FrozenGene commented on issue #4484: [FRONTEND][TF] conv2d_transpose 'SAME' support kernel more than 1x1 URL: https://github.com/apache/incubator-tvm/pull/4484#issuecomment-563760780 @optima2005 do you mind taking up this RFC? https://github.com/apache/incubator-tvm/issues/2682 Because I see you start to support 4D padding in Conv2dTranspose. I think the work should be similar with convolution, which could make us avoid `pad` operator and get potential performance.
---------------------------------------------------------------- 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
