optima2005 commented on issue #4400: implement conv3d op URL: https://github.com/apache/incubator-tvm/pull/4400#issuecomment-557778113 @masahi, glad to contribute! Many thanks for the review! For "unify 2D and 3D" cudnn convolution, I am wondering whether it is better to keep separation. The APIs are 2 groups in cudnn lib(2D vs ND), I guess whether there would be some specific optimazation for 2D group. What do you think about that? For other comments, I have revised in the latest commit, please check again, thanks a lot!
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
