run-3/checkpoint-1070/trainer_state.json

{
  "best_metric": 0.5135858119171381,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-3/checkpoint-1070",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1070,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.93,
      "learning_rate": 3.158795319612201e-05,
      "loss": 0.5306,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.4587918221950531,
      "eval_matthews_correlation": 0.45274767660246595,
      "eval_runtime": 0.731,
      "eval_samples_per_second": 1426.72,
      "eval_steps_per_second": 90.281,
      "step": 535
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.879222322330773e-06,
      "loss": 0.3002,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5364896059036255,
      "eval_matthews_correlation": 0.5135858119171381,
      "eval_runtime": 1.1712,
      "eval_samples_per_second": 890.526,
      "eval_steps_per_second": 56.352,
      "step": 1070
    }
  ],
  "logging_steps": 500,
  "max_steps": 1070,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 85817141396496.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 5.9296684069913246e-05,
    "num_train_epochs": 2,
    "per_device_train_batch_size": 16,
    "seed": 20
  }
}