run-3/checkpoint-1070/trainer_state.json

{
  "best_metric": 0.4998717031733986,
  "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,
      "grad_norm": 5.447902679443359,
      "learning_rate": 1.6748372574003986e-05,
      "loss": 0.5233,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.51726233959198,
      "eval_matthews_correlation": 0.3954714111531051,
      "eval_runtime": 0.832,
      "eval_samples_per_second": 1253.624,
      "eval_steps_per_second": 79.328,
      "step": 535
    },
    {
      "epoch": 1.87,
      "grad_norm": 4.607208251953125,
      "learning_rate": 2.0568176845268055e-06,
      "loss": 0.3131,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5029475688934326,
      "eval_matthews_correlation": 0.4998717031733986,
      "eval_runtime": 0.7988,
      "eval_samples_per_second": 1305.636,
      "eval_steps_per_second": 82.619,
      "step": 1070
    }
  ],
  "logging_steps": 500,
  "max_steps": 1070,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 85092192859320.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 3.143992746348117e-05,
    "num_train_epochs": 2,
    "per_device_train_batch_size": 16,
    "seed": 29
  }
}