{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3099381673356165,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 2e-05,
      "loss": 9.8671,
      "step": 50
    },
    {
      "epoch": 0.01,
      "learning_rate": 4e-05,
      "loss": 8.0656,
      "step": 100
    },
    {
      "epoch": 0.01,
      "learning_rate": 6e-05,
      "loss": 7.0772,
      "step": 150
    },
    {
      "epoch": 0.01,
      "learning_rate": 8e-05,
      "loss": 6.5354,
      "step": 200
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0001,
      "loss": 6.3969,
      "step": 250
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.00012,
      "loss": 6.3607,
      "step": 300
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.00014000000000000001,
      "loss": 6.3618,
      "step": 350
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.00016,
      "loss": 6.3356,
      "step": 400
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00017999999999999998,
      "loss": 6.3251,
      "step": 450
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 6.3114,
      "step": 500
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00022,
      "loss": 6.3068,
      "step": 550
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00024,
      "loss": 6.3395,
      "step": 600
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00026000000000000003,
      "loss": 6.3043,
      "step": 650
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.00028000000000000003,
      "loss": 6.3283,
      "step": 700
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0003,
      "loss": 6.2978,
      "step": 750
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00032,
      "loss": 6.2991,
      "step": 800
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00034,
      "loss": 6.2918,
      "step": 850
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00035999999999999997,
      "loss": 6.2905,
      "step": 900
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00038,
      "loss": 6.2749,
      "step": 950
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0004,
      "loss": 6.2729,
      "step": 1000
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00042,
      "loss": 6.2746,
      "step": 1050
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00044,
      "loss": 6.2571,
      "step": 1100
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00046,
      "loss": 6.2429,
      "step": 1150
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00048,
      "loss": 6.217,
      "step": 1200
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0005,
      "loss": 6.1977,
      "step": 1250
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0005200000000000001,
      "loss": 6.1434,
      "step": 1300
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00054,
      "loss": 6.0999,
      "step": 1350
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0005600000000000001,
      "loss": 6.0825,
      "step": 1400
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00058,
      "loss": 6.0578,
      "step": 1450
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0006,
      "loss": 6.0093,
      "step": 1500
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00062,
      "loss": 5.9856,
      "step": 1550
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00064,
      "loss": 5.9808,
      "step": 1600
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00066,
      "loss": 5.9318,
      "step": 1650
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00068,
      "loss": 5.9262,
      "step": 1700
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0007,
      "loss": 5.9321,
      "step": 1750
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0007199999999999999,
      "loss": 5.8779,
      "step": 1800
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00074,
      "loss": 5.7584,
      "step": 1850
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00076,
      "loss": 5.6457,
      "step": 1900
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0007800000000000001,
      "loss": 5.5759,
      "step": 1950
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0008,
      "loss": 5.4955,
      "step": 2000
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00082,
      "loss": 5.4367,
      "step": 2050
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00084,
      "loss": 5.3887,
      "step": 2100
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00086,
      "loss": 5.3198,
      "step": 2150
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00088,
      "loss": 5.2592,
      "step": 2200
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0009000000000000001,
      "loss": 5.2409,
      "step": 2250
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00092,
      "loss": 5.1493,
      "step": 2300
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00094,
      "loss": 5.1274,
      "step": 2350
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00096,
      "loss": 5.0972,
      "step": 2400
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00098,
      "loss": 5.0449,
      "step": 2450
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.001,
      "loss": 5.0472,
      "step": 2500
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00098,
      "loss": 5.0071,
      "step": 2550
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00096,
      "loss": 4.9817,
      "step": 2600
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00094,
      "loss": 4.9384,
      "step": 2650
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00092,
      "loss": 4.9402,
      "step": 2700
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0009000000000000001,
      "loss": 4.887,
      "step": 2750
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.00088,
      "loss": 4.8869,
      "step": 2800
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00086,
      "loss": 4.8926,
      "step": 2850
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00084,
      "loss": 4.8436,
      "step": 2900
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00082,
      "loss": 4.8625,
      "step": 2950
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0008,
      "loss": 4.8511,
      "step": 3000
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0007800000000000001,
      "loss": 4.815,
      "step": 3050
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00076,
      "loss": 4.7687,
      "step": 3100
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00074,
      "loss": 4.7673,
      "step": 3150
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0007199999999999999,
      "loss": 4.7485,
      "step": 3200
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0007,
      "loss": 4.7563,
      "step": 3250
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00068,
      "loss": 4.7166,
      "step": 3300
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00066,
      "loss": 4.7135,
      "step": 3350
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00064,
      "loss": 4.7097,
      "step": 3400
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00062,
      "loss": 4.6542,
      "step": 3450
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0006,
      "loss": 4.6362,
      "step": 3500
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00058,
      "loss": 4.6111,
      "step": 3550
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0005600000000000001,
      "loss": 4.6141,
      "step": 3600
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00054,
      "loss": 4.5965,
      "step": 3650
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0005200000000000001,
      "loss": 4.5671,
      "step": 3700
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0005,
      "loss": 4.5781,
      "step": 3750
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00048,
      "loss": 4.5353,
      "step": 3800
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00046,
      "loss": 4.5466,
      "step": 3850
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00044,
      "loss": 4.5191,
      "step": 3900
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00042,
      "loss": 4.5226,
      "step": 3950
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0004,
      "loss": 4.502,
      "step": 4000
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00038,
      "loss": 4.4773,
      "step": 4050
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00035999999999999997,
      "loss": 4.4743,
      "step": 4100
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00034,
      "loss": 4.473,
      "step": 4150
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00032,
      "loss": 4.4472,
      "step": 4200
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0003,
      "loss": 4.4057,
      "step": 4250
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00028000000000000003,
      "loss": 4.4335,
      "step": 4300
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00026000000000000003,
      "loss": 4.3862,
      "step": 4350
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00024,
      "loss": 4.3897,
      "step": 4400
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00022,
      "loss": 4.3902,
      "step": 4450
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 4.3782,
      "step": 4500
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00017999999999999998,
      "loss": 4.3618,
      "step": 4550
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00016,
      "loss": 4.3635,
      "step": 4600
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00014000000000000001,
      "loss": 4.3466,
      "step": 4650
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00012,
      "loss": 4.3191,
      "step": 4700
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0001,
      "loss": 4.3319,
      "step": 4750
    },
    {
      "epoch": 0.3,
      "learning_rate": 8e-05,
      "loss": 4.3091,
      "step": 4800
    },
    {
      "epoch": 0.3,
      "learning_rate": 6e-05,
      "loss": 4.3186,
      "step": 4850
    },
    {
      "epoch": 0.3,
      "learning_rate": 4e-05,
      "loss": 4.3197,
      "step": 4900
    },
    {
      "epoch": 0.31,
      "learning_rate": 2e-05,
      "loss": 4.2912,
      "step": 4950
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0,
      "loss": 4.3033,
      "step": 5000
    },
    {
      "epoch": 0.31,
      "step": 5000,
      "total_flos": 7.554075721728e+16,
      "train_loss": 5.339358209228515,
      "train_runtime": 13704.7632,
      "train_samples_per_second": 93.398,
      "train_steps_per_second": 0.365
    }
  ],
  "logging_steps": 50,
  "max_steps": 5000,
  "num_train_epochs": 1,
  "save_steps": 2000,
  "total_flos": 7.554075721728e+16,
  "trial_name": null,
  "trial_params": null
}