Dls.one_batch()

It pick a batch from train or valid set ???

Validation. You can do dls.train.one_batch() or dls.valid.one_batch() and get more specific

1 Like

thank you