run-0/checkpoint-2138/trainer_state.json

{
  "best_metric": 0.41870797137315424,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-0/checkpoint-2138",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 2138,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.47,
      "grad_norm": 5.35684061050415,
      "learning_rate": 2.775809181053764e-06,
      "loss": 0.5956,
      "step": 500
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.188376426696777,
      "learning_rate": 2.263099690648562e-06,
      "loss": 0.5375,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5379385948181152,
      "eval_matthews_correlation": 0.29871168199754417,
      "eval_runtime": 0.7649,
      "eval_samples_per_second": 1363.614,
      "eval_steps_per_second": 86.288,
      "step": 1069
    },
    {
      "epoch": 1.4,
      "grad_norm": 10.391807556152344,
      "learning_rate": 1.7503902002433598e-06,
      "loss": 0.4937,
      "step": 1500
    },
    {
      "epoch": 1.87,
      "grad_norm": 19.357559204101562,
      "learning_rate": 1.2376807098381578e-06,
      "loss": 0.4738,
      "step": 2000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.526983916759491,
      "eval_matthews_correlation": 0.41870797137315424,
      "eval_runtime": 0.7468,
      "eval_samples_per_second": 1396.551,
      "eval_steps_per_second": 88.372,
      "step": 2138
    }
  ],
  "logging_steps": 500,
  "max_steps": 3207,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 75492963513744.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 3.288518671458966e-06,
    "num_train_epochs": 3,
    "per_device_train_batch_size": 8,
    "seed": 24
  }
}