trainer_state.json

{
  "best_metric": 153.2258064516129,
  "best_model_checkpoint": "./checkpoint-1",
  "epoch": 1.0,
  "global_step": 2,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "learning_rate": 0.0,
      "loss": 1.565,
      "step": 1
    },
    {
      "epoch": 0.5,
      "eval_loss": 1.6191972494125366,
      "eval_runtime": 41.7522,
      "eval_samples_per_second": 0.096,
      "eval_steps_per_second": 0.048,
      "eval_wer": 153.2258064516129,
      "step": 1
    },
    {
      "epoch": 1.0,
      "learning_rate": 1e-05,
      "loss": 1.3028,
      "step": 2
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.6191972494125366,
      "eval_runtime": 37.2469,
      "eval_samples_per_second": 0.107,
      "eval_steps_per_second": 0.054,
      "eval_wer": 153.2258064516129,
      "step": 2
    },
    {
      "epoch": 1.0,
      "step": 2,
      "total_flos": 196951080960000.0,
      "train_loss": 1.4339025020599365,
      "train_runtime": 108.1566,
      "train_samples_per_second": 0.074,
      "train_steps_per_second": 0.018
    }
  ],
  "max_steps": 2,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 196951080960000.0,
  "trial_name": null,
  "trial_params": null
}