Open
Conversation
Allow channelwise convolution prefuse pass to match half_type in addition to float_type.
Contributor
There was a problem hiding this comment.
Pull request overview
Extends the GPU channelwise-convolution prefuse pattern so FP16 (half) convolutions can be recognized and rewritten to gpu::channelwise_conv, matching the existing FP32 behavior.
Changes:
- Allow
find_channelwise_convolutioningpu::prefuse_opsto matchshape::half_typeinputs in addition toshape::float_type.
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
Comment on lines
+304
to
+305
| if(input->get_shape().type() != shape::float_type and | ||
| input->get_shape().type() != shape::half_type) |
There was a problem hiding this comment.
This condition calls input->get_shape().type() twice and hard-codes two comparisons. For readability and to avoid repeating work, consider caching the type in a local (e.g., auto t = …) and checking membership (e.g., via contains({shape::float_type, shape::half_type}, t)) as done in other GPU type gates (e.g., src/targets/gpu/lowering.cpp:349).
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## develop #4808 +/- ##
===========================================
+ Coverage 92.46% 92.52% +0.05%
===========================================
Files 583 583
Lines 29551 29967 +416
===========================================
+ Hits 27323 27724 +401
- Misses 2228 2243 +15 🚀 New features to boost your workflow:
|
Collaborator
Check flagged results 🔆 * No develop baseline was found for this PR's branch point; compared against the latest available develop run instead. |
Collaborator
|
Contributor
Author
Collaborator
|
On Navi48, fp16 is sometimes slower:
|
weizhu12-amd
force-pushed
the
enable_fp16_channelwise
branch
from
ef2fc84 to
85b943f
Compare
Allow channelwise convolution prefuse pass to match half_type in addition to float_type. For small shape(CWH < 48k), still use mlir_conv.
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
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.
Motivation
Extend the channelwise convolution prefuse pass to match half_type in addition to float_type.
Technical Details
Allow the channelwise convolution prefuse pass to recognize half_type.
Changelog Category
Add a
CHANGELOG.mdentry for any option other thanNot Applicable