search

lucabergamini / VAEGAN-PYTORCH

VAEGAN from "Autoencoding beyond pixels using a learned similarity metric" implemented in Pytorch. Clean, clear and with comments.
182 stars 34 forks source link

Inplace operation #5

Open zideliu opened 2 years ago

zideliu commented 2 years ago

I couldn't run main.py, because of inplace operation

File "main.py", line 202, in loss_decoder.backward(retain)graph=True)

one of the Variables needed for gradient computation has been modified by an inplace operation:[torch.cuda.FloatTensor [1024,128]], which is output 0 of AsStridedBackward0, is at version3; expected version 2 instead

kzpsycheuu commented 2 years ago

Same error, have you found the way to solve it?

zideliu commented 2 years ago

Same error, have you found the way to solve it?

change pytorch version may be help you

changzhuoqing commented 2 years ago

Same error, have you found the way to solve it?

change pytorch version may be help you Sorry to bother you, I would like to ask which version of pytorch can work normally.