run-5/checkpoint-633/trainer_state.json

{
  "best_metric": 0.9139908256880734,
  "best_model_checkpoint": "bert-base-uncased-finetuned-sst2/run-5/checkpoint-633",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 633,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9002293577981652,
      "eval_loss": 0.24879497289657593,
      "eval_runtime": 2.3485,
      "eval_samples_per_second": 371.303,
      "eval_steps_per_second": 23.419,
      "step": 211
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9059633027522935,
      "eval_loss": 0.2516523599624634,
      "eval_runtime": 2.3894,
      "eval_samples_per_second": 364.945,
      "eval_steps_per_second": 23.018,
      "step": 422
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.7729095220565796,
      "learning_rate": 2.651478911079213e-05,
      "loss": 0.2165,
      "step": 500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9139908256880734,
      "eval_loss": 0.3265123963356018,
      "eval_runtime": 2.3971,
      "eval_samples_per_second": 363.775,
      "eval_steps_per_second": 22.944,
      "step": 633
    }
  ],
  "logging_steps": 500,
  "max_steps": 844,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 335486123357640.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 6.50537267718272e-05,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 32,
    "seed": 2
  }
}