run-4/checkpoint-1605/trainer_state.json

{
  "best_metric": 0.5033485542799104,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-cola/run-4/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,
      "learning_rate": 3.177733965790665e-05,
      "loss": 0.5312,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5129074454307556,
      "eval_matthews_correlation": 0.42602448983421265,
      "eval_runtime": 0.5399,
      "eval_samples_per_second": 1932.01,
      "eval_steps_per_second": 122.256,
      "step": 535
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.739845293487197e-05,
      "loss": 0.3098,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5732018351554871,
      "eval_matthews_correlation": 0.482985877914811,
      "eval_runtime": 0.724,
      "eval_samples_per_second": 1440.52,
      "eval_steps_per_second": 91.155,
      "step": 1070
    },
    {
      "epoch": 2.8,
      "learning_rate": 3.019566211837284e-06,
      "loss": 0.1857,
      "step": 1500
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.7172908782958984,
      "eval_matthews_correlation": 0.5033485542799104,
      "eval_runtime": 0.6596,
      "eval_samples_per_second": 1581.298,
      "eval_steps_per_second": 100.063,
      "step": 1605
    }
  ],
  "logging_steps": 500,
  "max_steps": 1605,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 128263369826388.0,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 4.615622638094134e-05,
    "num_train_epochs": 3,
    "per_device_train_batch_size": 16,
    "seed": 31
  }
}