What's the difference between ds_tfms and dl_tfms?

I am looking into the parameters of ImageDataBunch and I always forget which one is the parameter where I need to pass get_transform().

Can someone explain what’s the difference between ds_tfms and dl_tfms? It would help to know what’s the meaning of ds and dl :slight_smile:

ds stands by DataSet while dl stands by DataLoader.

As far as I known, datasets contains the data: train, validation and optionally test sets. DataLoader iterates over this data in order to provide to the model chunks of specific size to fit it (those chunks are called mini-batches)

1 Like