trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "global_step": 2414,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 0.00027499999999999996,
      "loss": 2.2345,
      "step": 603
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.330522060394287,
      "eval_runtime": 78.1023,
      "eval_samples_per_second": 478.232,
      "eval_steps_per_second": 59.781,
      "step": 603
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002499585406301824,
      "loss": 1.9491,
      "step": 1207
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.3658900260925293,
      "eval_runtime": 78.301,
      "eval_samples_per_second": 477.018,
      "eval_steps_per_second": 59.629,
      "step": 1207
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0002249585406301824,
      "loss": 1.7348,
      "step": 1810
    },
    {
      "epoch": 3.0,
      "eval_loss": 2.463245391845703,
      "eval_runtime": 77.97,
      "eval_samples_per_second": 479.043,
      "eval_steps_per_second": 59.882,
      "step": 1810
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.00019991708126036484,
      "loss": 1.5141,
      "step": 2414
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.605247735977173,
      "eval_runtime": 77.3626,
      "eval_samples_per_second": 482.804,
      "eval_steps_per_second": 60.352,
      "step": 2414
    }
  ],
  "max_steps": 7236,
  "num_train_epochs": 12,
  "total_flos": 9.504695164518605e+16,
  "trial_name": null,
  "trial_params": null
}