Jackie

https://github.com/lucidrains/byol-pytorch/blob/2aa84ee18fafecaf35637da4657f92619e83876d/byol_pytorch/byol_pytorch.py#L178 throw this error: torchvision.transforms.transforms.RandomGrayscale is not a Module subclass should use torchvision.transforms.Compose ?

out of memory

2

I got cuda out of memory every time I continued training. But I won't get error if I load initial weight and train from the first epoch. I used my...

Did you implement threshold tc mentioned in formula (4) in this repo? I think the line below only implements Δ https://github.com/cornell-zhang/dnn-gating/blob/31666fadf35789b433c79eec8669a3a2df818bd4/utils/cg_utils.py#L173

Did you implement two separate BN for base path (Xp) and conditional path (Xr) as figure 4 mentioned in your paper? I only find one BN here. https://github.com/cornell-zhang/dnn-gating/blob/31666fadf35789b433c79eec8669a3a2df818bd4/utils/cg_utils.py#L173 Also it...