[OpenVINO backend] Fix and enable numpy.rot90

[PES2UG23CS205]

This PR fixes and enables numpy.rot90 support for the OpenVINO backend in Keras.

The implementation follows NumPy semantics and aligns with existing OpenVINO backend patterns by expressing rot90 using supported OpenVINO ops.

What was done

Implemented numpy.rot90 in keras/src/backend/openvino/numpy.py

Normalized and validated k and axes to match NumPy behavior

Implemented rotation using transpose and reverse OpenVINO ops

Removed NumPyTestRot90 from excluded_concrete_tests.txt

Fixed OpenVINO test failures related to rot90

[gemini-code-assist]

Summary of Changes

Hello @PES2UG23CS205, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly enhances the Keras OpenVINO backend by introducing robust and correctly functioning support for the numpy.rot90 operation. The changes ensure that array rotations are performed accurately and efficiently, adhering to NumPy's specifications while leveraging OpenVINO's optimized operations. This integration resolves previous test failures and expands the backend's compatibility with common numerical operations.

Highlights

• Numpy.rot90 Support: Enabled and fixed numpy.rot90 functionality for the OpenVINO backend in Keras, aligning its behavior with standard NumPy semantics.
• Implementation Details: The rot90 operation is now implemented using native OpenVINO transpose and reverse operations, ensuring efficient execution on the OpenVINO backend.
• Parameter Handling: The k and axes parameters for rot90 are now properly normalized and validated to match NumPy's expected behavior.
• Test Coverage: The NumPyTestRot90 test case has been removed from the list of excluded concrete tests, indicating that the rot90 implementation is now passing all relevant tests and is fully functional.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

The pull request removes the NumPyTestRot90 entry from the keras/src/backend/openvino/excluded_concrete_tests.txt file, indicating that this test is no longer excluded for the OpenVINO backend. There were no review comments provided.

[hertschuh]

@PES2UG23CS205

Something weird happened here. Github shows that both files are fully changed. My guess is that the line breaks changed from Unix still to DOS style. Can you undo this and rebase? Thanks!

[PES2UG23CS205]

Thanks for pointing this out.

You’re right, this looks like a line-ending (CRLF vs LF) issue.
I’ll undo the line-ending changes, rebase on top of master, and push an updated commit shortly.

[PES2UG23CS205]

Thanks for pointing this out it was a CRLF vs LF issue.
I’ve reset the files, rebased on top of master, and force-pushed a clean commit with only the intended changes.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 25.92593% with 20 lines in your changes missing coverage. Please review.
✅ Project coverage is 82.67%. Comparing base (0e84cb5) to head (d97932a).
⚠️ Report is 5 commits behind head on master.

Files with missing lines	Patch %	Lines
keras/src/backend/openvino/numpy.py	25.92%	20 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master   #21967      +/-   ##
==========================================
- Coverage   82.68%   82.67%   -0.02%     
==========================================
  Files         588      588              
  Lines       61491    61582      +91     
  Branches     9629     9640      +11     
==========================================
+ Hits        50846    50915      +69     
- Misses       8152     8174      +22     
  Partials     2493     2493              

Flag	Coverage Δ
keras	82.50% <25.92%> (-0.01%)	⬇️
keras-jax	61.46% <0.00%> (-0.06%)	⬇️
keras-numpy	56.73% <0.00%> (-0.05%)	⬇️
keras-openvino	37.45% <25.92%> (+0.06%)	⬆️
keras-tensorflow	63.63% <3.70%> (-0.06%)	⬇️
keras-torch	62.38% <3.70%> (-0.05%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[PES2UG23CS205]

I attempted to run api_gen.py locally, but hit a known TensorFlow/JAX dependency conflict (ml_dtypes version mismatch).
Since CI is already generating the API changes successfully, could a maintainer please apply the api_gen updates or advise on the preferred workflow here?

[hertschuh]

Rename x to array. This is what's making the tests fail.

[PES2UG23CS205]

Thanks for the review!

I’ve renamed the argument from x to array to match NumPy’s signature and pushed an updated commit.

[hertschuh]

I attempted to run api_gen.py locally, but hit a known TensorFlow/JAX dependency conflict (ml_dtypes version mismatch). Since CI is already generating the API changes successfully, could a maintainer please apply the api_gen updates or advise on the preferred workflow here?

I can run it myself. Just curious though, can't you upgrade TensorFlow?

[PES2UG23CS205]

I’ve fixed the OpenVINO flip implementation to derive the rank from the underlying OpenVINO tensor instead of OpenVINOKerasTensor, and pushed the update.