trainer_state.json

{
  "best_metric": 0.9530666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0562/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.7113024592399597,
      "learning_rate": 2.8858192987669303e-05,
      "loss": 1.1467,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.932,
      "eval_loss": 0.2794320285320282,
      "eval_runtime": 10.1815,
      "eval_samples_per_second": 368.314,
      "eval_steps_per_second": 5.795,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.660823941230774,
      "learning_rate": 2.5606601717798212e-05,
      "loss": 0.1499,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9442666666666667,
      "eval_loss": 0.1933838576078415,
      "eval_runtime": 10.014,
      "eval_samples_per_second": 374.474,
      "eval_steps_per_second": 5.892,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.12702327966690063,
      "learning_rate": 2.074025148547635e-05,
      "loss": 0.0502,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9466666666666667,
      "eval_loss": 0.19058921933174133,
      "eval_runtime": 10.0413,
      "eval_samples_per_second": 373.458,
      "eval_steps_per_second": 5.876,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.187829494476318,
      "learning_rate": 1.5e-05,
      "loss": 0.0177,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9514666666666667,
      "eval_loss": 0.18742536008358002,
      "eval_runtime": 10.0582,
      "eval_samples_per_second": 372.83,
      "eval_steps_per_second": 5.866,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.04160599410533905,
      "learning_rate": 9.259748514523654e-06,
      "loss": 0.0083,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9514666666666667,
      "eval_loss": 0.19787722826004028,
      "eval_runtime": 10.0284,
      "eval_samples_per_second": 373.938,
      "eval_steps_per_second": 5.883,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.04516437277197838,
      "learning_rate": 4.393398282201788e-06,
      "loss": 0.0053,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9528,
      "eval_loss": 0.19119274616241455,
      "eval_runtime": 10.0628,
      "eval_samples_per_second": 372.66,
      "eval_steps_per_second": 5.863,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.1485859908123557e-06,
      "loss": 0.0039,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9530666666666666,
      "eval_loss": 0.192170649766922,
      "eval_runtime": 10.7807,
      "eval_samples_per_second": 347.843,
      "eval_steps_per_second": 5.473,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}