46.477989196777344,
      "learning_rate": 4.962019382530521e-06,
      "loss": 10.7146,
      "step": 15
    },
    {
      "epoch": 0.44755244755244755,
      "grad_norm": 50.21609115600586,
      "learning_rate": 4.9453690018345144e-06,
      "loss": 12.2147,
      "step": 16
    },
    {
      "epoch": 0.4755244755244755,
      "grad_norm": 25.672412872314453,
      "learning_rate": 4.925739315689991e-06,
      "loss": 8.7403,
      "step": 17
    },
    {
      "epoch": 0.5034965034965035,
      "grad_norm": 26.379234313964844,
      "learning_rate": 4.903154239845798e-06,
      "loss": 8.7188,
      "step": 18
    },
    {
      "epoch": 0.5314685314685315,
      "grad_norm": 27.213178634643555,
      "learning_rate": 4.8776412907378845e-06,
      "loss": 8.7172,
      "step": 19
    },
    {
      "epoch": 0.5594405594405595,
      "grad_norm": 26.93313217163086,
      "learning_rate": 4.849231551964771e-06,
      "loss": 8.7001,
      "step": 20
    },
    {
      "epoch": 0.5874125874125874,
      "grad_norm": 25.692018508911133,
      "learning_rate": 4.817959636416969e-06,
      "loss": 8.8583,
      "step": 21
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 29.379714965820312,
      "learning_rate": 4.783863644106502e-06,
      "loss": 9.2555,
      "step": 22
    },
    {
      "epoch": 0.6433566433566433,
      "grad_norm": 28.893835067749023,
      "learning_rate": 4.746985115747918e-06,
      "loss": 9.768,
      "step": 23
    },
    {
      "epoch": 0.6713286713286714,
      "grad_norm": 37.284706115722656,
      "learning_rate": 4.707368982147318e-06,
      "loss": 11.1789,
      "step": 24
    },
    {
      "epoch": 0.6993006993006993,
      "grad_norm": 18.487119674682617,
      "learning_rate": 4.665063509461098e-06,
      "loss": 7.7158,
      "step": 25
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 19.818843841552734,
      "learning_rate": 4.620120240391065e-06,
      "loss": 8.1177,
      "step": 26
    },
    {
      "epoch": 0.7552447552447552,
      "grad_norm": 21.235286712646484,
      "learning_rate": 4.572593931387604e-06,
      "loss": 8.4652,
      "step": 27
    },
    {
      "epoch": 0.7832167832167832,
      "grad_norm": 20.11646270751953,
      "learning_rate": 4.522542485937369e-06,
      "loss": 8.5836,
      "step": 28
    },
    {
      "epoch": 0.8111888111888111,
      "grad_norm": 24.192258834838867,
      "learning_rate": 4.470026884016805e-06,
      "loss": 8.669,
      "step": 29
    },
    {
      "epoch": 0.8391608391608392,
      "grad_norm": 21.79612159729004,
      "learning_rate": 4.415111107797445e-06,
      "loss": 8.9857,
      "step": 30
    },
    {
      "epoch": 0.8671328671328671,
      "grad_norm": 23.46041488647461,
      "learning_rate": 4.357862063693486e-06,
      "loss": 8.8214,
      "step": 31
    },
    {
      "epoch": 0.8951048951048951,
      "grad_norm": 32.813968658447266,
      "learning_rate": 4.2983495008466285e-06,
      "loss": 10.271,
      "step": 32
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 16.506298065185547,
      "learning_rate": 4.236645926147493e-06,
      "loss": 7.6578,
      "step": 33
    },
    {
      "epoch": 0.951048951048951,
      "grad_norm": 18.15751838684082,
      "learning_rate": 4.172826515897146e-06,
      "loss": 8.3214,
      "step": 34
    },
    {
      "epoch": 0.9790209790209791,
      "grad_norm": 21.8306884765625,
      "learning_rate": 4.106969024216348e-06,
      "loss": 8.4322,
      "step": 35
    },
    {
      "epoch": 1.006993006993007,
      "grad_norm": 27.69145965576172,
      "learning_rate": 4.039153688314146e-06,
      "loss": 9.1098,
      "step": 36
    },
    {
      "epoch": 1.034965034965035,
      "grad_norm": 15.433505058288574,
      "learning_rate": 3.969463130731183e-06,
      "loss": 7.4845,
      "step": 37
    },
    {
      "epoch": 1.062937062937063,
      "grad_norm": 15.924324989318848,
      "learning_rate": 3.897982258676867e-06,
      "loss": 7.397,
      "step": 38
    },
    {
      "epoch": 1.0909090909090908,
      "grad_norm": 15.257765769958496,
      "learning_rate": 3.824798160583012e-06,
      "loss": 7.9702,
      "step": 39
    },
    {
      "epoch": 1.118881118881119,
      "grad_norm": 15.685321807861328,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 7.8997,
      "step": 40
    },
    {
      "epoch": 1.1468531468531469,
      "grad_norm": 16.57723617553711,
      "learning_rate": 3.6736789069647273e-06,
      "loss": 8.1123,
      "step": 41
    },
    {
      "epoch": 1.1748251748251748,
      "grad_norm": 18.624141693115234,
      "learning_rate": 3.595927866972694e-06,
      "loss": 8.1582,
      "step": 42
    },
    {
      "epoch": 1.2027972027972027,
      "grad_norm": 18.69766616821289,
      "learning_rate": 3.516841607689501e-06,
      "loss": 8.0007,
      "step": 43
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 26.333940505981445,
      "learning_rate": 3.436516483539781e-06,
      "loss": 8.7443,
      "step": 44
    },
    {
      "epoch": 1.2587412587412588,
      "grad_norm": 13.822927474975586,
      "learning_rate": 3.3550503583141726e-06,
      "loss": 7.5874,
      "step": 45
    },
    {
      "epoch": 1.2867132867132867,
      "grad_norm": 14.370196342468262,
      "learning_rate": 3.272542485937369e-06,
      "loss": 7.3716,
      "step": 46
    },
    {
      "epoch": 1.3146853146853146,
      "grad_norm": 15.154383659362793,
      "learning_rate": 3.189093389542498e-06,
      "loss": 7.6832,
      "step": 47
    },
    {
      "epoch": 1.3426573426573427,
      "grad_norm": 16.412900924682617,
      "learning_rate": 3.1048047389991693e-06,
      "loss": 7.9016,
      "step": 48
    },
    {
      "epoch": 1.3706293706293706,
      "grad_norm": 16.24167251586914,
      "learning_rate": 3.019779227044398e-06,
      "loss": 7.6034,
      "step": 49
    },
    {
      "epoch": 1.3986013986013985,
      "grad_norm": 15.842921257019043,
      "learning_rate": 2.9341204441673267e-06,
      "loss": 7.9604,
      "step": 50
    },
    {
      "epoch": 1.3986013986013985,
      "eval_loss": 0.9812448024749756,
      "eval_runtime": 4.7027,
      "eval_samples_per_second": 12.971,
      "eval_steps_per_second": 3.402,
      "step": 50
    },
    {
      "epoch": 1.4265734265734267,
      "grad_norm": 19.437496185302734,
      "learning_rate": 2.847932752400164e-06,
      "loss": 7.9052,
      "step": 51
    },
    {
      "epoch": 1.4545454545454546,
      "grad_norm": 26.167152404785156,
      "learning_rate": 2.761321158169134e-06,
      "loss": 9.7995,
      "step": 52
    },
    {
      "epoch": 1.4825174825174825,
      "grad_norm": 13.2216157913208,
      "learning_rate": 2.6743911843603134e-06,
      "loss": 7.3562,
      "step": 53
    },
    {
      "epoch": 1.5104895104895104,
      "grad_norm": 15.14995002746582,
      "learning_rate": 2.587248741756253e-06,
      "loss": 7.6083,
      "step": 54
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 14.735421180725098,
      "learning_rate": 2.5e-06,
      "loss": 7.493,
      "step": 55
    },
    {
      "epoch": 1.5664335664335665,
      "grad_norm": 16.057226181030273,
      "learning_rate": 2.4127512582437486e-06,
      "loss": 7.7143,
      "step": 56
    },
    {
      "epoch": 1.5944055944055944,
      "grad_norm": 17.881492614746094,
      "learning_rate": 2.325608815639687e-06,
      "loss": 7.9213,
      "step": 57
    },
    {
      "epoch": 1.6223776223776225,
      "grad_norm": 18.69293785095215,
      "learning_rate": 2.238678841830867e-06,
      "loss": 8.0768,
      "step": 58
    },
    {
      "epoch": 1.6503496503496504,
      "grad_norm": 20.132923126220703,
      "learning_rate": 2.1520672475998374e-06,
      "loss": 8.4112,
      "step": 59
    },
    {
      "epoch": 1.6783216783216783,
      "grad_norm": 22.801490783691406,
      "learning_rate": 2.0658795558326745e-06,
      "loss": 8.7676,
      "step": 60
    },
    {
      "epoch": 1.7062937062937062,
      "grad_norm": 12.906396865844727,
      "learning_rate": 1.9802207729556023e-06,
      "loss": 7.2744,
      "step": 61
    },
    {
      "epoch": 1.7342657342657342,
      "grad_norm": 14.026016235351562,
      "learning_rate": 1.895195261000831e-06,
      "loss": 7.0761,
      "step": 62
    },
    {
      "epoch": 1.762237762237762,
      "grad_norm": 13.737709999084473,
      "learning_rate": 1.8109066104575023e-06,
      "loss": 7.4639,
      "step": 63
    },
    {
      "epoch": 1.7902097902097902,
      "grad_norm": 15.159075736999512,
      "learning_rate": 1.7274575140626318e-06,
      "loss": 7.1799,
      "step": 64
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 16.306058883666992,
      "learning_rate": 1.6449496416858285e-06,
      "loss": 7.3868,
      "step": 65
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 16.63066291809082,
      "learning_rate": 1.56348351646022e-06,
      "loss": 7.9422,
      "step": 66
    },
    {
      "epoch": 1.8741258741258742,
      "grad_norm": 19.77796173095703,
      "learning_rate": 1.4831583923105e-06,
      "loss": 8.3306,
      "step": 67
    },
    {
      "epoch": 1.902097902097902,
      "grad_norm": 23.449113845825195,
      "learning_rate": 1.4040721330273063e-06,
      "loss": 8.1032,
      "step": 68
    },
    {
      "epoch": 1.93006993006993,
      "grad_norm": 13.089604377746582,
      "learning_rate": 1.3263210930352737e-06,
      "loss": 6.8976,
      "step": 69
    },
    {
      "epoch": 1.958041958041958,
      "grad_norm": 15.323660850524902,
      "learning_rate": 1.2500000000000007e-06,
      "loss": 7.1423,
      "step": 70
    },
    {
      "epoch": 1.986013986013986,
      "grad_norm": 17.456003189086914,
      "learning_rate": 1.1752018394169882e-06,
      "loss": 8.0554,
      "step": 71
    },
    {
      "epoch": 2.013986013986014,
      "grad_norm": 17.351194381713867,
      "learning_rate": 1.1020177413231334e-06,
      "loss": 7.5417,
      "step": 72
    },
    {
      "epoch": 2.041958041958042,
      "grad_norm": 12.989124298095703,
      "learning_rate": 1.0305368692688175e-06,
      "loss": 6.9213,
      "step": 73
    },
    {
      "epoch": 2.06993006993007,
      "grad_norm": 13.786518096923828,
      "learning_rate": 9.608463116858544e-07,
      "loss": 6.993,
      "step": 74
    },
    {
      "epoch": 2.097902097902098,
      "grad_norm": 14.11694622039795,
      "learning_rate": 8.930309757836517e-07,
      "loss": 7.277,
      "step": 75
    },
    {
      "epoch": 2.125874125874126,
      "grad_norm": 15.72767162322998,
      "learning_rate": 8.271734841028553e-07,
      "loss": 7.5282,
      "step": 76
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 17.46666717529297,
      "learning_rate": 7.633540738525066e-07,
      "loss": 7.2148,
      "step": 77
    },
    {
      "epoch": 2.1818181818181817,
      "grad_norm": 17.60526466369629,
      "learning_rate": 7.016504991533727e-07,
      "loss": 7.6238,
      "step": 78
    },
    {
      "epoch": 2.20979020979021,
      "grad_norm": 18.603511810302734,
      "learning_rate": 6.421379363065142e-07,
      "loss": 8.0623,
      "step": 79
    },
    {
      "epoch": 2.237762237762238,
      "grad_norm": 18.113256454467773,
      "learning_rate": 5.848888922025553e-07,
      "loss": 7.6123,
      "step": 80
    },
    {
      "epoch": 2.265734265734266,
      "grad_norm": 12.54310417175293,
      "learning_rate": 5.299731159831953e-07,
      "loss": 7.2639,
      "step": 81
    },
    {
      "epoch": 2.2937062937062938,
      "grad_norm": 14.775126457214355,
      "learning_rate": 4.774575140626317e-07,
      "loss": 6.9113,
      "step": 82
    },
    {
      "epoch": 2.3216783216783217,
      "grad_norm": 13.778559684753418,
      "learning_rate": 4.27406068612396e-07,
      "loss": 7.214,
      "step": 83
    },
    {
      "epoch": 2.3496503496503496,
      "grad_norm": 15.231853485107422,
      "learning_rate": 3.798797596089351e-07,
      "loss": 7.3562,
      "step": 84
    },
    {
      "epoch": 2.3776223776223775,
      "grad_norm": 16.52434539794922,
      "learning_rate": 3.3493649053890325e-07,
      "loss": 7.6841,
      "step": 85
    },
    {
      "epoch": 2.4055944055944054,
      "grad_norm": 16.546457290649414,
      "learning_rate": 2.9263101785268253e-07,
      "loss": 7.2925,
      "step": 86
    },
    {
      "epoch": 2.4335664335664333,
      "grad_norm": 20.887779235839844,
      "learning_rate": 2.53014884252083e-07,
      "loss": 8.0049,
      "step": 87
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 24.01591682434082,
      "learning_rate": 2.1613635589349756e-07,
      "loss": 8.0908,
      "step": 88
    },
    {
      "epoch": 2.4895104895104896,
      "grad_norm": 13.272303581237793,
      "learning_rate": 1.8204036358303173e-07,
      "loss": 6.8586,
      "step": 89
    },
    {
      "epoch": 2.5174825174825175,
      "grad_norm": 13.375730514526367,
      "learning_rate": 1.507684480352292e-07,
      "loss": 6.9535,
      "step": 90
    },
    {
      "epoch": 2.5454545454545454,
      "grad_norm": 15.659381866455078,
      "learning_rate": 1.223587092621162e-07,
      "loss": 7.2084,
      "step": 91
    },
    {
      "epoch": 2.5734265734265733,
      "grad_norm": 15.646546363830566,
      "learning_rate": 9.684576015420277e-08,
      "loss": 6.8886,
      "step": 92
    },
    {
      "epoch": 2.6013986013986012,
      "grad_norm": 17.0341796875,
      "learning_rate": 7.426068431000883e-08,
      "loss": 7.3088,
      "step": 93
    },
    {
      "epoch": 2.629370629370629,
      "grad_norm": 17.291650772094727,
      "learning_rate": 5.463099816548578e-08,
      "loss": 7.3454,
      "step": 94
    },
    {
      "epoch": 2.6573426573426575,
      "grad_norm": 18.05946922302246,
      "learning_rate": 3.798061746947995e-08,
      "loss": 7.7567,
      "step": 95
    },
    {
      "epoch": 2.6853146853146854,
      "grad_norm": 16.397621154785156,
      "learning_rate": 2.4329828146074096e-08,
      "loss": 7.8379,
      "step": 96
    },
    {
      "epoch": 2.7132867132867133,
      "grad_norm": 14.185662269592285,
      "learning_rate": 1.3695261579316776e-08,
      "loss": 6.869,
      "step": 97
    },
    {
      "epoch": 2.7412587412587412,
      "grad_norm": 13.866539001464844,
      "learning_rate": 6.089874350439507e-09,
      "loss": 6.8378,
      "step": 98
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 15.03514575958252,
      "learning_rate": 1.5229324522605949e-09,
      "loss": 7.0574,
      "step": 99
    },
    {
      "epoch": 2.797202797202797,
      "grad_norm": 15.289778709411621,
      "learning_rate": 0.0,
      "loss": 7.0866,
      "step": 100
    },
    {
      "epoch": 2.797202797202797,
      "eval_loss": 0.9307088255882263,
      "eval_runtime": 4.6988,
      "eval_samples_per_second": 12.982,
      "eval_steps_per_second": 3.405,
      "step": 100
    }
  ],
  "logging_steps": 1,
  "max_steps": 100,
  "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": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.4053816950954394e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}