run-5/checkpoint-1605/trainer_state.json

{
  "best_metric": 0.48594322653502614,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-5/checkpoint-1605",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1605,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.93,
      "grad_norm": 2.427436590194702,
      "learning_rate": 6.891375784861594e-05,
      "loss": 0.5578,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5254166722297668,
      "eval_matthews_correlation": 0.4033324209850535,
      "eval_runtime": 0.7104,
      "eval_samples_per_second": 1468.246,
      "eval_steps_per_second": 92.909,
      "step": 535
    },
    {
      "epoch": 1.87,
      "grad_norm": 9.167702674865723,
      "learning_rate": 4.790346582159888e-05,
      "loss": 0.356,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5107694864273071,
      "eval_matthews_correlation": 0.4744648934358445,
      "eval_runtime": 1.1779,
      "eval_samples_per_second": 885.459,
      "eval_steps_per_second": 56.031,
      "step": 1070
    },
    {
      "epoch": 2.8,
      "grad_norm": 5.444707870483398,
      "learning_rate": 2.689317379458183e-05,
      "loss": 0.2179,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7949007749557495,
      "eval_matthews_correlation": 0.48594322653502614,
      "eval_runtime": 1.1637,
      "eval_samples_per_second": 896.282,
      "eval_steps_per_second": 56.716,
      "step": 1605
    }
  ],
  "logging_steps": 500,
  "max_steps": 2140,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 128086660386384.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 8.9924049875633e-05,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 16,
    "seed": 26
  }
}