use torch to compute discrepancies in OnnxDiscrepancyCheck#2540
Merged
Conversation
Contributor
There was a problem hiding this comment.
Pull request overview
This PR updates OnnxDiscrepancyCheck to use PyTorch (instead of NumPy) for element-wise discrepancy computations so the pass can handle bfloat16 models, and adds unit tests for inferring ONNX weight dtypes to align reference-model precision.
Changes:
- Added ONNX weight dtype inference + ONNX→torch dtype mapping utilities and used them to cast the reference HF model to match ONNX weight precision.
- Replaced NumPy-based absolute-difference computation with a torch-based implementation (double precision).
- Added unit tests covering ONNX weight dtype inference and dtype mapping.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 4 comments.
| File | Description |
|---|---|
olive/passes/onnx/discrepancy_check.py |
Adds ONNX weight dtype inference + switches discrepancy math to torch; casts reference model to match inferred ONNX precision. |
test/passes/onnx/test_discrepancy_check.py |
Adds unit tests for ONNX weight dtype inference and ONNX→torch dtype mapping. |
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Collaborator
|
/azp run |
|
Azure Pipelines successfully started running 1 pipeline(s). |
jambayk
approved these changes
Jun 24, 2026
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.
Describe your changes
OnnxDiscrepancyCheck is using numpy to compute discrepancies, it must be switched to torch for bfloat16.