Can't get dls.vocab appropriate texts

Hi, so I have a dataset with its own train and test folders. I created a datablock and I’m trying to get the vocab after executing the datablock but i get random letters instead. I’ve tried writting a function to get the parent label but still no avail. below is screenshot of my code

I tried using Imagedataloader.from_folder which worked fine but still having a problem using datablock.

You’re using multicategory block, which expects a list. You can see an example here where I use the multicategory block with PETs

(I should note that multicategory in this context means multiple labels present in each y, if it’s a single label in each y you should use CategoryBlock)

Thanks alot @muellerzr . I was using MultiCategoryBlock instead of CategoryBlock.

Hey @muellerzr . I’m using grandparent funtion
splitter=GrandparentSplitter(train_name=(train, test), valid_name=preds)
where train is a directory for train files, test is a directory for test files and preds is a directory for prediction images. is using grandparent this way valid and if not how can I get the three directories inside a datablock.