How to use KFoldDataset in dev1.x

[cape-zck]

Thank you for your outstanding contributions. Recently I want to use mmrazor to call mmcls in order to train the model using K-fold cross-validation. However, I find the PR about KFoldDataset using in master like:

    train=dict(
        type='KFoldDataset',
        dataset=dict(
            type=dataset_type,
            data_prefix='data/imagenet/train',
            pipeline=train_pipeline),
        # Modify the `fold` to use different split. For 5-fold cross-validation,
        # five experiments need to be executed (fold=0, fold=1, fold=2, ...)
        fold=0,
        num_splits=5,
        # seed = 1,   # If set seed, shuffle samples before spliting dataset.
    )

I'm having a hard time figuring out how to use this KFoldDataset class in dev1.x. It seems that I can realize it using tools/kfold-cross-valid.py.But if I want to modify the config,can you give me some ideas? Thank you！