checkpoint-422/trainer_state.json

{
  "best_metric": 0.7583549419422403,
  "best_model_checkpoint": "outputs/time_lm_nli/checkpoint-422",
  "epoch": 2.0,
  "global_step": 422,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 9.666666666666666e-07,
      "loss": 0.5392,
      "step": 211
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7,
      "eval_f1": 0.6998718119161117,
      "eval_loss": 0.4048246741294861,
      "eval_runtime": 5.0256,
      "eval_samples_per_second": 596.939,
      "eval_steps_per_second": 74.617,
      "step": 211
    },
    {
      "epoch": 2.0,
      "learning_rate": 9.333333333333333e-07,
      "loss": 0.3725,
      "step": 422
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.759,
      "eval_f1": 0.7583549419422403,
      "eval_loss": 0.3489450216293335,
      "eval_runtime": 5.103,
      "eval_samples_per_second": 587.885,
      "eval_steps_per_second": 73.486,
      "step": 422
    }
  ],
  "max_steps": 6330,
  "num_train_epochs": 30,
  "total_flos": 2864746003045248.0,
  "trial_name": null,
  "trial_params": null
}