{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 130.20671834625324,
  "global_step": 12500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5.21,
      "learning_rate": 1e-05,
      "learning_rate_embeddings": 1e-05,
      "loss": 8.4554,
      "step": 500
    },
    {
      "epoch": 10.41,
      "learning_rate": 2e-05,
      "learning_rate_embeddings": 2e-05,
      "loss": 5.9825,
      "step": 1000
    },
    {
      "epoch": 15.62,
      "learning_rate": 3e-05,
      "learning_rate_embeddings": 3e-05,
      "loss": 5.1246,
      "step": 1500
    },
    {
      "epoch": 20.83,
      "learning_rate": 4e-05,
      "learning_rate_embeddings": 4e-05,
      "loss": 4.648,
      "step": 2000
    },
    {
      "epoch": 26.04,
      "learning_rate": 5e-05,
      "learning_rate_embeddings": 5e-05,
      "loss": 4.333,
      "step": 2500
    },
    {
      "epoch": 26.04,
      "eval_loss": 4.06018590927124,
      "eval_runtime": 21.6357,
      "eval_samples_per_second": 251.945,
      "eval_steps_per_second": 1.987,
      "step": 2500
    },
    {
      "epoch": 31.25,
      "learning_rate": 6e-05,
      "learning_rate_embeddings": 6e-05,
      "loss": 4.0801,
      "step": 3000
    },
    {
      "epoch": 36.45,
      "learning_rate": 7.000000000000001e-05,
      "learning_rate_embeddings": 7.000000000000001e-05,
      "loss": 3.8748,
      "step": 3500
    },
    {
      "epoch": 41.66,
      "learning_rate": 8e-05,
      "learning_rate_embeddings": 8e-05,
      "loss": 3.7065,
      "step": 4000
    },
    {
      "epoch": 46.87,
      "learning_rate": 8.999999999999999e-05,
      "learning_rate_embeddings": 8.999999999999999e-05,
      "loss": 3.5779,
      "step": 4500
    },
    {
      "epoch": 52.08,
      "learning_rate": 0.0001,
      "learning_rate_embeddings": 0.0001,
      "loss": 3.4682,
      "step": 5000
    },
    {
      "epoch": 52.08,
      "eval_loss": 3.347400665283203,
      "eval_runtime": 7.4364,
      "eval_samples_per_second": 733.017,
      "eval_steps_per_second": 5.782,
      "step": 5000
    },
    {
      "epoch": 57.29,
      "learning_rate": 0.00011,
      "learning_rate_embeddings": 0.00011,
      "loss": 3.3567,
      "step": 5500
    },
    {
      "epoch": 62.5,
      "learning_rate": 0.00012,
      "learning_rate_embeddings": 0.00012,
      "loss": 3.2615,
      "step": 6000
    },
    {
      "epoch": 67.7,
      "learning_rate": 0.00013000000000000002,
      "learning_rate_embeddings": 0.00013000000000000002,
      "loss": 3.1695,
      "step": 6500
    },
    {
      "epoch": 72.91,
      "learning_rate": 0.00014000000000000001,
      "learning_rate_embeddings": 0.00014000000000000001,
      "loss": 3.0914,
      "step": 7000
    },
    {
      "epoch": 78.12,
      "learning_rate": 0.00015,
      "learning_rate_embeddings": 0.00015,
      "loss": 3.0173,
      "step": 7500
    },
    {
      "epoch": 78.12,
      "eval_loss": 3.027360200881958,
      "eval_runtime": 7.56,
      "eval_samples_per_second": 721.027,
      "eval_steps_per_second": 5.688,
      "step": 7500
    },
    {
      "epoch": 83.33,
      "learning_rate": 0.00016,
      "learning_rate_embeddings": 0.00016,
      "loss": 2.9281,
      "step": 8000
    },
    {
      "epoch": 88.54,
      "learning_rate": 0.00017,
      "learning_rate_embeddings": 0.00017,
      "loss": 2.8527,
      "step": 8500
    },
    {
      "epoch": 93.74,
      "learning_rate": 0.00017999999999999998,
      "learning_rate_embeddings": 0.00017999999999999998,
      "loss": 2.7853,
      "step": 9000
    },
    {
      "epoch": 98.95,
      "learning_rate": 0.00019,
      "learning_rate_embeddings": 0.00019,
      "loss": 2.7177,
      "step": 9500
    },
    {
      "epoch": 104.17,
      "learning_rate": 0.0002,
      "learning_rate_embeddings": 0.0002,
      "loss": 2.6528,
      "step": 10000
    },
    {
      "epoch": 104.17,
      "eval_loss": 2.862852096557617,
      "eval_runtime": 7.3882,
      "eval_samples_per_second": 737.797,
      "eval_steps_per_second": 5.82,
      "step": 10000
    },
    {
      "epoch": 109.37,
      "learning_rate": 0.00021,
      "learning_rate_embeddings": 0.00021,
      "loss": 2.5833,
      "step": 10500
    },
    {
      "epoch": 114.58,
      "learning_rate": 0.00022,
      "learning_rate_embeddings": 0.00022,
      "loss": 2.5229,
      "step": 11000
    },
    {
      "epoch": 119.79,
      "learning_rate": 0.00023,
      "learning_rate_embeddings": 0.00023,
      "loss": 2.4533,
      "step": 11500
    },
    {
      "epoch": 124.99,
      "learning_rate": 0.00024,
      "learning_rate_embeddings": 0.00024,
      "loss": 2.3875,
      "step": 12000
    },
    {
      "epoch": 130.21,
      "learning_rate": 0.00025,
      "learning_rate_embeddings": 0.00025,
      "loss": 2.3269,
      "step": 12500
    },
    {
      "epoch": 130.21,
      "eval_loss": 2.7988603115081787,
      "eval_runtime": 7.5079,
      "eval_samples_per_second": 726.039,
      "eval_steps_per_second": 5.727,
      "step": 12500
    }
  ],
  "max_steps": 250000,
  "num_train_epochs": 2605,
  "total_flos": 2.1066408787968e+17,
  "trial_name": null,
  "trial_params": null
}