run-13/checkpoint-633/trainer_state.json

{
  "best_metric": 0.9197247706422018,
  "best_model_checkpoint": "bert-base-uncased-finetuned-sst2/run-13/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": 0.911697247706422,
      "eval_loss": 0.22412490844726562,
      "eval_runtime": 2.4274,
      "eval_samples_per_second": 359.236,
      "eval_steps_per_second": 22.658,
      "step": 211
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9105504587155964,
      "eval_loss": 0.25786927342414856,
      "eval_runtime": 2.4527,
      "eval_samples_per_second": 355.532,
      "eval_steps_per_second": 22.425,
      "step": 422
    },
    {
      "epoch": 2.37,
      "grad_norm": 30.697444915771484,
      "learning_rate": 2.3450215142725074e-05,
      "loss": 0.2117,
      "step": 500
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9197247706422018,
      "eval_loss": 0.3487926125526428,
      "eval_runtime": 2.4283,
      "eval_samples_per_second": 359.093,
      "eval_steps_per_second": 22.649,
      "step": 633
    }
  ],
  "logging_steps": 500,
  "max_steps": 844,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 422743410424080.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 5.753483017575571e-05,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 32,
    "seed": 7
  }
}