run-4/checkpoint-2138/trainer_state.json

{
  "best_metric": 0.0,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-4/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,
      "learning_rate": 9.621960181020146e-07,
      "loss": 0.6281,
      "step": 500
    },
    {
      "epoch": 0.47,
      "learning_rate": 6.684853898657462e-07,
      "loss": 0.5843,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.747747616294781e-07,
      "loss": 0.5742,
      "step": 1500
    },
    {
      "epoch": 0.94,
      "learning_rate": 8.106413339321002e-08,
      "loss": 0.5737,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5868913531303406,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.7093,
      "eval_samples_per_second": 1470.475,
      "eval_steps_per_second": 93.05,
      "step": 2138
    }
  ],
  "logging_steps": 500,
  "max_steps": 2138,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 32612852608176.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.2559066463382826e-06,
    "num_train_epochs": 1,
    "per_device_train_batch_size": 4,
    "seed": 28
  }
}