trainer_state.json

{
  "best_metric": 43.05182737972331,
  "best_model_checkpoint": "./small-trsc-3/checkpoint-800",
  "epoch": 0.2820874471086037,
  "eval_steps": 400,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0003526093088857546,
      "grad_norm": 248945.40625,
      "learning_rate": 2e-08,
      "loss": 0.1794,
      "step": 1
    },
    {
      "epoch": 0.08815232722143865,
      "grad_norm": 137488.28125,
      "learning_rate": 5e-06,
      "loss": 0.1084,
      "step": 250
    },
    {
      "epoch": 0.14104372355430184,
      "eval_loss": 0.06996609270572662,
      "eval_runtime": 525.4559,
      "eval_samples_per_second": 0.658,
      "eval_steps_per_second": 0.084,
      "eval_wer": 50.691720008259345,
      "step": 400
    },
    {
      "epoch": 0.1763046544428773,
      "grad_norm": 100189.40625,
      "learning_rate": 1e-05,
      "loss": 0.0741,
      "step": 500
    },
    {
      "epoch": 0.2644569816643159,
      "grad_norm": 82682.015625,
      "learning_rate": 6.4285714285714295e-06,
      "loss": 0.0616,
      "step": 750
    },
    {
      "epoch": 0.2820874471086037,
      "eval_loss": 0.05677231401205063,
      "eval_runtime": 520.6337,
      "eval_samples_per_second": 0.665,
      "eval_steps_per_second": 0.085,
      "eval_wer": 43.05182737972331,
      "step": 800
    }
  ],
  "logging_steps": 250,
  "max_steps": 1200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 400,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.846946562048e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}