checkpoint-1000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9050036258158087,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.29,
      "learning_rate": 3.7125e-06,
      "loss": 12.2713,
      "step": 100
    },
    {
      "epoch": 0.58,
      "learning_rate": 7.4625e-06,
      "loss": 6.2026,
      "step": 200
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.1212499999999998e-05,
      "loss": 4.1374,
      "step": 300
    },
    {
      "epoch": 1.16,
      "learning_rate": 1.49625e-05,
      "loss": 3.6755,
      "step": 400
    },
    {
      "epoch": 1.45,
      "learning_rate": 1.8712499999999997e-05,
      "loss": 3.3332,
      "step": 500
    },
    {
      "epoch": 1.45,
      "eval_loss": 3.292029857635498,
      "eval_runtime": 137.2974,
      "eval_samples_per_second": 33.65,
      "eval_steps_per_second": 4.21,
      "eval_wer": 1.0,
      "step": 500
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.2462499999999997e-05,
      "loss": 3.1418,
      "step": 600
    },
    {
      "epoch": 2.03,
      "learning_rate": 2.6212499999999997e-05,
      "loss": 3.0879,
      "step": 700
    },
    {
      "epoch": 2.32,
      "learning_rate": 2.99625e-05,
      "loss": 3.0216,
      "step": 800
    },
    {
      "epoch": 2.61,
      "learning_rate": 3.37125e-05,
      "loss": 2.9595,
      "step": 900
    },
    {
      "epoch": 2.91,
      "learning_rate": 3.7462499999999996e-05,
      "loss": 2.9269,
      "step": 1000
    },
    {
      "epoch": 2.91,
      "eval_loss": 2.941540241241455,
      "eval_runtime": 178.4388,
      "eval_samples_per_second": 25.891,
      "eval_steps_per_second": 3.239,
      "eval_wer": 0.9966346448020559,
      "step": 1000
    }
  ],
  "max_steps": 17200,
  "num_train_epochs": 50,
  "total_flos": 2.9636799545011507e+18,
  "trial_name": null,
  "trial_params": null
}