Hello,
I’m struggling to create the databunch for the food-101 dataset.My data set is saved in google drive and follows this structure:
Images\class[1]
…
Images\class[n]
but len(data.classes) it is 22. (it should be 101)
This is my code:
**
bs=1024
tfms = get_transforms()
data = (ImageList.from_folder(path=path/‘food-101/images’)
.split_by_rand_pct>(0.5)
.label_from_folder()
.transform(tfms, size=32)
.databunch(bs=bs)
)
**
(path/‘food-101/images’).ls(). #Shows the 101 classes