trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.76923076923077,
  "eval_steps": 100,
  "global_step": 480,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.08,
      "eval_loss": 10.369086265563965,
      "eval_runtime": 5.0307,
      "eval_samples_per_second": 38.762,
      "eval_steps_per_second": 4.97,
      "eval_wer": 1.0,
      "step": 100
    },
    {
      "epoch": 6.15,
      "eval_loss": 3.5670173168182373,
      "eval_runtime": 4.9472,
      "eval_samples_per_second": 39.416,
      "eval_steps_per_second": 5.053,
      "eval_wer": 1.0,
      "step": 200
    },
    {
      "epoch": 9.23,
      "eval_loss": 3.113863945007324,
      "eval_runtime": 4.8913,
      "eval_samples_per_second": 39.866,
      "eval_steps_per_second": 5.111,
      "eval_wer": 0.9484029484029484,
      "step": 300
    },
    {
      "epoch": 12.31,
      "eval_loss": 3.094938039779663,
      "eval_runtime": 4.8858,
      "eval_samples_per_second": 39.912,
      "eval_steps_per_second": 5.117,
      "eval_wer": 0.9484029484029484,
      "step": 400
    },
    {
      "epoch": 14.77,
      "step": 480,
      "total_flos": 2.1645545063012813e+18,
      "train_loss": 6.593796793619791,
      "train_runtime": 936.6313,
      "train_samples_per_second": 16.447,
      "train_steps_per_second": 0.512
    }
  ],
  "logging_steps": 500,
  "max_steps": 480,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 400,
  "total_flos": 2.1645545063012813e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}