checkpoint-500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.1,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 1.9600000000000003e-06,
      "loss": 4.1002,
      "step": 100
    },
    {
      "epoch": 0.04,
      "learning_rate": 3.96e-06,
      "loss": 3.4625,
      "step": 200
    },
    {
      "epoch": 0.06,
      "learning_rate": 5.9600000000000005e-06,
      "loss": 2.4907,
      "step": 300
    },
    {
      "epoch": 0.08,
      "learning_rate": 7.960000000000002e-06,
      "loss": 1.5881,
      "step": 400
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.960000000000001e-06,
      "loss": 1.1583,
      "step": 500
    },
    {
      "epoch": 0.1,
      "eval_combined_wer": 0.3216761184625079,
      "eval_f1_score": 0.07455306200076076,
      "eval_label_f1": 0.1414986686953214,
      "eval_loss": 1.0361335277557373,
      "eval_runtime": 337.7385,
      "eval_samples_per_second": 2.961,
      "eval_steps_per_second": 0.047,
      "eval_wer": 0.20672983948946044,
      "step": 500
    }
  ],
  "logging_steps": 100,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "total_flos": 1.864612970496e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}