run-0/checkpoint-1070/trainer_state.json

{
  "best_metric": 0.4809059900216985,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-0/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": 3.8641507625579834,
      "learning_rate": 3.8123782979849475e-05,
      "loss": 0.553,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.48692557215690613,
      "eval_matthews_correlation": 0.40760885904937505,
      "eval_runtime": 0.3431,
      "eval_samples_per_second": 3040.316,
      "eval_steps_per_second": 192.388,
      "step": 535
    },
    {
      "epoch": 1.87,
      "grad_norm": 10.785917282104492,
      "learning_rate": 4.6818680852446724e-06,
      "loss": 0.3085,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5644384026527405,
      "eval_matthews_correlation": 0.4809059900216985,
      "eval_runtime": 0.392,
      "eval_samples_per_second": 2660.922,
      "eval_steps_per_second": 168.38,
      "step": 1070
    }
  ],
  "logging_steps": 500,
  "max_steps": 1070,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 85536683075904.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 7.156569787445428e-05,
    "num_train_epochs": 2,
    "per_device_train_batch_size": 16,
    "seed": 15
  }
}