run-4/checkpoint-2138/trainer_state.json

{
  "best_metric": 0.08036809130702588,
  "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,
      "grad_norm": 4.632926940917969,
      "learning_rate": 1.2946011611760917e-06,
      "loss": 0.6259,
      "step": 500
    },
    {
      "epoch": 0.47,
      "grad_norm": 5.080592155456543,
      "learning_rate": 1.2310780423745168e-06,
      "loss": 0.6096,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.2639031410217285,
      "learning_rate": 1.1675549235729423e-06,
      "loss": 0.5698,
      "step": 1500
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.567307472229004,
      "learning_rate": 1.1040318047713676e-06,
      "loss": 0.5589,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5995281934738159,
      "eval_matthews_correlation": 0.08036809130702588,
      "eval_runtime": 0.3414,
      "eval_samples_per_second": 3055.256,
      "eval_steps_per_second": 193.334,
      "step": 2138
    }
  ],
  "logging_steps": 500,
  "max_steps": 10690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 32410011903984.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.3581242799776662e-06,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 4,
    "seed": 5
  }
}