{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.961038961038961,
  "eval_steps": 500,
  "global_step": 114,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 5e-06,
      "loss": 8.853,
      "step": 1
    },
    {
      "epoch": 0.05,
      "learning_rate": 1e-05,
      "loss": 8.3278,
      "step": 2
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 5.9423,
      "step": 3
    },
    {
      "epoch": 0.1,
      "learning_rate": 2e-05,
      "loss": 2.6977,
      "step": 4
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.9995921928281893e-05,
      "loss": 2.0802,
      "step": 5
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.9983691039261358e-05,
      "loss": 1.6523,
      "step": 6
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.9963317308626916e-05,
      "loss": 1.0093,
      "step": 7
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.99348173534855e-05,
      "loss": 1.2197,
      "step": 8
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.989821441880933e-05,
      "loss": 0.9122,
      "step": 9
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.9853538358476933e-05,
      "loss": 0.7479,
      "step": 10
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.9800825610923937e-05,
      "loss": 0.927,
      "step": 11
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.9740119169423337e-05,
      "loss": 0.5945,
      "step": 12
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.9671468547019575e-05,
      "loss": 0.6373,
      "step": 13
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.9594929736144978e-05,
      "loss": 0.5594,
      "step": 14
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.9510565162951538e-05,
      "loss": 0.61,
      "step": 15
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.941844363639525e-05,
      "loss": 0.5487,
      "step": 16
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.9318640292114526e-05,
      "loss": 0.5727,
      "step": 17
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.92112365311485e-05,
      "loss": 0.4994,
      "step": 18
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.9096319953545186e-05,
      "loss": 0.6802,
      "step": 19
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.8973984286913584e-05,
      "loss": 0.5732,
      "step": 20
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.8844329309978146e-05,
      "loss": 0.4517,
      "step": 21
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.8707460771197773e-05,
      "loss": 0.4944,
      "step": 22
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.856349030251589e-05,
      "loss": 0.4534,
      "step": 23
    },
    {
      "epoch": 0.62,
      "learning_rate": 1.8412535328311813e-05,
      "loss": 0.4643,
      "step": 24
    },
    {
      "epoch": 0.65,
      "learning_rate": 1.825471896962774e-05,
      "loss": 0.3858,
      "step": 25
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 0.3925,
      "step": 26
    },
    {
      "epoch": 0.7,
      "learning_rate": 1.7919022459222754e-05,
      "loss": 0.4107,
      "step": 27
    },
    {
      "epoch": 0.73,
      "learning_rate": 1.7741416106390828e-05,
      "loss": 0.4385,
      "step": 28
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.7557495743542586e-05,
      "loss": 0.3052,
      "step": 29
    },
    {
      "epoch": 0.78,
      "learning_rate": 1.736741137876405e-05,
      "loss": 0.2988,
      "step": 30
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.7171318047589637e-05,
      "loss": 0.3424,
      "step": 31
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.696937568655294e-05,
      "loss": 0.4287,
      "step": 32
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.6761749002740195e-05,
      "loss": 0.394,
      "step": 33
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.6548607339452853e-05,
      "loss": 0.4674,
      "step": 34
    },
    {
      "epoch": 0.91,
      "learning_rate": 1.6330124538088705e-05,
      "loss": 0.3405,
      "step": 35
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.6106478796354382e-05,
      "loss": 0.3606,
      "step": 36
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.5877852522924733e-05,
      "loss": 0.2745,
      "step": 37
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.5644432188667695e-05,
      "loss": 0.3599,
      "step": 38
    },
    {
      "epoch": 1.01,
      "learning_rate": 1.5406408174555978e-05,
      "loss": 0.2998,
      "step": 39
    },
    {
      "epoch": 1.04,
      "learning_rate": 1.5163974616389621e-05,
      "loss": 0.2837,
      "step": 40
    },
    {
      "epoch": 1.06,
      "learning_rate": 1.491732924645604e-05,
      "loss": 0.204,
      "step": 41
    },
    {
      "epoch": 1.09,
      "learning_rate": 1.4666673232256738e-05,
      "loss": 0.2637,
      "step": 42
    },
    {
      "epoch": 1.12,
      "learning_rate": 1.4412211012432213e-05,
      "loss": 0.2783,
      "step": 43
    },
    {
      "epoch": 1.14,
      "learning_rate": 1.4154150130018867e-05,
      "loss": 0.2416,
      "step": 44
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.3892701063173917e-05,
      "loss": 0.2795,
      "step": 45
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.362807705350641e-05,
      "loss": 0.2432,
      "step": 46
    },
    {
      "epoch": 1.22,
      "learning_rate": 1.3360493932154301e-05,
      "loss": 0.2224,
      "step": 47
    },
    {
      "epoch": 1.25,
      "learning_rate": 1.3090169943749475e-05,
      "loss": 0.2528,
      "step": 48
    },
    {
      "epoch": 1.27,
      "learning_rate": 1.2817325568414299e-05,
      "loss": 0.2461,
      "step": 49
    },
    {
      "epoch": 1.3,
      "learning_rate": 1.2542183341934873e-05,
      "loss": 0.2683,
      "step": 50
    },
    {
      "epoch": 1.32,
      "learning_rate": 1.2264967674257647e-05,
      "loss": 0.2396,
      "step": 51
    },
    {
      "epoch": 1.35,
      "learning_rate": 1.1985904666457455e-05,
      "loss": 0.2429,
      "step": 52
    },
    {
      "epoch": 1.38,
      "learning_rate": 1.170522192632624e-05,
      "loss": 0.2568,
      "step": 53
    },
    {
      "epoch": 1.4,
      "learning_rate": 1.1423148382732854e-05,
      "loss": 0.2612,
      "step": 54
    },
    {
      "epoch": 1.43,
      "learning_rate": 1.1139914098905406e-05,
      "loss": 0.2221,
      "step": 55
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.08557500847884e-05,
      "loss": 0.2526,
      "step": 56
    },
    {
      "epoch": 1.48,
      "learning_rate": 1.0570888108627682e-05,
      "loss": 0.2413,
      "step": 57
    },
    {
      "epoch": 1.51,
      "learning_rate": 1.0285560507936962e-05,
      "loss": 0.1982,
      "step": 58
    },
    {
      "epoch": 1.53,
      "learning_rate": 1e-05,
      "loss": 0.1877,
      "step": 59
    },
    {
      "epoch": 1.56,
      "learning_rate": 9.71443949206304e-06,
      "loss": 0.2348,
      "step": 60
    },
    {
      "epoch": 1.58,
      "learning_rate": 9.42911189137232e-06,
      "loss": 0.2131,
      "step": 61
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.144249915211605e-06,
      "loss": 0.2035,
      "step": 62
    },
    {
      "epoch": 1.64,
      "learning_rate": 8.860085901094595e-06,
      "loss": 0.2275,
      "step": 63
    },
    {
      "epoch": 1.66,
      "learning_rate": 8.576851617267151e-06,
      "loss": 0.2132,
      "step": 64
    },
    {
      "epoch": 1.69,
      "learning_rate": 8.294778073673762e-06,
      "loss": 0.2492,
      "step": 65
    },
    {
      "epoch": 1.71,
      "learning_rate": 8.014095333542548e-06,
      "loss": 0.2203,
      "step": 66
    },
    {
      "epoch": 1.74,
      "learning_rate": 7.735032325742355e-06,
      "loss": 0.212,
      "step": 67
    },
    {
      "epoch": 1.77,
      "learning_rate": 7.4578166580651335e-06,
      "loss": 0.2255,
      "step": 68
    },
    {
      "epoch": 1.79,
      "learning_rate": 7.182674431585703e-06,
      "loss": 0.2004,
      "step": 69
    },
    {
      "epoch": 1.82,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.2618,
      "step": 70
    },
    {
      "epoch": 1.84,
      "learning_rate": 6.639506067845698e-06,
      "loss": 0.205,
      "step": 71
    },
    {
      "epoch": 1.87,
      "learning_rate": 6.3719229464935915e-06,
      "loss": 0.1988,
      "step": 72
    },
    {
      "epoch": 1.9,
      "learning_rate": 6.107298936826086e-06,
      "loss": 0.2342,
      "step": 73
    },
    {
      "epoch": 1.92,
      "learning_rate": 5.845849869981137e-06,
      "loss": 0.2351,
      "step": 74
    },
    {
      "epoch": 1.95,
      "learning_rate": 5.587788987567785e-06,
      "loss": 0.2241,
      "step": 75
    },
    {
      "epoch": 1.97,
      "learning_rate": 5.333326767743263e-06,
      "loss": 0.2031,
      "step": 76
    },
    {
      "epoch": 2.0,
      "learning_rate": 5.082670753543961e-06,
      "loss": 0.1787,
      "step": 77
    },
    {
      "epoch": 2.03,
      "learning_rate": 4.836025383610382e-06,
      "loss": 0.1162,
      "step": 78
    },
    {
      "epoch": 2.05,
      "learning_rate": 4.593591825444028e-06,
      "loss": 0.1533,
      "step": 79
    },
    {
      "epoch": 2.08,
      "learning_rate": 4.355567811332311e-06,
      "loss": 0.1156,
      "step": 80
    },
    {
      "epoch": 2.1,
      "learning_rate": 4.12214747707527e-06,
      "loss": 0.1691,
      "step": 81
    },
    {
      "epoch": 2.13,
      "learning_rate": 3.893521203645618e-06,
      "loss": 0.1713,
      "step": 82
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.6698754619112974e-06,
      "loss": 0.1608,
      "step": 83
    },
    {
      "epoch": 2.18,
      "learning_rate": 3.4513926605471504e-06,
      "loss": 0.1553,
      "step": 84
    },
    {
      "epoch": 2.21,
      "learning_rate": 3.2382509972598087e-06,
      "loss": 0.1298,
      "step": 85
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.0306243134470668e-06,
      "loss": 0.1794,
      "step": 86
    },
    {
      "epoch": 2.26,
      "learning_rate": 2.8286819524103657e-06,
      "loss": 0.1235,
      "step": 87
    },
    {
      "epoch": 2.29,
      "learning_rate": 2.6325886212359496e-06,
      "loss": 0.1465,
      "step": 88
    },
    {
      "epoch": 2.31,
      "learning_rate": 2.4425042564574186e-06,
      "loss": 0.1605,
      "step": 89
    },
    {
      "epoch": 2.34,
      "learning_rate": 2.2585838936091753e-06,
      "loss": 0.1318,
      "step": 90
    },
    {
      "epoch": 2.36,
      "learning_rate": 2.0809775407772505e-06,
      "loss": 0.158,
      "step": 91
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.9098300562505266e-06,
      "loss": 0.1413,
      "step": 92
    },
    {
      "epoch": 2.42,
      "learning_rate": 1.74528103037226e-06,
      "loss": 0.2054,
      "step": 93
    },
    {
      "epoch": 2.44,
      "learning_rate": 1.587464671688187e-06,
      "loss": 0.1365,
      "step": 94
    },
    {
      "epoch": 2.47,
      "learning_rate": 1.436509697484111e-06,
      "loss": 0.133,
      "step": 95
    },
    {
      "epoch": 2.49,
      "learning_rate": 1.2925392288022299e-06,
      "loss": 0.154,
      "step": 96
    },
    {
      "epoch": 2.52,
      "learning_rate": 1.1556706900218572e-06,
      "loss": 0.191,
      "step": 97
    },
    {
      "epoch": 2.55,
      "learning_rate": 1.0260157130864178e-06,
      "loss": 0.1685,
      "step": 98
    },
    {
      "epoch": 2.57,
      "learning_rate": 9.036800464548157e-07,
      "loss": 0.1512,
      "step": 99
    },
    {
      "epoch": 2.6,
      "learning_rate": 7.887634688515e-07,
      "loss": 0.1472,
      "step": 100
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.813597078854772e-07,
      "loss": 0.1433,
      "step": 101
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.815563636047539e-07,
      "loss": 0.1774,
      "step": 102
    },
    {
      "epoch": 2.68,
      "learning_rate": 4.894348370484648e-07,
      "loss": 0.161,
      "step": 103
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.0507026385502747e-07,
      "loss": 0.1705,
      "step": 104
    },
    {
      "epoch": 2.73,
      "learning_rate": 3.2853145298042954e-07,
      "loss": 0.1515,
      "step": 105
    },
    {
      "epoch": 2.75,
      "learning_rate": 2.5988083057666534e-07,
      "loss": 0.165,
      "step": 106
    },
    {
      "epoch": 2.78,
      "learning_rate": 1.9917438907606556e-07,
      "loss": 0.1492,
      "step": 107
    },
    {
      "epoch": 2.81,
      "learning_rate": 1.464616415230702e-07,
      "loss": 0.1548,
      "step": 108
    },
    {
      "epoch": 2.83,
      "learning_rate": 1.0178558119067316e-07,
      "loss": 0.1562,
      "step": 109
    },
    {
      "epoch": 2.86,
      "learning_rate": 6.51826465144978e-08,
      "loss": 0.1532,
      "step": 110
    },
    {
      "epoch": 2.88,
      "learning_rate": 3.668269137308666e-08,
      "loss": 0.1501,
      "step": 111
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.630896073864352e-08,
      "loss": 0.1639,
      "step": 112
    },
    {
      "epoch": 2.94,
      "learning_rate": 4.0780717181077015e-09,
      "loss": 0.1539,
      "step": 113
    },
    {
      "epoch": 2.96,
      "learning_rate": 0.0,
      "loss": 0.1866,
      "step": 114
    },
    {
      "epoch": 2.96,
      "step": 114,
      "total_flos": 118111256969216.0,
      "train_loss": 0.5400874392505277,
      "train_runtime": 1384.0957,
      "train_samples_per_second": 21.317,
      "train_steps_per_second": 0.082
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 114,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "total_flos": 118111256969216.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}