checkpoint-500/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.4525018129079044,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.29,
      "learning_rate": 7.151470588235293e-05,
      "loss": 5.1135,
      "step": 100
    },
    {
      "epoch": 0.58,
      "learning_rate": 6.710294117647058e-05,
      "loss": 3.0957,
      "step": 200
    },
    {
      "epoch": 0.87,
      "learning_rate": 6.269117647058824e-05,
      "loss": 3.0078,
      "step": 300
    },
    {
      "epoch": 1.16,
      "learning_rate": 5.827941176470588e-05,
      "loss": 2.9785,
      "step": 400
    },
    {
      "epoch": 1.45,
      "learning_rate": 5.3867647058823525e-05,
      "loss": 2.9069,
      "step": 500
    },
    {
      "epoch": 1.45,
      "eval_loss": 2.9046826362609863,
      "eval_runtime": 128.7688,
      "eval_samples_per_second": 35.878,
      "eval_steps_per_second": 4.489,
      "eval_wer": 1.0,
      "step": 500
    }
  ],
  "max_steps": 1720,
  "num_train_epochs": 5,
  "total_flos": 1.4827194756605722e+18,
  "trial_name": null,
  "trial_params": null
}