trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1025,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.4400977995110025,
      "grad_norm": NaN,
      "learning_rate": 9.98e-05,
      "loss": 0.0,
      "step": 500
    },
    {
      "epoch": 2.4400977995110025,
      "eval_loss": Infinity,
      "eval_runtime": 372.0966,
      "eval_samples_per_second": 4.397,
      "eval_steps_per_second": 0.551,
      "eval_wer": 0.4636784155319565,
      "step": 500
    },
    {
      "epoch": 4.880195599022005,
      "grad_norm": NaN,
      "learning_rate": 4.952380952380953e-06,
      "loss": 0.0,
      "step": 1000
    },
    {
      "epoch": 4.880195599022005,
      "eval_loss": Infinity,
      "eval_runtime": 372.7362,
      "eval_samples_per_second": 4.389,
      "eval_steps_per_second": 0.55,
      "eval_wer": 0.4636784155319565,
      "step": 1000
    }
  ],
  "logging_steps": 500,
  "max_steps": 1025,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.4044119614033756e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}