Cudnn cudnnConvolutionBiasActivationForward question and suggest

AI & Data Science Deep Learning (Training & Inference) cuDNN
1

cudnnConvolutionBiasActivationForward not support x point to the same buffer as z or y
I understand x != y, but why x can not be same with z (z and x are both read only)? Some skip layer may be this case.

2

@AakankshaS

3

Hi @liujunjie ,
Apologies for delayed response.
This is happening because dim(x) != dim(z)
Hope this helps.

Thanks!