-
Notifications
You must be signed in to change notification settings - Fork 66
Open
Description
Hi, thanks for sharing your research.
I need to stack vertices and faces in a batch, but I note that all your experiments in the repository have batch size=None. Have you tried stacking vertices and faces in a batch, adding padding like pytorch3d to avoid shape differences? Can batch size and padding produce an error or bad behavior in diffusion-net?
Metadata
Metadata
Assignees
Labels
No labels