RuntimeError: Inferred shape and existing shape differ in dimension 2 #626
Unanswered
ElMahdiElAnnabi
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi everyone,
i have build a brevitas version of squeezenet (Pytorch code) and exported it as onnx to finn. Unfortunately i become every time when i try to do the streamlining this error (RuntimeError: Inferred shape and existing shape differ in dimension 2: (16 ) & (36)) I believe, it is caused by MaxPool2d when ceil_mode= False. With ceil_mode=true i become uneven pads, which are not supported in finn as another error said. Is there any suggestion how to overcome this dilemma? here are the model in pytorch and onnx
thanks in advance
model.zip
Beta Was this translation helpful? Give feedback.
All reactions