run-3/checkpoint-2638/trainer_state.json

{
  "best_metric": 0.7350265352539803,
  "best_model_checkpoint": "distilbert-base-uncased-finetuned-multiclass-classification/run-3/checkpoint-2638",
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2638,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18953752843062927,
      "grad_norm": 9.870668411254883,
      "learning_rate": 1.490581283927063e-05,
      "loss": 5.8661,
      "step": 500
    },
    {
      "epoch": 0.37907505686125853,
      "grad_norm": 11.824650764465332,
      "learning_rate": 1.4318507368850198e-05,
      "loss": 4.7233,
      "step": 1000
    },
    {
      "epoch": 0.5686125852918877,
      "grad_norm": 11.662109375,
      "learning_rate": 1.3731201898429762e-05,
      "loss": 3.8044,
      "step": 1500
    },
    {
      "epoch": 0.7581501137225171,
      "grad_norm": 11.159646987915039,
      "learning_rate": 1.3143896428009328e-05,
      "loss": 3.0524,
      "step": 2000
    },
    {
      "epoch": 0.9476876421531463,
      "grad_norm": 13.040279388427734,
      "learning_rate": 1.2556590957588894e-05,
      "loss": 2.5575,
      "step": 2500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7350265352539803,
      "eval_loss": 2.0373177528381348,
      "eval_runtime": 41.2753,
      "eval_samples_per_second": 63.912,
      "eval_steps_per_second": 3.998,
      "step": 2638
    }
  ],
  "logging_steps": 500,
  "max_steps": 13190,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 84227996516352.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.5493118309691066e-05,
    "num_train_epochs": 5,
    "per_device_train_batch_size": 4,
    "seed": 29
  }
}