Use the JTE API to set stacked table stats...

[cantonios]

… to the maximum of the input table specs.

This allows setting parameters like max_ids_per_partition and max_unique_ids_per_partition, suggested_coo_buffer_size for stacked tables with auto-stacking.

Although the heuristic may not be optimal, this at least provides a method for directly setting the values in the stacked tables, and is consistent with the default values if nothing is set.

Uses the jax_tpu_embedding API for future-proofing.

[gemini-code-assist]

Summary of Changes

Hello @cantonios, 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 refines the handling of stacked table statistics within the JAX TPU embedding layer. It introduces a mechanism to derive key parameters, such as maximum IDs per partition and suggested buffer sizes, by taking the maximum values from the constituent input table specifications. This ensures that auto-stacked tables are configured with appropriate and consistent values, utilizing the JTE API for the update process.

Highlights

• Stacked Table Statistics Update: The pull request updates how statistics for stacked tables are determined, specifically max_ids_per_partition, max_unique_ids_per_partition, and suggested_coo_buffer_size_per_device.
• Maximum Value Aggregation: These statistics are now set to the maximum values observed across all individual input table specifications within a stack, ensuring consistent parameterization for auto-stacking.
• JTE API Integration: The changes leverage the jax_tpu_embedding (JTE) API, using embedding.SparseDenseMatmulInputStats and embedding.update_preprocessing_parameters for a more robust and future-proof approach.
• Code Refactoring: The previous dataclasses.replace method for updating StackTableSpec has been replaced with explicit calculation and aggregation of statistics, leading to the removal of the dataclasses import.

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 refactors the logic for setting stacked table statistics to use the jax_tpu_embedding API, which is a good improvement for future-proofing and alignment with the library's intended usage. The change correctly computes the maximum values for max_ids_per_partition, max_unique_ids_per_partition, and the newly supported suggested_coo_buffer_size from the input table specs. The implementation is solid, with one minor suggestion to improve conciseness.