{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.42158516020236086,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.1470233384070705e-06,
      "loss": 1.4705,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 5.4552635448851985e-06,
      "loss": 1.5012,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 6.289174995846284e-06,
      "loss": 1.385,
      "step": 30
    },
    {
      "epoch": 0.02,
      "learning_rate": 6.865549773769684e-06,
      "loss": 1.2729,
      "step": 40
    },
    {
      "epoch": 0.02,
      "learning_rate": 7.306472495116047e-06,
      "loss": 1.2978,
      "step": 50
    },
    {
      "epoch": 0.03,
      "learning_rate": 7.663646266610644e-06,
      "loss": 1.4354,
      "step": 60
    },
    {
      "epoch": 0.03,
      "learning_rate": 7.963863644365277e-06,
      "loss": 1.4443,
      "step": 70
    },
    {
      "epoch": 0.03,
      "learning_rate": 8.222815896602431e-06,
      "loss": 1.5108,
      "step": 80
    },
    {
      "epoch": 0.04,
      "learning_rate": 8.450488154497406e-06,
      "loss": 1.3964,
      "step": 90
    },
    {
      "epoch": 0.04,
      "learning_rate": 8.653629372258186e-06,
      "loss": 1.2203,
      "step": 100
    },
    {
      "epoch": 0.05,
      "learning_rate": 8.837015420566862e-06,
      "loss": 1.3174,
      "step": 110
    },
    {
      "epoch": 0.05,
      "learning_rate": 9.004150364012388e-06,
      "loss": 1.1764,
      "step": 120
    },
    {
      "epoch": 0.05,
      "learning_rate": 9.157681445346895e-06,
      "loss": 1.1009,
      "step": 130
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.299657503090295e-06,
      "loss": 1.0422,
      "step": 140
    },
    {
      "epoch": 0.06,
      "learning_rate": 9.43169675964485e-06,
      "loss": 1.0693,
      "step": 150
    },
    {
      "epoch": 0.07,
      "learning_rate": 9.555099634066188e-06,
      "loss": 0.9296,
      "step": 160
    },
    {
      "epoch": 0.07,
      "learning_rate": 9.670926871527948e-06,
      "loss": 1.0445,
      "step": 170
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.780055047302923e-06,
      "loss": 1.008,
      "step": 180
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.883216879176116e-06,
      "loss": 0.9062,
      "step": 190
    },
    {
      "epoch": 0.08,
      "learning_rate": 9.98103107994822e-06,
      "loss": 1.0707,
      "step": 200
    },
    {
      "epoch": 0.09,
      "learning_rate": 1e-05,
      "loss": 1.0204,
      "step": 210
    },
    {
      "epoch": 0.09,
      "learning_rate": 1e-05,
      "loss": 0.8036,
      "step": 220
    },
    {
      "epoch": 0.1,
      "learning_rate": 1e-05,
      "loss": 0.9934,
      "step": 230
    },
    {
      "epoch": 0.1,
      "learning_rate": 1e-05,
      "loss": 0.9144,
      "step": 240
    },
    {
      "epoch": 0.11,
      "learning_rate": 1e-05,
      "loss": 0.8351,
      "step": 250
    },
    {
      "epoch": 0.11,
      "learning_rate": 1e-05,
      "loss": 0.8797,
      "step": 260
    },
    {
      "epoch": 0.11,
      "learning_rate": 1e-05,
      "loss": 0.7679,
      "step": 270
    },
    {
      "epoch": 0.12,
      "learning_rate": 1e-05,
      "loss": 0.9388,
      "step": 280
    },
    {
      "epoch": 0.12,
      "learning_rate": 1e-05,
      "loss": 0.9333,
      "step": 290
    },
    {
      "epoch": 0.13,
      "learning_rate": 1e-05,
      "loss": 0.8698,
      "step": 300
    },
    {
      "epoch": 0.13,
      "learning_rate": 1e-05,
      "loss": 0.7842,
      "step": 310
    },
    {
      "epoch": 0.13,
      "learning_rate": 1e-05,
      "loss": 0.8371,
      "step": 320
    },
    {
      "epoch": 0.14,
      "learning_rate": 1e-05,
      "loss": 0.7694,
      "step": 330
    },
    {
      "epoch": 0.14,
      "learning_rate": 1e-05,
      "loss": 0.8426,
      "step": 340
    },
    {
      "epoch": 0.15,
      "learning_rate": 1e-05,
      "loss": 0.7972,
      "step": 350
    },
    {
      "epoch": 0.15,
      "learning_rate": 1e-05,
      "loss": 0.8183,
      "step": 360
    },
    {
      "epoch": 0.16,
      "learning_rate": 1e-05,
      "loss": 0.7462,
      "step": 370
    },
    {
      "epoch": 0.16,
      "learning_rate": 1e-05,
      "loss": 0.7919,
      "step": 380
    },
    {
      "epoch": 0.16,
      "learning_rate": 1e-05,
      "loss": 0.8173,
      "step": 390
    },
    {
      "epoch": 0.17,
      "learning_rate": 1e-05,
      "loss": 0.7654,
      "step": 400
    },
    {
      "epoch": 0.17,
      "learning_rate": 1e-05,
      "loss": 0.7601,
      "step": 410
    },
    {
      "epoch": 0.18,
      "learning_rate": 1e-05,
      "loss": 0.788,
      "step": 420
    },
    {
      "epoch": 0.18,
      "learning_rate": 1e-05,
      "loss": 0.7488,
      "step": 430
    },
    {
      "epoch": 0.19,
      "learning_rate": 1e-05,
      "loss": 0.7721,
      "step": 440
    },
    {
      "epoch": 0.19,
      "learning_rate": 1e-05,
      "loss": 0.8289,
      "step": 450
    },
    {
      "epoch": 0.19,
      "learning_rate": 1e-05,
      "loss": 0.7367,
      "step": 460
    },
    {
      "epoch": 0.2,
      "learning_rate": 1e-05,
      "loss": 0.8422,
      "step": 470
    },
    {
      "epoch": 0.2,
      "learning_rate": 1e-05,
      "loss": 0.7383,
      "step": 480
    },
    {
      "epoch": 0.21,
      "learning_rate": 1e-05,
      "loss": 0.7729,
      "step": 490
    },
    {
      "epoch": 0.21,
      "learning_rate": 1e-05,
      "loss": 0.7537,
      "step": 500
    },
    {
      "epoch": 0.22,
      "learning_rate": 1e-05,
      "loss": 0.7956,
      "step": 510
    },
    {
      "epoch": 0.22,
      "learning_rate": 1e-05,
      "loss": 0.8438,
      "step": 520
    },
    {
      "epoch": 0.22,
      "learning_rate": 1e-05,
      "loss": 0.7865,
      "step": 530
    },
    {
      "epoch": 0.23,
      "learning_rate": 1e-05,
      "loss": 0.7995,
      "step": 540
    },
    {
      "epoch": 0.23,
      "learning_rate": 1e-05,
      "loss": 0.6958,
      "step": 550
    },
    {
      "epoch": 0.24,
      "learning_rate": 1e-05,
      "loss": 0.805,
      "step": 560
    },
    {
      "epoch": 0.24,
      "learning_rate": 1e-05,
      "loss": 0.7969,
      "step": 570
    },
    {
      "epoch": 0.24,
      "learning_rate": 1e-05,
      "loss": 0.7798,
      "step": 580
    },
    {
      "epoch": 0.25,
      "learning_rate": 1e-05,
      "loss": 0.7764,
      "step": 590
    },
    {
      "epoch": 0.25,
      "learning_rate": 1e-05,
      "loss": 0.7392,
      "step": 600
    },
    {
      "epoch": 0.26,
      "learning_rate": 1e-05,
      "loss": 0.7266,
      "step": 610
    },
    {
      "epoch": 0.26,
      "learning_rate": 1e-05,
      "loss": 0.7475,
      "step": 620
    },
    {
      "epoch": 0.27,
      "learning_rate": 1e-05,
      "loss": 0.7397,
      "step": 630
    },
    {
      "epoch": 0.27,
      "learning_rate": 1e-05,
      "loss": 0.72,
      "step": 640
    },
    {
      "epoch": 0.27,
      "learning_rate": 1e-05,
      "loss": 0.7729,
      "step": 650
    },
    {
      "epoch": 0.28,
      "learning_rate": 1e-05,
      "loss": 0.7378,
      "step": 660
    },
    {
      "epoch": 0.28,
      "learning_rate": 1e-05,
      "loss": 0.7306,
      "step": 670
    },
    {
      "epoch": 0.29,
      "learning_rate": 1e-05,
      "loss": 0.8145,
      "step": 680
    },
    {
      "epoch": 0.29,
      "learning_rate": 1e-05,
      "loss": 0.69,
      "step": 690
    },
    {
      "epoch": 0.3,
      "learning_rate": 1e-05,
      "loss": 0.7128,
      "step": 700
    },
    {
      "epoch": 0.3,
      "learning_rate": 1e-05,
      "loss": 0.7634,
      "step": 710
    },
    {
      "epoch": 0.3,
      "learning_rate": 1e-05,
      "loss": 0.7451,
      "step": 720
    },
    {
      "epoch": 0.31,
      "learning_rate": 1e-05,
      "loss": 0.7208,
      "step": 730
    },
    {
      "epoch": 0.31,
      "learning_rate": 1e-05,
      "loss": 0.8274,
      "step": 740
    },
    {
      "epoch": 0.32,
      "learning_rate": 1e-05,
      "loss": 0.6745,
      "step": 750
    },
    {
      "epoch": 0.32,
      "learning_rate": 1e-05,
      "loss": 0.6813,
      "step": 760
    },
    {
      "epoch": 0.32,
      "learning_rate": 1e-05,
      "loss": 0.7841,
      "step": 770
    },
    {
      "epoch": 0.33,
      "learning_rate": 1e-05,
      "loss": 0.7459,
      "step": 780
    },
    {
      "epoch": 0.33,
      "learning_rate": 1e-05,
      "loss": 0.7265,
      "step": 790
    },
    {
      "epoch": 0.34,
      "learning_rate": 1e-05,
      "loss": 0.6608,
      "step": 800
    },
    {
      "epoch": 0.34,
      "learning_rate": 1e-05,
      "loss": 0.6852,
      "step": 810
    },
    {
      "epoch": 0.35,
      "learning_rate": 1e-05,
      "loss": 0.6859,
      "step": 820
    },
    {
      "epoch": 0.35,
      "learning_rate": 1e-05,
      "loss": 0.7223,
      "step": 830
    },
    {
      "epoch": 0.35,
      "learning_rate": 1e-05,
      "loss": 0.7669,
      "step": 840
    },
    {
      "epoch": 0.36,
      "learning_rate": 1e-05,
      "loss": 0.7634,
      "step": 850
    },
    {
      "epoch": 0.36,
      "learning_rate": 1e-05,
      "loss": 0.7344,
      "step": 860
    },
    {
      "epoch": 0.37,
      "learning_rate": 1e-05,
      "loss": 0.7292,
      "step": 870
    },
    {
      "epoch": 0.37,
      "learning_rate": 1e-05,
      "loss": 0.7836,
      "step": 880
    },
    {
      "epoch": 0.38,
      "learning_rate": 1e-05,
      "loss": 0.6891,
      "step": 890
    },
    {
      "epoch": 0.38,
      "learning_rate": 1e-05,
      "loss": 0.7374,
      "step": 900
    },
    {
      "epoch": 0.38,
      "learning_rate": 1e-05,
      "loss": 0.749,
      "step": 910
    },
    {
      "epoch": 0.39,
      "learning_rate": 1e-05,
      "loss": 0.7338,
      "step": 920
    },
    {
      "epoch": 0.39,
      "learning_rate": 1e-05,
      "loss": 0.7597,
      "step": 930
    },
    {
      "epoch": 0.4,
      "learning_rate": 1e-05,
      "loss": 0.6914,
      "step": 940
    },
    {
      "epoch": 0.4,
      "learning_rate": 1e-05,
      "loss": 0.6781,
      "step": 950
    },
    {
      "epoch": 0.4,
      "learning_rate": 1e-05,
      "loss": 0.7396,
      "step": 960
    },
    {
      "epoch": 0.41,
      "learning_rate": 1e-05,
      "loss": 0.7296,
      "step": 970
    },
    {
      "epoch": 0.41,
      "learning_rate": 1e-05,
      "loss": 0.7038,
      "step": 980
    },
    {
      "epoch": 0.42,
      "learning_rate": 1e-05,
      "loss": 0.7363,
      "step": 990
    },
    {
      "epoch": 0.42,
      "learning_rate": 1e-05,
      "loss": 0.7154,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 251028716912640.0,
  "trial_name": null,
  "trial_params": null
}