run-0/checkpoint-2138/trainer_state.json

{
  "best_metric": 0.429931898675694,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-0/checkpoint-2138",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2138,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23,
      "grad_norm": 7.291004657745361,
      "learning_rate": 1.351268050159641e-05,
      "loss": 0.5812,
      "step": 500
    },
    {
      "epoch": 0.47,
      "grad_norm": 14.129287719726562,
      "learning_rate": 1.284964416687627e-05,
      "loss": 0.5391,
      "step": 1000
    },
    {
      "epoch": 0.7,
      "grad_norm": 4.682578086853027,
      "learning_rate": 1.2186607832156134e-05,
      "loss": 0.5557,
      "step": 1500
    },
    {
      "epoch": 0.94,
      "grad_norm": 1.5674272775650024,
      "learning_rate": 1.1523571497435995e-05,
      "loss": 0.5335,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5571191906929016,
      "eval_matthews_correlation": 0.429931898675694,
      "eval_runtime": 0.7111,
      "eval_samples_per_second": 1466.739,
      "eval_steps_per_second": 92.814,
      "step": 2138
    }
  ],
  "logging_steps": 500,
  "max_steps": 10690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 32615957312832.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.4175716836316547e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 4,
    "seed": 20
  }
}