vinx13 edited a comment on pull request #9465: URL: https://github.com/apache/tvm/pull/9465#issuecomment-963464553
I'm confused. Currently TOPI implementations expect input of NCHW and kernel of IOHW for `conv2d_transpose`, so the shape checking in Relay in actually correct, isn't it? (while `kernel_layout` in Relay attrs is still `OIHW` it has different meaning for `conv2d_transpose`) https://github.com/apache/tvm/blob/main/python/tvm/topi/nn/conv2d_transpose.py#L33-L37 -- 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. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
