CU-Net icon indicating copy to clipboard operation
CU-Net copied to clipboard

Published 20 hours ago • verified publisher icon zhiqiangdon

Encounter Segmentation fault when running loss.backword()

Open cenaliu opened this issue 5 years ago • 1 comments

HI, thanks for your impressive code. When I used your code to train a native dense unet, loss.backword caused Segmentation fault, while the forward is ok. Can you help me, Thanks very much!

cenaliu avatar Aug 30 '18 13:08 cenaliu

Please make sure your Pytorch version matches the corresponding script. Script names with "prev-version" require Pytorch v0.1.12.

zhiqiangdon avatar Sep 01 '18 20:09 zhiqiangdon