NVIDIA/TransformerEngine

Can TE optimize the find cudnn?

MARD1NO opened this issue · 1 comments

Maybe user do not install cuda/cudnn toolkit independently, but pytorch will pack their own cudnn in installation package.

I think TE can set the default cudnn path by pytorch install dir, otherwise user need to specify the cudnn include directory by setting ENV_VAR CUDNN_INCLUDE_DIR.

Hi @MARD1NO, could you please give #831 a try? Thanks!