misc: Customize kv lens buffer size for sparse attention

[Edenzzzz]

📌 Description

🔍 Related Issues

Fixes #1367

🚀 Pull Request Checklist

Thank you for contributing to FlashInfer! Before we review your pull request, please make sure the following items are complete.

✅ Pre-commit Checks

• I have installed pre-commit by running pip install pre-commit (or used your preferred method).
• I have installed the hooks with pre-commit install.
• I have run the hooks manually with pre-commit run --all-files and fixed any reported issues.

If you are unsure about how to set up pre-commit, see the pre-commit documentation.

🧪 Tests

• Tests have been added or updated as needed.
• All tests are passing (unittest, etc.).

Reviewer Notes

[gemini-code-assist]

Summary of Changes

Hello @Edenzzzz, 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 enhances the flexibility of the sparse attention mechanism by making the size of its internal key-value (KV) lens buffers configurable. This change allows for better memory management and performance tuning for different sparse attention use cases, addressing a previous limitation where these buffer sizes were fixed.

Highlights

• Configurable KV Lens Buffer Size: I've introduced a new kv_lens_buffer_size parameter to the BlockSparseAttentionWrapper's constructor in flashinfer/sparse.py. This parameter, defaulting to 32768, allows users to customize the size of the internal buffers used for sparse attention.
• Dynamic Buffer Allocation: The _vector_sparse_indptr_buffer and _kv_lens_buffer within the BlockSparseAttentionWrapper now dynamically allocate their size based on the new kv_lens_buffer_size parameter, replacing previously hardcoded values.

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 in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

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 issue 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 or fill out our survey to provide feedback.

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 the kv_lens_buffer_size parameter to the BlockSparseAttentionWrapper class, allowing for customization of the KV lens buffer size. The implementation is straightforward and correct.

[gemini-code-assist]

The description for kv_lens_buffer_size states the size is (num_kv_heads * MB), but for BlockSparseAttentionWrapper, the required size for _kv_lens_buffer is MB (number of block rows) and for _vector_sparse_indptr_buffer it's MB + 1. The (num_kv_heads * MB) seems to be from VariableBlockSparseAttentionWrapper and is misleading here.

Also, the parameter is used for both _kv_lens_buffer and _vector_sparse_indptr_buffer, so the docstring could be more precise by mentioning both.

Suggested change

	kv_lens_buffer_size : int
	The size of the kv lens buffer (num_kv_heads * MB), defaults to 32768.
	"""
	kv_lens_buffer_size : int
	The buffer size for KV lens and sparse indptr (>= MB). Defaults to 32768.

[KevinZeng08]

Can we allocate dynamically in the plan function without user specification?

[Edenzzzz]

It would be problematic if we want to support cuda graph later

[KevinZeng08]

Got it. Your PR works for me, thanks for your fix!