flatten conv2d when input_width==kernel_width - #1435
Merged
Conversation
linkerzhang
reviewed
Jul 18, 2019
linkerzhang
approved these changes
Jul 19, 2019
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description:
This change implements a NCHWc optimization for the case input_width==kernel_width. These operations can be flattened to effectively a 1D convolution using striding to allow the NCHWc kernels to better process data.
Motivation and Context
This change addresses a performance bug seen in some internal non-image models using convolution using the above properties. The NCHW MlasConv detected this pattern and converted the convolution to a simple GEMM call. With this fix, models run with the NCHWc optimizer enabled run at the same speed as with "-o 2".
The existing MLAS unit test already covers the new code path for NCHWc thanks to reusing parts of the NCHW tests.