trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.7699115044247788,
  "eval_steps": 50,
  "global_step": 100,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.44,
      "grad_norm": 9.63073444366455,
      "learning_rate": 5e-06,
      "loss": 0.7305,
      "step": 25
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.945356607437134,
      "learning_rate": 1e-05,
      "loss": 0.1912,
      "step": 50
    },
    {
      "epoch": 0.88,
      "eval_loss": 3.5328922271728516,
      "eval_runtime": 111.5535,
      "eval_samples_per_second": 0.887,
      "eval_steps_per_second": 0.063,
      "eval_wer": 1.2171099928109275,
      "step": 50
    },
    {
      "epoch": 1.33,
      "grad_norm": 2.5617167949676514,
      "learning_rate": 5e-06,
      "loss": 0.0998,
      "step": 75
    },
    {
      "epoch": 1.77,
      "grad_norm": 1.9796466827392578,
      "learning_rate": 0.0,
      "loss": 0.0828,
      "step": 100
    },
    {
      "epoch": 1.77,
      "eval_loss": 3.6638808250427246,
      "eval_runtime": 106.6363,
      "eval_samples_per_second": 0.928,
      "eval_steps_per_second": 0.066,
      "eval_wer": 1.216391085549964,
      "step": 100
    },
    {
      "epoch": 1.77,
      "step": 100,
      "total_flos": 9.23473281024e+17,
      "train_loss": 0.2760725998878479,
      "train_runtime": 1580.9218,
      "train_samples_per_second": 2.024,
      "train_steps_per_second": 0.063
    }
  ],
  "logging_steps": 25,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 50,
  "total_flos": 9.23473281024e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}