run-1/checkpoint-2138/trainer_state.json

{
  "best_metric": 0.15833456940979748,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-1/checkpoint-2138",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2138,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "grad_norm": 2.2052836418151855,
      "learning_rate": 5.527461334920752e-05,
      "loss": 0.6278,
      "step": 500
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5090547800064087,
      "learning_rate": 5.1842260309639244e-05,
      "loss": 0.6192,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.86550760269165,
      "learning_rate": 4.840990727007098e-05,
      "loss": 0.6103,
      "step": 1500
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.385833740234375,
      "learning_rate": 4.49775542305027e-05,
      "loss": 0.5852,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.6027922034263611,
      "eval_matthews_correlation": 0.15833456940979748,
      "eval_runtime": 0.7524,
      "eval_samples_per_second": 1386.172,
      "eval_steps_per_second": 87.716,
      "step": 2138
    }
  ],
  "logging_steps": 500,
  "max_steps": 8552,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 32633550639216.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 5.87069663887758e-05,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 4,
    "seed": 11
  }
}