{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 41.666666666666664,
  "eval_steps": 500,
  "global_step": 1500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.28,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 13.9056,
      "step": 10
    },
    {
      "epoch": 0.56,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 14.53,
      "step": 20
    },
    {
      "epoch": 0.83,
      "learning_rate": 3e-06,
      "loss": 14.6503,
      "step": 30
    },
    {
      "epoch": 1.11,
      "learning_rate": 4.000000000000001e-06,
      "loss": 13.8701,
      "step": 40
    },
    {
      "epoch": 1.39,
      "learning_rate": 5e-06,
      "loss": 12.7866,
      "step": 50
    },
    {
      "epoch": 1.67,
      "learning_rate": 6e-06,
      "loss": 12.0329,
      "step": 60
    },
    {
      "epoch": 1.94,
      "learning_rate": 7.000000000000001e-06,
      "loss": 11.6797,
      "step": 70
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.000000000000001e-06,
      "loss": 10.1652,
      "step": 80
    },
    {
      "epoch": 2.5,
      "learning_rate": 9e-06,
      "loss": 8.5016,
      "step": 90
    },
    {
      "epoch": 2.78,
      "learning_rate": 1e-05,
      "loss": 7.3051,
      "step": 100
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 6.5266,
      "step": 110
    },
    {
      "epoch": 3.33,
      "learning_rate": 1.2e-05,
      "loss": 4.9159,
      "step": 120
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 4.1691,
      "step": 130
    },
    {
      "epoch": 3.89,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 2.9038,
      "step": 140
    },
    {
      "epoch": 4.17,
      "learning_rate": 1.5e-05,
      "loss": 2.2194,
      "step": 150
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.97,
      "step": 160
    },
    {
      "epoch": 4.72,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 1.4839,
      "step": 170
    },
    {
      "epoch": 5.0,
      "learning_rate": 1.8e-05,
      "loss": 1.2807,
      "step": 180
    },
    {
      "epoch": 5.28,
      "learning_rate": 1.9e-05,
      "loss": 1.0124,
      "step": 190
    },
    {
      "epoch": 5.56,
      "learning_rate": 2e-05,
      "loss": 0.6787,
      "step": 200
    },
    {
      "epoch": 5.83,
      "learning_rate": 2.1e-05,
      "loss": 0.5985,
      "step": 210
    },
    {
      "epoch": 6.11,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.463,
      "step": 220
    },
    {
      "epoch": 6.39,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.4044,
      "step": 230
    },
    {
      "epoch": 6.67,
      "learning_rate": 2.4e-05,
      "loss": 0.3232,
      "step": 240
    },
    {
      "epoch": 6.94,
      "learning_rate": 2.5e-05,
      "loss": 0.3017,
      "step": 250
    },
    {
      "epoch": 7.22,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.2714,
      "step": 260
    },
    {
      "epoch": 7.5,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.2657,
      "step": 270
    },
    {
      "epoch": 7.78,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.1915,
      "step": 280
    },
    {
      "epoch": 8.06,
      "learning_rate": 2.9e-05,
      "loss": 0.1732,
      "step": 290
    },
    {
      "epoch": 8.33,
      "learning_rate": 3e-05,
      "loss": 0.1623,
      "step": 300
    },
    {
      "epoch": 8.61,
      "learning_rate": 3.1e-05,
      "loss": 0.1401,
      "step": 310
    },
    {
      "epoch": 8.89,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.1644,
      "step": 320
    },
    {
      "epoch": 9.17,
      "learning_rate": 3.3e-05,
      "loss": 0.1303,
      "step": 330
    },
    {
      "epoch": 9.44,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.1083,
      "step": 340
    },
    {
      "epoch": 9.72,
      "learning_rate": 3.5e-05,
      "loss": 0.1249,
      "step": 350
    },
    {
      "epoch": 10.0,
      "learning_rate": 3.6e-05,
      "loss": 0.1037,
      "step": 360
    },
    {
      "epoch": 10.28,
      "learning_rate": 3.7e-05,
      "loss": 0.1048,
      "step": 370
    },
    {
      "epoch": 10.56,
      "learning_rate": 3.8e-05,
      "loss": 0.0925,
      "step": 380
    },
    {
      "epoch": 10.83,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.0976,
      "step": 390
    },
    {
      "epoch": 11.11,
      "learning_rate": 4e-05,
      "loss": 0.0796,
      "step": 400
    },
    {
      "epoch": 11.39,
      "learning_rate": 4.1e-05,
      "loss": 0.0811,
      "step": 410
    },
    {
      "epoch": 11.67,
      "learning_rate": 4.2e-05,
      "loss": 0.0789,
      "step": 420
    },
    {
      "epoch": 11.94,
      "learning_rate": 4.3e-05,
      "loss": 0.0788,
      "step": 430
    },
    {
      "epoch": 12.22,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0674,
      "step": 440
    },
    {
      "epoch": 12.5,
      "learning_rate": 4.5e-05,
      "loss": 0.0632,
      "step": 450
    },
    {
      "epoch": 12.78,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.0779,
      "step": 460
    },
    {
      "epoch": 13.06,
      "learning_rate": 4.7e-05,
      "loss": 0.0596,
      "step": 470
    },
    {
      "epoch": 13.33,
      "learning_rate": 4.8e-05,
      "loss": 0.0568,
      "step": 480
    },
    {
      "epoch": 13.61,
      "learning_rate": 4.9e-05,
      "loss": 0.0647,
      "step": 490
    },
    {
      "epoch": 13.89,
      "learning_rate": 5e-05,
      "loss": 0.0574,
      "step": 500
    },
    {
      "epoch": 14.17,
      "learning_rate": 4.961538461538462e-05,
      "loss": 0.0584,
      "step": 510
    },
    {
      "epoch": 14.44,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.0468,
      "step": 520
    },
    {
      "epoch": 14.72,
      "learning_rate": 4.884615384615385e-05,
      "loss": 0.055,
      "step": 530
    },
    {
      "epoch": 15.0,
      "learning_rate": 4.846153846153846e-05,
      "loss": 0.05,
      "step": 540
    },
    {
      "epoch": 15.28,
      "learning_rate": 4.8076923076923084e-05,
      "loss": 0.0502,
      "step": 550
    },
    {
      "epoch": 15.56,
      "learning_rate": 4.76923076923077e-05,
      "loss": 0.048,
      "step": 560
    },
    {
      "epoch": 15.83,
      "learning_rate": 4.730769230769231e-05,
      "loss": 0.0465,
      "step": 570
    },
    {
      "epoch": 16.11,
      "learning_rate": 4.692307692307693e-05,
      "loss": 0.0477,
      "step": 580
    },
    {
      "epoch": 16.39,
      "learning_rate": 4.653846153846154e-05,
      "loss": 0.0451,
      "step": 590
    },
    {
      "epoch": 16.67,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0345,
      "step": 600
    },
    {
      "epoch": 16.94,
      "learning_rate": 4.576923076923077e-05,
      "loss": 0.0443,
      "step": 610
    },
    {
      "epoch": 17.22,
      "learning_rate": 4.538461538461539e-05,
      "loss": 0.034,
      "step": 620
    },
    {
      "epoch": 17.5,
      "learning_rate": 4.5e-05,
      "loss": 0.0422,
      "step": 630
    },
    {
      "epoch": 17.78,
      "learning_rate": 4.461538461538462e-05,
      "loss": 0.0385,
      "step": 640
    },
    {
      "epoch": 18.06,
      "learning_rate": 4.423076923076923e-05,
      "loss": 0.0339,
      "step": 650
    },
    {
      "epoch": 18.33,
      "learning_rate": 4.384615384615385e-05,
      "loss": 0.0322,
      "step": 660
    },
    {
      "epoch": 18.61,
      "learning_rate": 4.346153846153846e-05,
      "loss": 0.0319,
      "step": 670
    },
    {
      "epoch": 18.89,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0358,
      "step": 680
    },
    {
      "epoch": 19.17,
      "learning_rate": 4.269230769230769e-05,
      "loss": 0.0365,
      "step": 690
    },
    {
      "epoch": 19.44,
      "learning_rate": 4.230769230769231e-05,
      "loss": 0.0349,
      "step": 700
    },
    {
      "epoch": 19.72,
      "learning_rate": 4.192307692307693e-05,
      "loss": 0.034,
      "step": 710
    },
    {
      "epoch": 20.0,
      "learning_rate": 4.1538461538461544e-05,
      "loss": 0.031,
      "step": 720
    },
    {
      "epoch": 20.28,
      "learning_rate": 4.115384615384615e-05,
      "loss": 0.0394,
      "step": 730
    },
    {
      "epoch": 20.56,
      "learning_rate": 4.0769230769230773e-05,
      "loss": 0.0266,
      "step": 740
    },
    {
      "epoch": 20.83,
      "learning_rate": 4.038461538461539e-05,
      "loss": 0.0274,
      "step": 750
    },
    {
      "epoch": 21.11,
      "learning_rate": 4e-05,
      "loss": 0.0227,
      "step": 760
    },
    {
      "epoch": 21.39,
      "learning_rate": 3.961538461538462e-05,
      "loss": 0.0299,
      "step": 770
    },
    {
      "epoch": 21.67,
      "learning_rate": 3.923076923076923e-05,
      "loss": 0.0262,
      "step": 780
    },
    {
      "epoch": 21.94,
      "learning_rate": 3.884615384615385e-05,
      "loss": 0.0218,
      "step": 790
    },
    {
      "epoch": 22.22,
      "learning_rate": 3.846153846153846e-05,
      "loss": 0.0253,
      "step": 800
    },
    {
      "epoch": 22.5,
      "learning_rate": 3.807692307692308e-05,
      "loss": 0.0336,
      "step": 810
    },
    {
      "epoch": 22.78,
      "learning_rate": 3.769230769230769e-05,
      "loss": 0.0288,
      "step": 820
    },
    {
      "epoch": 23.06,
      "learning_rate": 3.730769230769231e-05,
      "loss": 0.0233,
      "step": 830
    },
    {
      "epoch": 23.33,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0276,
      "step": 840
    },
    {
      "epoch": 23.61,
      "learning_rate": 3.653846153846154e-05,
      "loss": 0.0184,
      "step": 850
    },
    {
      "epoch": 23.89,
      "learning_rate": 3.615384615384615e-05,
      "loss": 0.029,
      "step": 860
    },
    {
      "epoch": 24.17,
      "learning_rate": 3.5769230769230774e-05,
      "loss": 0.0197,
      "step": 870
    },
    {
      "epoch": 24.44,
      "learning_rate": 3.538461538461539e-05,
      "loss": 0.0224,
      "step": 880
    },
    {
      "epoch": 24.72,
      "learning_rate": 3.5e-05,
      "loss": 0.0248,
      "step": 890
    },
    {
      "epoch": 25.0,
      "learning_rate": 3.461538461538462e-05,
      "loss": 0.0286,
      "step": 900
    },
    {
      "epoch": 25.28,
      "learning_rate": 3.4230769230769234e-05,
      "loss": 0.0241,
      "step": 910
    },
    {
      "epoch": 25.56,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0217,
      "step": 920
    },
    {
      "epoch": 25.83,
      "learning_rate": 3.346153846153846e-05,
      "loss": 0.021,
      "step": 930
    },
    {
      "epoch": 26.11,
      "learning_rate": 3.307692307692308e-05,
      "loss": 0.0229,
      "step": 940
    },
    {
      "epoch": 26.39,
      "learning_rate": 3.269230769230769e-05,
      "loss": 0.0246,
      "step": 950
    },
    {
      "epoch": 26.67,
      "learning_rate": 3.230769230769231e-05,
      "loss": 0.0238,
      "step": 960
    },
    {
      "epoch": 26.94,
      "learning_rate": 3.192307692307692e-05,
      "loss": 0.0216,
      "step": 970
    },
    {
      "epoch": 27.22,
      "learning_rate": 3.153846153846154e-05,
      "loss": 0.0257,
      "step": 980
    },
    {
      "epoch": 27.5,
      "learning_rate": 3.115384615384615e-05,
      "loss": 0.0193,
      "step": 990
    },
    {
      "epoch": 27.78,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.017,
      "step": 1000
    },
    {
      "epoch": 28.06,
      "learning_rate": 3.0384615384615382e-05,
      "loss": 0.0176,
      "step": 1010
    },
    {
      "epoch": 28.33,
      "learning_rate": 3e-05,
      "loss": 0.0227,
      "step": 1020
    },
    {
      "epoch": 28.61,
      "learning_rate": 2.9615384615384616e-05,
      "loss": 0.0185,
      "step": 1030
    },
    {
      "epoch": 28.89,
      "learning_rate": 2.9230769230769234e-05,
      "loss": 0.0206,
      "step": 1040
    },
    {
      "epoch": 29.17,
      "learning_rate": 2.8846153846153845e-05,
      "loss": 0.0194,
      "step": 1050
    },
    {
      "epoch": 29.44,
      "learning_rate": 2.846153846153846e-05,
      "loss": 0.0229,
      "step": 1060
    },
    {
      "epoch": 29.72,
      "learning_rate": 2.807692307692308e-05,
      "loss": 0.0186,
      "step": 1070
    },
    {
      "epoch": 30.0,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.015,
      "step": 1080
    },
    {
      "epoch": 30.28,
      "learning_rate": 2.7307692307692305e-05,
      "loss": 0.0198,
      "step": 1090
    },
    {
      "epoch": 30.56,
      "learning_rate": 2.6923076923076923e-05,
      "loss": 0.0162,
      "step": 1100
    },
    {
      "epoch": 30.83,
      "learning_rate": 2.6538461538461538e-05,
      "loss": 0.0163,
      "step": 1110
    },
    {
      "epoch": 31.11,
      "learning_rate": 2.6153846153846157e-05,
      "loss": 0.0257,
      "step": 1120
    },
    {
      "epoch": 31.39,
      "learning_rate": 2.5769230769230768e-05,
      "loss": 0.017,
      "step": 1130
    },
    {
      "epoch": 31.67,
      "learning_rate": 2.5384615384615383e-05,
      "loss": 0.0131,
      "step": 1140
    },
    {
      "epoch": 31.94,
      "learning_rate": 2.5e-05,
      "loss": 0.0224,
      "step": 1150
    },
    {
      "epoch": 32.22,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.0218,
      "step": 1160
    },
    {
      "epoch": 32.5,
      "learning_rate": 2.423076923076923e-05,
      "loss": 0.014,
      "step": 1170
    },
    {
      "epoch": 32.78,
      "learning_rate": 2.384615384615385e-05,
      "loss": 0.0165,
      "step": 1180
    },
    {
      "epoch": 33.06,
      "learning_rate": 2.3461538461538464e-05,
      "loss": 0.017,
      "step": 1190
    },
    {
      "epoch": 33.33,
      "learning_rate": 2.307692307692308e-05,
      "loss": 0.0177,
      "step": 1200
    },
    {
      "epoch": 33.61,
      "learning_rate": 2.2692307692307694e-05,
      "loss": 0.015,
      "step": 1210
    },
    {
      "epoch": 33.89,
      "learning_rate": 2.230769230769231e-05,
      "loss": 0.0185,
      "step": 1220
    },
    {
      "epoch": 34.17,
      "learning_rate": 2.1923076923076924e-05,
      "loss": 0.0201,
      "step": 1230
    },
    {
      "epoch": 34.44,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.0134,
      "step": 1240
    },
    {
      "epoch": 34.72,
      "learning_rate": 2.1153846153846154e-05,
      "loss": 0.019,
      "step": 1250
    },
    {
      "epoch": 35.0,
      "learning_rate": 2.0769230769230772e-05,
      "loss": 0.0156,
      "step": 1260
    },
    {
      "epoch": 35.28,
      "learning_rate": 2.0384615384615387e-05,
      "loss": 0.0171,
      "step": 1270
    },
    {
      "epoch": 35.56,
      "learning_rate": 2e-05,
      "loss": 0.0114,
      "step": 1280
    },
    {
      "epoch": 35.83,
      "learning_rate": 1.9615384615384617e-05,
      "loss": 0.0208,
      "step": 1290
    },
    {
      "epoch": 36.11,
      "learning_rate": 1.923076923076923e-05,
      "loss": 0.0145,
      "step": 1300
    },
    {
      "epoch": 36.39,
      "learning_rate": 1.8846153846153846e-05,
      "loss": 0.0124,
      "step": 1310
    },
    {
      "epoch": 36.67,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.0197,
      "step": 1320
    },
    {
      "epoch": 36.94,
      "learning_rate": 1.8076923076923076e-05,
      "loss": 0.018,
      "step": 1330
    },
    {
      "epoch": 37.22,
      "learning_rate": 1.7692307692307694e-05,
      "loss": 0.0218,
      "step": 1340
    },
    {
      "epoch": 37.5,
      "learning_rate": 1.730769230769231e-05,
      "loss": 0.0111,
      "step": 1350
    },
    {
      "epoch": 37.78,
      "learning_rate": 1.6923076923076924e-05,
      "loss": 0.0124,
      "step": 1360
    },
    {
      "epoch": 38.06,
      "learning_rate": 1.653846153846154e-05,
      "loss": 0.0154,
      "step": 1370
    },
    {
      "epoch": 38.33,
      "learning_rate": 1.6153846153846154e-05,
      "loss": 0.0196,
      "step": 1380
    },
    {
      "epoch": 38.61,
      "learning_rate": 1.576923076923077e-05,
      "loss": 0.0123,
      "step": 1390
    },
    {
      "epoch": 38.89,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.0128,
      "step": 1400
    },
    {
      "epoch": 39.17,
      "learning_rate": 1.5e-05,
      "loss": 0.0234,
      "step": 1410
    },
    {
      "epoch": 39.44,
      "learning_rate": 1.4615384615384617e-05,
      "loss": 0.0123,
      "step": 1420
    },
    {
      "epoch": 39.72,
      "learning_rate": 1.423076923076923e-05,
      "loss": 0.0121,
      "step": 1430
    },
    {
      "epoch": 40.0,
      "learning_rate": 1.3846153846153847e-05,
      "loss": 0.0167,
      "step": 1440
    },
    {
      "epoch": 40.28,
      "learning_rate": 1.3461538461538462e-05,
      "loss": 0.0146,
      "step": 1450
    },
    {
      "epoch": 40.56,
      "learning_rate": 1.3076923076923078e-05,
      "loss": 0.0205,
      "step": 1460
    },
    {
      "epoch": 40.83,
      "learning_rate": 1.2692307692307691e-05,
      "loss": 0.0125,
      "step": 1470
    },
    {
      "epoch": 41.11,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.0104,
      "step": 1480
    },
    {
      "epoch": 41.39,
      "learning_rate": 1.1923076923076925e-05,
      "loss": 0.021,
      "step": 1490
    },
    {
      "epoch": 41.67,
      "learning_rate": 1.153846153846154e-05,
      "loss": 0.0121,
      "step": 1500
    }
  ],
  "logging_steps": 10,
  "max_steps": 1800,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 812050808832000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}