Training data versus validation data in ImageClassifierData.from_csv

Hi,

ImageClassifierData.from_csv takes following parameters.
path, folder, csv_fname, bs=64, tfms=(None, None), val_idxs=None, suffix=’’, test_name=None, continuous=False, skip_header=True, num_workers=8, cat_separator=’ ’

we set the val_idxs using the get_cv_idxs.

Are we sure that the training data excludes the images listed in the array val_idxs ?