[OpenVINO] Fix ndim() to return Python int instead of graph tensor

[goyaladitya05]

ndim() in the OpenVINO backend was building an OV graph node to represent the rank, returning an OpenVINOKerasTensor. However, the number of dimensions is a static, compile-time property of a tensor's type, not a runtime value. It never changes during inference and must be a Python int for use in Python control flow.

All other backends return an int.

This PR adds a fix which reads the rank directly from the OV partial shape, get_partial_shape().rank.get_length() which is always known at graph-build time.

[gemini-code-assist]

Code Review

This pull request removes a test from the OpenVINO excluded tests list and refactors the ndim function in the OpenVINO backend to use get_partial_shape().rank.get_length(). While the refactor simplifies the code, the current implementation is fragile because get_length() will raise a RuntimeError if the tensor has a dynamic rank. It is recommended to handle dynamic ranks by checking is_static or returning None to ensure robustness.

[goyaladitya05]

/gemini review

[gemini-code-assist]

Code Review

This pull request refactors the ndim implementation in the OpenVINO backend to use the tensor's partial shape rank and re-enables the test_normalization_kpl test. The review feedback suggests optimizing the ndim function by checking for the attribute directly on the input object to avoid unnecessary backend operations and recommends clarifying the error message regarding static rank requirements.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 83.33333% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 83.29%. Comparing base (6c5956d) to head (91935a7).
⚠️ Report is 6 commits behind head on master.

Files with missing lines	Patch %	Lines
keras/src/backend/openvino/numpy.py	83.33%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #22608   +/-   ##
=======================================
  Coverage   83.28%   83.29%           
=======================================
  Files         596      596           
  Lines       68102    68139   +37     
  Branches    10610    10614    +4     
=======================================
+ Hits        56720    56753   +33     
- Misses       8636     8639    +3     
- Partials     2746     2747    +1     

Flag	Coverage Δ
keras	83.09% <83.33%> (+<0.01%)	⬆️
keras-jax	59.66% <16.66%> (+<0.01%)	⬆️
keras-numpy	55.33% <0.00%> (+<0.01%)	⬆️
keras-openvino	53.41% <83.33%> (+0.02%)	⬆️
keras-tensorflow	61.03% <16.66%> (+<0.01%)	⬆️
keras-torch	59.85% <16.66%> (+<0.01%)	⬆️

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.

[hertschuh]

Thanks for the fix!