feat(tree_attn): implement Triton kernel

[alumkal]

Description

This PR implements a Triton kernel that supports tree attention.

Type of Change

• Bug fix (non-breaking change that fixes an issue)
• New feature (non-breaking change that adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not
  work as expected)
• Documentation update
• Code refactoring (no functional changes)
• Performance improvement
• Test coverage improvement

Checklist

• I have read the Contributing Guide
• I have run formatting tools (pre-commit or manual)
• I have run relevant unit tests and they pass
• I have added tests for new functionality
• I have updated documentation if needed
• My branch is up to date with main
• This PR introduces breaking changes (if yes, fill out details below)
• If this PR changes documentation, I have built and previewed it locally with
  jb build docs
• No critical issues raised by AI reviewers (/gemini review)

[gemini-code-assist]

Summary of Changes

Hello @alumkal, 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 tree attention mechanism by introducing a high-performance Triton kernel. This new kernel is designed to accelerate computations by optimizing GPU resource utilization. The changes ensure a seamless integration into the existing attention infrastructure, allowing users to opt into the Triton backend for potential performance gains while maintaining compatibility with the original implementation. The update also includes necessary data preparation steps and robust testing to validate the new functionality.

Highlights

• Triton Kernel Implementation: A new Triton-based kernel has been implemented for tree attention, designed to improve performance by leveraging GPU-specific optimizations.
• Conditional Backend Selection: The system now supports conditional switching between the existing FlexAttention and the new Triton kernel for tree attention, controlled by the AREAL_USE_TRITON_TREE_ATTN environment variable.
• Data Precomputation for Triton: New logic has been added to precompute specific data structures, such as packed masks and KV/Q block indices, from the tree structure (TrieNode) into a TreeAttentionData object, which is required by the Triton kernel.
• Integration into Attention Modules: The Triton kernel has been integrated into both the FSDP and Megatron attention modules, allowing them to utilize the new high-performance backend when enabled.
• Comprehensive Testing: Existing unit tests for tree training have been updated and parameterized to ensure correctness and gradient consistency across both the FlexAttention and the new Triton attention backends.

🧠 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 introduces a Triton kernel for tree attention to improve performance, which is a great addition. The implementation is well-structured with conditional imports to handle Triton's availability and includes tests for both the flex and Triton backends. My review focuses on improving code maintainability by reducing duplication in the attention modules and enhancing the performance of the CPU-based pre-computation steps for the Triton kernel.

[nuzant]

Please format the code following CONTRIBUTING.md, and all tests should pass before merging.