Fix TFP compatibility with Keras 3 (Issue #1945)

[aviralgarg05]

Description

This PR addresses Issue #1945, where tensorflow_probability (TFP) was failing to import in Keras 3 environments due to a missing internal TensorFlow attribute: register_load_context_function.

Changes

1. Compatibility Patch: Introduced a monkey-patch in affected example files to alias tf._api.v2.compat.v2.__internal__.register_load_context_function to register_call_context_function if it's missing. This is a robust workaround for current TFP/Keras 3 version mismatches.
2. Execution Guard: Wrapped top-level script execution (training loops, data loading) in if __name__ == "__main__": blocks. This ensures the files can be safely imported by other scripts or test suites without triggering full training runs.
3. Affected Files:
   • examples/generative/vq_vae.py
   • examples/generative/real_nvp.py
   • examples/keras_recipes/bayesian_neural_networks.py

Verification Results

• Reproduction: Confirmed the import error in a Keras 3 environment.
• Fix: Verified that the modified scripts successfully import tensorflow_probability and build models without crashing.
• Style: All modified files were reformatted using black.

Related Issues

Fixes #1945

[gemini-code-assist]

Summary of Changes

Hello @aviralgarg05, 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 addresses a critical compatibility problem preventing tensorflow_probability from importing correctly in Keras 3 environments. The solution involves a targeted monkey-patch to resolve a missing internal TensorFlow attribute, ensuring seamless integration. Additionally, the changes enhance the robustness of example scripts by preventing their automatic execution when imported, making them more modular and testable.

Highlights

• TFP Compatibility Patch: Introduced a monkey-patch in affected example files to alias tf._api.v2.compat.v2.internal.register_load_context_function to register_call_context_function if the former is missing, resolving TFP import issues with Keras 3 / TensorFlow 2.19+.
• Execution Guard for Examples: Encapsulated top-level script execution (training loops, data loading, and evaluation) within if name == "main": blocks in example files, preventing unintended execution upon import.
• Code Formatting: All modified files were reformatted using black to maintain consistent code style.

🧠 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

This pull request correctly addresses the TFP compatibility issue with a monkey-patch and improves the example scripts by wrapping execution logic in if __name__ == "__main__": guards.

However, there's a critical issue in both examples/generative/vq_vae.py and examples/keras_recipes/bayesian_neural_networks.py. In both files, some of the execution logic that depends on trained models or data splits has been left outside the if __name__ == "__main__": block. This will cause NameError exceptions at runtime because variables defined within the if block are not accessible globally. All top-level execution code needs to be moved inside the main guard to resolve this.

[aviralgarg05]

Pls review @sachinprasadhs