trainer_state.json

{
  "best_metric": 32.52656434474616,
  "best_model_checkpoint": "./whisper-tiny-en/checkpoint-500",
  "epoch": 17.857142857142858,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.89,
      "grad_norm": 34.12553405761719,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 3.184,
      "step": 25
    },
    {
      "epoch": 1.79,
      "grad_norm": 14.376067161560059,
      "learning_rate": 9.200000000000002e-06,
      "loss": 1.352,
      "step": 50
    },
    {
      "epoch": 2.68,
      "grad_norm": 11.755908966064453,
      "learning_rate": 1e-05,
      "loss": 0.4052,
      "step": 75
    },
    {
      "epoch": 3.57,
      "grad_norm": 7.901417255401611,
      "learning_rate": 1e-05,
      "loss": 0.2839,
      "step": 100
    },
    {
      "epoch": 4.46,
      "grad_norm": 5.7967610359191895,
      "learning_rate": 1e-05,
      "loss": 0.1608,
      "step": 125
    },
    {
      "epoch": 5.36,
      "grad_norm": 4.712026596069336,
      "learning_rate": 1e-05,
      "loss": 0.0984,
      "step": 150
    },
    {
      "epoch": 6.25,
      "grad_norm": 1.4813052415847778,
      "learning_rate": 1e-05,
      "loss": 0.0546,
      "step": 175
    },
    {
      "epoch": 7.14,
      "grad_norm": 0.8922443985939026,
      "learning_rate": 1e-05,
      "loss": 0.0286,
      "step": 200
    },
    {
      "epoch": 8.04,
      "grad_norm": 2.358762502670288,
      "learning_rate": 1e-05,
      "loss": 0.0134,
      "step": 225
    },
    {
      "epoch": 8.93,
      "grad_norm": 0.38798490166664124,
      "learning_rate": 1e-05,
      "loss": 0.0066,
      "step": 250
    },
    {
      "epoch": 9.82,
      "grad_norm": 0.184628427028656,
      "learning_rate": 1e-05,
      "loss": 0.0033,
      "step": 275
    },
    {
      "epoch": 10.71,
      "grad_norm": 0.10512322932481766,
      "learning_rate": 1e-05,
      "loss": 0.0021,
      "step": 300
    },
    {
      "epoch": 11.61,
      "grad_norm": 0.11653382331132889,
      "learning_rate": 1e-05,
      "loss": 0.002,
      "step": 325
    },
    {
      "epoch": 12.5,
      "grad_norm": 0.07005073130130768,
      "learning_rate": 1e-05,
      "loss": 0.0014,
      "step": 350
    },
    {
      "epoch": 13.39,
      "grad_norm": 0.05834714323282242,
      "learning_rate": 1e-05,
      "loss": 0.0011,
      "step": 375
    },
    {
      "epoch": 14.29,
      "grad_norm": 0.04799078404903412,
      "learning_rate": 1e-05,
      "loss": 0.0009,
      "step": 400
    },
    {
      "epoch": 15.18,
      "grad_norm": 0.045657627284526825,
      "learning_rate": 1e-05,
      "loss": 0.0009,
      "step": 425
    },
    {
      "epoch": 16.07,
      "grad_norm": 0.03895680606365204,
      "learning_rate": 1e-05,
      "loss": 0.0007,
      "step": 450
    },
    {
      "epoch": 16.96,
      "grad_norm": 0.039002932608127594,
      "learning_rate": 1e-05,
      "loss": 0.0007,
      "step": 475
    },
    {
      "epoch": 17.86,
      "grad_norm": 0.03650466352701187,
      "learning_rate": 1e-05,
      "loss": 0.0006,
      "step": 500
    },
    {
      "epoch": 17.86,
      "eval_loss": 0.6624370813369751,
      "eval_runtime": 48.1951,
      "eval_samples_per_second": 2.345,
      "eval_steps_per_second": 0.166,
      "eval_wer": 32.52656434474616,
      "eval_wer_ortho": 32.510795805058606,
      "step": 500
    }
  ],
  "logging_steps": 25,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 18,
  "save_steps": 500,
  "total_flos": 1.9569551781888e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}