New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update Convolution layer/op weight layout to not depend on input format #6412

Merged
merged 15 commits into from Sep 12, 2018

Conversation

Projects
None yet
3 participants
@AlexDBlack
Copy link
Member

AlexDBlack commented Sep 11, 2018

Fixes: #6393

}
ConvolutionUtils<T>::mkldnn_conv2d(*block.getMKLDNNStream(), {input, weights, bias}, output, {kH,kW,sH,sW,pH,pW,dH,dW,isSameMode,isNCHW});
} else {
#endif

This comment has been minimized.

@saudet

saudet Sep 11, 2018

Member

Please don't do that, MKL-DNN does support strides:
https://intel.github.io/mkl-dnn/understanding_memory_formats.html

@AlexDBlack AlexDBlack merged commit 45c84e1 into master Sep 12, 2018

0 of 2 checks passed

codeclimate Code Climate encountered an error attempting to analyze this pull request.
Details
continuous-integration/jenkins/pr-head This commit cannot be built
Details

@AlexDBlack AlexDBlack deleted the shyrma_weights_format branch Sep 12, 2018

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment