trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 0.6447093486785889,
      "eval_runtime": 4.5652,
      "eval_samples_per_second": 7.01,
      "eval_steps_per_second": 0.219,
      "eval_wer": 0.4030808729139923,
      "step": 4
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.6389366984367371,
      "eval_runtime": 4.147,
      "eval_samples_per_second": 7.716,
      "eval_steps_per_second": 0.241,
      "eval_wer": 0.3992297817715019,
      "step": 8
    },
    {
      "epoch": 2.5,
      "learning_rate": 9.000000000000001e-07,
      "loss": 0.4891,
      "step": 10
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.6194185018539429,
      "eval_runtime": 4.1704,
      "eval_samples_per_second": 7.673,
      "eval_steps_per_second": 0.24,
      "eval_wer": 0.392811296534018,
      "step": 12
    },
    {
      "epoch": 3.0,
      "step": 12,
      "total_flos": 1.1081679372288e+17,
      "train_loss": 0.4858972082535426,
      "train_runtime": 84.4207,
      "train_samples_per_second": 4.549,
      "train_steps_per_second": 0.142
    }
  ],
  "max_steps": 12,
  "num_train_epochs": 3,
  "total_flos": 1.1081679372288e+17,
  "trial_name": null,
  "trial_params": null
}