trainer_state.json

{
  "best_metric": 0.940987712295531,
  "best_model_checkpoint": "model_saves/deberta-large_spell_5k_6_p3/checkpoint-268",
  "epoch": 4.0,
  "global_step": 1072,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.940987712295531,
      "eval_loss": 0.4161001443862915,
      "eval_runtime": 4.6558,
      "eval_samples_per_second": 941.182,
      "eval_steps_per_second": 7.517,
      "step": 268
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9409485794787509,
      "eval_loss": 0.4395348131656647,
      "eval_runtime": 4.6611,
      "eval_samples_per_second": 940.112,
      "eval_steps_per_second": 7.509,
      "step": 536
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9400429342904102,
      "eval_loss": 0.4750344753265381,
      "eval_runtime": 4.6519,
      "eval_samples_per_second": 941.975,
      "eval_steps_per_second": 7.524,
      "step": 804
    },
    {
      "epoch": 3.73,
      "learning_rate": 1e-05,
      "loss": 0.2544,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9394671228435023,
      "eval_loss": 0.5293166041374207,
      "eval_runtime": 4.6693,
      "eval_samples_per_second": 938.468,
      "eval_steps_per_second": 7.496,
      "step": 1072
    },
    {
      "epoch": 4.0,
      "step": 1072,
      "total_flos": 1.609706900553728e+16,
      "train_loss": 0.24813842417588874,
      "train_runtime": 470.0935,
      "train_samples_per_second": 1094.591,
      "train_steps_per_second": 8.551
    }
  ],
  "max_steps": 4020,
  "num_train_epochs": 15,
  "total_flos": 1.609706900553728e+16,
  "trial_name": null,
  "trial_params": null
}