Post-Release 0.52.0 Merge & Bump Master to 0.53.0.dev

[vkarampudi]

Overview

This pull request merges the 0.52.0 release updates into master and bumps the development version to 0.53.0.dev across package metadata and version strings, maintaining standard TFX / TensorFlow post-release branching conventions.

---

Summary of Changes

1. Development Version Bump (0.53.0.dev)

• Bumped __version__ from 0.49.0.dev to 0.53.0.dev across:
  • fairness_indicators/version.py
  • tensorboard_plugin/tensorboard_plugin_fairness_indicators/version.py

2. Release Notes & Documentation (RELEASE.md)

• Preserved the # Current Version (Still in Development) header at the top of RELEASE.md and tensorboard_plugin/RELEASE.md for future development.
• Inserted the complete # Version 0.52.0 release notes and updated compatibility tables right below Current Version while preserving all historical release sections.

3. Python Support Matrix (3.10 – 3.13)

• Deprecated and removed Python 3.9 across package Trove classifiers and CI workflows.
• Updated setup classifiers and GitHub Actions workflows (build.yml, docs.yml, test.yml) to officially test and support Python ['3.10', '3.11', '3.12', '3.13'].

4. Dependency & Ecosystem Alignment

Aligned constraints in setup.py and tensorboard_plugin/setup.py with TF 2.21 / TFX 0.52:

• TensorFlow: tensorflow>=2.21,<2.22 (and tf-keras>=2.21,<2.22 in plugin).
• TF Model Analysis (TFMA): tensorflow-model-analysis>=0.52.0,<0.53.0.
• TF Data Validation (TFDV): tensorflow-data-validation>=1.21.0,<1.22.0.
• Protobuf: protobuf>=6.0.0,<7.0.0 (aligned with TFDV 1.21.0 to prevent pip resolution conflicts).
• PEP 508 NumPy Markers: Added conditional NumPy constraints (numpy>=1.23.5,<2.0.0 for Python < 3.13 and numpy>=2.1.0 for Python >= 3.13) to support Python 3.13 while preventing C-extension crashes in older wheels.

5. Code Compatibility & Test Discovery

• TFMA 0.52.0 Protobuf Config Import Fix: Updated test suites and utilities (example_model_test.py, util.py, util_test.py, plugin_test.py) to import config_pb2 directly from tensorflow_model_analysis.proto and use config_pb2.EvalConfig().
• Test Discovery: Added pythonpath = ["."] to pyproject.toml and pytest.ini for clean local pytest execution.
• Ruff Linting: Removed unused imports (F401) to ensure 100% clean pre-commit checks.

6. Build & PyPI Automation Improvements (build.yml)

• Added release: types: [published] and tags: ['*'] triggers so publishing a GitHub release automatically triggers PyPI publishing.
• Simplified packaging to build pristine universal distributions (py3-none-any.whl) without multi-job artifact merging collisions.
• Added skip-existing: true and secret token / OIDC authentication support to handle partial upload resumption cleanly.