{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 131.57894736842104,
  "global_step": 17500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.76,
      "learning_rate": 1e-05,
      "learning_rate_embeddings": 1e-05,
      "loss": 9.5582,
      "step": 500
    },
    {
      "epoch": 7.52,
      "learning_rate": 2e-05,
      "learning_rate_embeddings": 2e-05,
      "loss": 8.0322,
      "step": 1000
    },
    {
      "epoch": 11.28,
      "learning_rate": 3e-05,
      "learning_rate_embeddings": 3e-05,
      "loss": 7.3149,
      "step": 1500
    },
    {
      "epoch": 15.04,
      "learning_rate": 4e-05,
      "learning_rate_embeddings": 4e-05,
      "loss": 6.6378,
      "step": 2000
    },
    {
      "epoch": 18.8,
      "learning_rate": 5e-05,
      "learning_rate_embeddings": 5e-05,
      "loss": 6.0786,
      "step": 2500
    },
    {
      "epoch": 18.8,
      "eval_loss": 5.710835933685303,
      "eval_runtime": 23.8206,
      "eval_samples_per_second": 315.399,
      "eval_steps_per_second": 2.477,
      "step": 2500
    },
    {
      "epoch": 22.56,
      "learning_rate": 6e-05,
      "learning_rate_embeddings": 6e-05,
      "loss": 5.6654,
      "step": 3000
    },
    {
      "epoch": 26.32,
      "learning_rate": 7.000000000000001e-05,
      "learning_rate_embeddings": 7.000000000000001e-05,
      "loss": 5.3417,
      "step": 3500
    },
    {
      "epoch": 30.08,
      "learning_rate": 8e-05,
      "learning_rate_embeddings": 8e-05,
      "loss": 5.0624,
      "step": 4000
    },
    {
      "epoch": 33.83,
      "learning_rate": 8.999999999999999e-05,
      "learning_rate_embeddings": 8.999999999999999e-05,
      "loss": 4.8364,
      "step": 4500
    },
    {
      "epoch": 37.59,
      "learning_rate": 0.0001,
      "learning_rate_embeddings": 0.0001,
      "loss": 4.6372,
      "step": 5000
    },
    {
      "epoch": 37.59,
      "eval_loss": 4.484490394592285,
      "eval_runtime": 10.1862,
      "eval_samples_per_second": 737.569,
      "eval_steps_per_second": 5.792,
      "step": 5000
    },
    {
      "epoch": 41.35,
      "learning_rate": 0.00011,
      "learning_rate_embeddings": 0.00011,
      "loss": 4.4734,
      "step": 5500
    },
    {
      "epoch": 45.11,
      "learning_rate": 0.00012,
      "learning_rate_embeddings": 0.00012,
      "loss": 4.322,
      "step": 6000
    },
    {
      "epoch": 48.87,
      "learning_rate": 0.00013000000000000002,
      "learning_rate_embeddings": 0.00013000000000000002,
      "loss": 4.1937,
      "step": 6500
    },
    {
      "epoch": 52.63,
      "learning_rate": 0.00014000000000000001,
      "learning_rate_embeddings": 0.00014000000000000001,
      "loss": 4.0804,
      "step": 7000
    },
    {
      "epoch": 56.39,
      "learning_rate": 0.00015,
      "learning_rate_embeddings": 0.00015,
      "loss": 3.9741,
      "step": 7500
    },
    {
      "epoch": 56.39,
      "eval_loss": 3.9806618690490723,
      "eval_runtime": 10.0742,
      "eval_samples_per_second": 745.765,
      "eval_steps_per_second": 5.857,
      "step": 7500
    },
    {
      "epoch": 60.15,
      "learning_rate": 0.00016,
      "learning_rate_embeddings": 0.00016,
      "loss": 3.8851,
      "step": 8000
    },
    {
      "epoch": 63.91,
      "learning_rate": 0.00017,
      "learning_rate_embeddings": 0.00017,
      "loss": 3.796,
      "step": 8500
    },
    {
      "epoch": 67.67,
      "learning_rate": 0.00017999999999999998,
      "learning_rate_embeddings": 0.00017999999999999998,
      "loss": 3.7154,
      "step": 9000
    },
    {
      "epoch": 71.43,
      "learning_rate": 0.00019,
      "learning_rate_embeddings": 0.00019,
      "loss": 3.6434,
      "step": 9500
    },
    {
      "epoch": 75.19,
      "learning_rate": 0.0002,
      "learning_rate_embeddings": 0.0002,
      "loss": 3.5794,
      "step": 10000
    },
    {
      "epoch": 75.19,
      "eval_loss": 3.74064564704895,
      "eval_runtime": 10.1022,
      "eval_samples_per_second": 743.703,
      "eval_steps_per_second": 5.84,
      "step": 10000
    },
    {
      "epoch": 78.95,
      "learning_rate": 0.00021,
      "learning_rate_embeddings": 0.00021,
      "loss": 3.5056,
      "step": 10500
    },
    {
      "epoch": 82.71,
      "learning_rate": 0.00022,
      "learning_rate_embeddings": 0.00022,
      "loss": 3.4434,
      "step": 11000
    },
    {
      "epoch": 86.47,
      "learning_rate": 0.00023,
      "learning_rate_embeddings": 0.00023,
      "loss": 3.3841,
      "step": 11500
    },
    {
      "epoch": 90.23,
      "learning_rate": 0.00024,
      "learning_rate_embeddings": 0.00024,
      "loss": 3.3289,
      "step": 12000
    },
    {
      "epoch": 93.98,
      "learning_rate": 0.00025,
      "learning_rate_embeddings": 0.00025,
      "loss": 3.2738,
      "step": 12500
    },
    {
      "epoch": 93.98,
      "eval_loss": 3.6198933124542236,
      "eval_runtime": 10.0449,
      "eval_samples_per_second": 747.942,
      "eval_steps_per_second": 5.874,
      "step": 12500
    },
    {
      "epoch": 97.74,
      "learning_rate": 0.00026000000000000003,
      "learning_rate_embeddings": 0.00026000000000000003,
      "loss": 3.2181,
      "step": 13000
    },
    {
      "epoch": 101.5,
      "learning_rate": 0.00027,
      "learning_rate_embeddings": 0.00027,
      "loss": 3.1679,
      "step": 13500
    },
    {
      "epoch": 105.26,
      "learning_rate": 0.00028000000000000003,
      "learning_rate_embeddings": 0.00028000000000000003,
      "loss": 3.1144,
      "step": 14000
    },
    {
      "epoch": 109.02,
      "learning_rate": 0.00029,
      "learning_rate_embeddings": 0.00029,
      "loss": 3.067,
      "step": 14500
    },
    {
      "epoch": 112.78,
      "learning_rate": 0.0003,
      "learning_rate_embeddings": 0.0003,
      "loss": 3.0177,
      "step": 15000
    },
    {
      "epoch": 112.78,
      "eval_loss": 3.5627167224884033,
      "eval_runtime": 10.0076,
      "eval_samples_per_second": 750.731,
      "eval_steps_per_second": 5.896,
      "step": 15000
    },
    {
      "epoch": 116.54,
      "learning_rate": 0.00031,
      "learning_rate_embeddings": 0.00031,
      "loss": 2.9764,
      "step": 15500
    },
    {
      "epoch": 120.3,
      "learning_rate": 0.00032,
      "learning_rate_embeddings": 0.00032,
      "loss": 2.9304,
      "step": 16000
    },
    {
      "epoch": 124.06,
      "learning_rate": 0.00033,
      "learning_rate_embeddings": 0.00033,
      "loss": 2.8881,
      "step": 16500
    },
    {
      "epoch": 127.82,
      "learning_rate": 0.00034,
      "learning_rate_embeddings": 0.00034,
      "loss": 2.8463,
      "step": 17000
    },
    {
      "epoch": 131.58,
      "learning_rate": 0.00035,
      "learning_rate_embeddings": 0.00035,
      "loss": 2.8012,
      "step": 17500
    },
    {
      "epoch": 131.58,
      "eval_loss": 3.5356414318084717,
      "eval_runtime": 10.0661,
      "eval_samples_per_second": 746.367,
      "eval_steps_per_second": 5.861,
      "step": 17500
    }
  ],
  "max_steps": 250000,
  "num_train_epochs": 1880,
  "total_flos": 2.9264707584e+17,
  "trial_name": null,
  "trial_params": null
}