Assertion `indexValue >= 0 && indexValue < tensor.sizes[dim]` failed about yowo

[wei-tim]

@nkise
Hi,
that's true, you need to modify this part if you train with your own dataset.

[okankop]

Currently I am refactoring the whole repo. So, there won't be these kind of bugs. More importantly, I will add the training details for the AVA dataset, and also update the article with AVA results.

Stay tuned!

[nkise]

@wei-tim it seems that you have hardcoded numbers of class in region_loss.py

        # try focal loss with gamma = 2
        FL = FocalLoss(class_num=24, gamma=2, size_average=False)
        loss_cls = self.class_scale * FL(cls, tcls)

[nkise]

@nkise
Hi,
that's true, you need to modify this part if you train with your own dataset.

So, as I understand, I need to modify clues_num to my number of classes?

[wei-tim]

@nkise

exactly.

[hfarhid]

Did this fix you problem? still have same issue after fixing the class_num