Results 8 comments of FraMan

From your explanation it's hard to give you a complete answer. You can simply replace the backbone on your code with EfficientNet adjsuting the final shape of your tensor equal...

I have the same problem with maskrcnn pytorch

Any news regarding this feature?

Hey, does the tutorial for motion work for you? I found the documentation not well written

Can you explain to me how did you set the environment to test the motion challenge and tutorial? I mean which steps did you follow from the beginning, really appreciate...

Update: It seems that with new versions of proto every Ray version have the same problem.