trainer_state.json

{
  "best_metric": 0.9412840036225808,
  "best_model_checkpoint": "model_saves/deberta-large_spell_5k_3_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": 0.9412840036225808,
      "eval_loss": 0.41558289527893066,
      "eval_runtime": 4.666,
      "eval_samples_per_second": 939.137,
      "eval_steps_per_second": 7.501,
      "step": 268
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9406187457373182,
      "eval_loss": 0.4373631179332733,
      "eval_runtime": 4.651,
      "eval_samples_per_second": 942.161,
      "eval_steps_per_second": 7.525,
      "step": 536
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9397745949753463,
      "eval_loss": 0.4800662696361542,
      "eval_runtime": 4.656,
      "eval_samples_per_second": 941.151,
      "eval_steps_per_second": 7.517,
      "step": 804
    },
    {
      "epoch": 3.73,
      "learning_rate": 1e-05,
      "loss": 0.2528,
      "step": 1000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9388465881774171,
      "eval_loss": 0.5268030166625977,
      "eval_runtime": 4.656,
      "eval_samples_per_second": 941.15,
      "eval_steps_per_second": 7.517,
      "step": 1072
    },
    {
      "epoch": 4.0,
      "step": 1072,
      "total_flos": 1.6093771678613504e+16,
      "train_loss": 0.24649164036138735,
      "train_runtime": 460.9187,
      "train_samples_per_second": 1116.379,
      "train_steps_per_second": 8.722
    }
  ],
  "max_steps": 4020,
  "num_train_epochs": 15,
  "total_flos": 1.6093771678613504e+16,
  "trial_name": null,
  "trial_params": null
}