Skip to content
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

Fix direct Depthwise convolution implementation #367

Merged
merged 2 commits into from
Dec 5, 2021

Conversation

staticfloat
Copy link
Contributor

There was a silly typo bug in the direct depthwise convolution algorithm
that broke multi-channel testsets. Perhaps more disturbingly, our
default test set did not catch this, because multi-channel workloads are
not run unless the fuzzing tests are run.

There was a silly typo bug in the direct depthwise convolution algorithm
that broke multi-channel testsets.  Perhaps more disturbingly, our
default test set did not catch this, because multi-channel workloads are
not run unless the fuzzing tests are run.
test/conv.jl Outdated Show resolved Hide resolved
@DhairyaLGandhi
Copy link
Member

Thanks for catching this @staticfloat , maybe we should have an optional GHA action for fuzzing tests. It lgtm, let me know if you're happy and we can merge this.

Co-authored-by: Carlo Lucibello <carlo.lucibello@unibocconi.it>
@staticfloat staticfloat merged commit 2436b32 into master Dec 5, 2021
@CarloLucibello CarloLucibello deleted the sf/depthwise_convolution_fix branch June 15, 2023 17:05
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants