trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "global_step": 436,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.92,
      "eval_loss": 0.1821640133857727,
      "eval_runtime": 40.8734,
      "eval_samples_per_second": 40.295,
      "eval_steps_per_second": 5.04,
      "eval_wer": 0.2605453988356654,
      "step": 100
    },
    {
      "epoch": 1.83,
      "eval_loss": 0.16198405623435974,
      "eval_runtime": 40.9812,
      "eval_samples_per_second": 40.189,
      "eval_steps_per_second": 5.027,
      "eval_wer": 0.2388928607905219,
      "step": 200
    },
    {
      "epoch": 2.75,
      "eval_loss": 0.15808044373989105,
      "eval_runtime": 41.0599,
      "eval_samples_per_second": 40.112,
      "eval_steps_per_second": 5.017,
      "eval_wer": 0.2318455724645082,
      "step": 300
    },
    {
      "epoch": 3.67,
      "eval_loss": 0.1534813940525055,
      "eval_runtime": 40.8801,
      "eval_samples_per_second": 40.289,
      "eval_steps_per_second": 5.039,
      "eval_wer": 0.22704524563374529,
      "step": 400
    },
    {
      "epoch": 4.0,
      "step": 436,
      "total_flos": 5.380629413554985e+18,
      "train_loss": 1.3154365469556335,
      "train_runtime": 794.3677,
      "train_samples_per_second": 17.513,
      "train_steps_per_second": 0.549
    }
  ],
  "max_steps": 436,
  "num_train_epochs": 4,
  "total_flos": 5.380629413554985e+18,
  "trial_name": null,
  "trial_params": null
}