Improve SmoothQuant test cases #3101
Merged
Conversation
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/3101
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 389b91d with merge base 5cbbd73 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
Xia-Weiwen
added
topic: not user facing
|
CC @namgyu-youn I cannot add you as a reviewer. |
jerryzh168
approved these changes
Sep 30, 2025
namgyu-youn
approved these changes
Sep 30, 2025
Thanks. Actually, the cases may fail if we don't. That means alpha=0.5 is worse than 0 with some random inputs. |
meta-cla
bot
added
the
CLA Signed
namgyu-youn
suggested changes
Sep 30, 2025
Yeah outlier is a key to produce SmoothQuant. Thanks for this change |
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Outdated suggestions cannot be applied.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.
Summary
The test cases for SmoothQuant have the following issues:
test_smoothquant_accuracyuses a single linear as the model, but only submodules can be converted toSmoothQuantObservedLinear. So, the single linear is never converted and the case does not actually test the results of SmoothQuant linear. When running withpytest -sv, the following warning messages are printed:convert: module is not SmoothQuantObservedLinear, skipping: <class 'torch.nn.modules.linear.Linear'>This PR fixes the issues.
Test plan
pytest -sv test/prototype/test_smoothquant.py