trainer_state.json

{
  "best_metric": 0.7698666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0244/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": 207.54774475097656,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 3.8548,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3384,
      "eval_loss": 3.7364556789398193,
      "eval_runtime": 7.9785,
      "eval_samples_per_second": 470.014,
      "eval_steps_per_second": 7.395,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 89.83267211914062,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 3.38,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5514666666666667,
      "eval_loss": 2.980492115020752,
      "eval_runtime": 7.9905,
      "eval_samples_per_second": 469.309,
      "eval_steps_per_second": 7.384,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 396.32965087890625,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 2.4638,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6517333333333334,
      "eval_loss": 1.957831621170044,
      "eval_runtime": 8.0088,
      "eval_samples_per_second": 468.234,
      "eval_steps_per_second": 7.367,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 292.3949890136719,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 1.6829,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.716,
      "eval_loss": 1.3760690689086914,
      "eval_runtime": 7.9735,
      "eval_samples_per_second": 470.307,
      "eval_steps_per_second": 7.4,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 89.15348052978516,
      "learning_rate": 8.571428571428571e-06,
      "loss": 1.2868,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7482666666666666,
      "eval_loss": 1.1482754945755005,
      "eval_runtime": 8.0124,
      "eval_samples_per_second": 468.026,
      "eval_steps_per_second": 7.364,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 370.8469543457031,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 1.0977,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7666666666666667,
      "eval_loss": 1.0217229127883911,
      "eval_runtime": 7.982,
      "eval_samples_per_second": 469.809,
      "eval_steps_per_second": 7.392,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 388.7190856933594,
      "learning_rate": 0.0,
      "loss": 1.0129,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7698666666666667,
      "eval_loss": 0.9665048122406006,
      "eval_runtime": 7.9522,
      "eval_samples_per_second": 471.565,
      "eval_steps_per_second": 7.419,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}