I remember definitely being able to pass a list to class_weight with keras (binary image segmentation specifically). For example:

class_weight = [1, 10] (1:10 class weighting)

But now it's saying it has to take a dictionary instead of a list.

weights = {0: 1, 1: 10}

Well, when I do that I get the error:

`class_weight` not supported for 3+ dimensional targets.

Has something changed? I didn't think I'd have so many issues here.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.