run-2/checkpoint-1605/trainer_state.json

{
  "best_metric": 0.38853046110587136,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-2/checkpoint-1605",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1605,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.93,
      "grad_norm": 5.214674472808838,
      "learning_rate": 1.8520669838507643e-06,
      "loss": 0.6015,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5738183259963989,
      "eval_matthews_correlation": 0.0,
      "eval_runtime": 0.3412,
      "eval_samples_per_second": 3056.962,
      "eval_steps_per_second": 193.441,
      "step": 535
    },
    {
      "epoch": 1.87,
      "grad_norm": 4.423364162445068,
      "learning_rate": 1.28741241560358e-06,
      "loss": 0.5393,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5461441874504089,
      "eval_matthews_correlation": 0.3010425673469933,
      "eval_runtime": 0.3638,
      "eval_samples_per_second": 2866.982,
      "eval_steps_per_second": 181.42,
      "step": 1070
    },
    {
      "epoch": 2.8,
      "grad_norm": 4.518457412719727,
      "learning_rate": 7.227578473563959e-07,
      "loss": 0.4865,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.5234982967376709,
      "eval_matthews_correlation": 0.38853046110587136,
      "eval_runtime": 0.3431,
      "eval_samples_per_second": 3039.887,
      "eval_steps_per_second": 192.361,
      "step": 1605
    }
  ],
  "logging_steps": 500,
  "max_steps": 2140,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 128596608126132.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 2.4167215520979487e-06,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 16,
    "seed": 33
  }
}