run-4/checkpoint-5052/trainer_state.json

{
  "best_metric": 0.9071100917431193,
  "best_model_checkpoint": "Tech-oriented/best_model_bert_uncasedbert-base-uncased-finetuned-sst2/run-4/checkpoint-5052",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 5052,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9025229357798165,
      "eval_loss": 0.38704702258110046,
      "eval_runtime": 2.3419,
      "eval_samples_per_second": 372.352,
      "eval_steps_per_second": 23.486,
      "step": 1684
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9059633027522935,
      "eval_loss": 0.4139329493045807,
      "eval_runtime": 2.3503,
      "eval_samples_per_second": 371.023,
      "eval_steps_per_second": 23.402,
      "step": 3368
    },
    {
      "epoch": 2.5,
      "grad_norm": 20.706912994384766,
      "learning_rate": 2.349164848148445e-06,
      "loss": 0.4762,
      "step": 4209
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9071100917431193,
      "eval_loss": 0.441041499376297,
      "eval_runtime": 2.3527,
      "eval_samples_per_second": 370.641,
      "eval_steps_per_second": 23.378,
      "step": 5052
    }
  ],
  "logging_steps": 4209,
  "max_steps": 5052,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 208766802986220.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.4078269054384274e-05,
    "num_train_epochs": 3,
    "per_device_train_batch_size": 4,
    "seed": 34,
    "weight_decay": 0.446398905246141
  }
}