VideoSwin Implementation #2
Description
I really appreciate your amazing work. However, it seems that the implementation of STA on VideoSwin has not been released here. It would be great if we could see how the fill-up operation for windowed attention layers is implemented. Thanks a lot!