FrozenGene edited a comment on pull request #7445: URL: https://github.com/apache/tvm/pull/7445#issuecomment-782072291
> > I worry about we can not ignore it simply. According to doc: https://www.tensorflow.org/api_docs/python/tf/nn/conv2d > > > Either the string "SAME" or "VALID" indicating the type of padding algorithm to use, or a list indicating the explicit paddings at the start and end of each dimension. When explicit padding is used and data_format is "NHWC", this should be in the form [[0, 0], [pad_top,pad_bottom], [pad_left, pad_right], [0, 0]]. When explicit padding used and data_format is "NCHW", this should be in the form [[0, 0], [0, 0],[pad_top, pad_bottom], [pad_left, pad_right]]. > > > > > > If `explicit_padding` is not None, we should apply its values. However, @trevor-m you could make a double check. > > Thanks for the review! That is true, I will update this PR to properly support explicit padding. Thanks @trevor-m Our convolution op ignore `explicit_padding` attribute simply is wrong too (Note: be careful of `conv2d_transpose`: https://www.tensorflow.org/api_docs/python/tf/nn/conv2d_transpose, its explicit padding is not the same as `conv2d`), not only just pooling op. You could correct it too.Thanks. ---------------------------------------------------------------- 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]
