checkpoint-4000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 444.0004,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 11.0,
      "learning_rate": 4.85e-06,
      "loss": 0.716,
      "step": 100
    },
    {
      "epoch": 22.0,
      "learning_rate": 9.85e-06,
      "loss": 0.0615,
      "step": 200
    },
    {
      "epoch": 33.0,
      "learning_rate": 1e-05,
      "loss": 0.0043,
      "step": 300
    },
    {
      "epoch": 44.0,
      "learning_rate": 1e-05,
      "loss": 0.002,
      "step": 400
    },
    {
      "epoch": 55.0,
      "learning_rate": 1e-05,
      "loss": 0.0008,
      "step": 500
    },
    {
      "epoch": 66.0,
      "learning_rate": 1e-05,
      "loss": 0.0006,
      "step": 600
    },
    {
      "epoch": 77.0,
      "learning_rate": 1e-05,
      "loss": 0.0004,
      "step": 700
    },
    {
      "epoch": 88.0,
      "learning_rate": 1e-05,
      "loss": 0.0003,
      "step": 800
    },
    {
      "epoch": 99.0,
      "learning_rate": 1e-05,
      "loss": 0.0003,
      "step": 900
    },
    {
      "epoch": 111.0,
      "learning_rate": 1e-05,
      "loss": 0.0002,
      "step": 1000
    },
    {
      "epoch": 111.0,
      "eval_loss": 0.21552081406116486,
      "eval_runtime": 10.773,
      "eval_samples_per_second": 1.671,
      "eval_steps_per_second": 0.093,
      "eval_wer": 9.75609756097561,
      "step": 1000
    },
    {
      "epoch": 122.0,
      "learning_rate": 1e-05,
      "loss": 0.0002,
      "step": 1100
    },
    {
      "epoch": 133.0,
      "learning_rate": 1e-05,
      "loss": 0.0002,
      "step": 1200
    },
    {
      "epoch": 144.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 1300
    },
    {
      "epoch": 155.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 1400
    },
    {
      "epoch": 166.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 1500
    },
    {
      "epoch": 177.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 1600
    },
    {
      "epoch": 188.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 1700
    },
    {
      "epoch": 199.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 1800
    },
    {
      "epoch": 211.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 1900
    },
    {
      "epoch": 222.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 2000
    },
    {
      "epoch": 222.0,
      "eval_loss": 0.2447899729013443,
      "eval_runtime": 14.1456,
      "eval_samples_per_second": 1.272,
      "eval_steps_per_second": 0.071,
      "eval_wer": 12.195121951219512,
      "step": 2000
    },
    {
      "epoch": 233.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 2100
    },
    {
      "epoch": 244.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 2200
    },
    {
      "epoch": 255.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 2300
    },
    {
      "epoch": 266.0,
      "learning_rate": 1e-05,
      "loss": 0.0001,
      "step": 2400
    },
    {
      "epoch": 277.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 2500
    },
    {
      "epoch": 288.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 2600
    },
    {
      "epoch": 299.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 2700
    },
    {
      "epoch": 311.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 2800
    },
    {
      "epoch": 322.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 2900
    },
    {
      "epoch": 333.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3000
    },
    {
      "epoch": 333.0,
      "eval_loss": 0.2674005329608917,
      "eval_runtime": 8.0249,
      "eval_samples_per_second": 2.243,
      "eval_steps_per_second": 0.125,
      "eval_wer": 13.414634146341465,
      "step": 3000
    },
    {
      "epoch": 344.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3100
    },
    {
      "epoch": 355.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3200
    },
    {
      "epoch": 366.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3300
    },
    {
      "epoch": 377.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3400
    },
    {
      "epoch": 388.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3500
    },
    {
      "epoch": 399.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3600
    },
    {
      "epoch": 411.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3700
    },
    {
      "epoch": 422.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3800
    },
    {
      "epoch": 433.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 3900
    },
    {
      "epoch": 444.0,
      "learning_rate": 1e-05,
      "loss": 0.0,
      "step": 4000
    },
    {
      "epoch": 444.0,
      "eval_loss": 0.29429247975349426,
      "eval_runtime": 7.7984,
      "eval_samples_per_second": 2.308,
      "eval_steps_per_second": 0.128,
      "eval_wer": 15.853658536585366,
      "step": 4000
    }
  ],
  "max_steps": 10000,
  "num_train_epochs": 9223372036854775807,
  "total_flos": 7.84131409575936e+18,
  "trial_name": null,
  "trial_params": null
}