The number of subfolders doesn't match the number of specified classes when using custom Dataset

[foreverls]

I am following the tutorial here to use my own dataset for finetuning.
https://mmclassification.readthedocs.io/en/latest/tutorials/new_dataset.html#an-example-of-customized-dataset

The dataset structure is similar to below, but only two classes(Female/Male). and same for test and val folder.

the config looks like

dataset_type = 'CustomDataset'
classes = ['Female', 'Male']  # The category names of your dataset

data = dict(
    train=dict(
        type=dataset_type,
        data_prefix='data/gender_dataset/train',
        classes=classes,
        pipeline=train_pipeline
    ),
    val=dict(
        type=dataset_type,
        data_prefix='data/gender_dataset/val',
        classes=classes,
        pipeline=test_pipeline
    ),
    test=dict(
        type=dataset_type,
        data_prefix='data/gender_dataset/test',
        classes=classes,
        pipeline=test_pipeline
    )
)

but when I train it, it will show error: AssertionError: The number of subfolders (3) doesn't match the number of specified classes (2). Please check the data folder.
May I know what is wrong with the folder structure? Shall I also create an annotation file even with such folder structure? thanks a lot!

[Ezra-Yu]

The dataset structure is similar to below, but only two classes(Female/Male). and same for test and val folder.

Check there are only 2 sub-folders, both in your train/val/test dataset root.

[foreverls]

thanks for replying! here is my folder structure.

[foreverls]

I guess I know why...there is a hidden folder inside...Thanks!