Add errors when dtype is anything other than int32 for ptr metatdata

[pavanimajety]

This PR adds a very simple error check for the metadata arguments qo_indptr, kv_indptr and kv_lens array.

📌 Description

This PR adds a very simple error check for the metadata arguments qo_indptr, kv_indptr and kv_lens array.

🔍 Related Issues

🚀 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).
• [N/A] 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

• [N/A] Tests have been added or updated as needed.
• All tests are passing (unittest, etc.).

Reviewer Notes

[gemini-code-assist]

Summary of Changes

Hello @pavanimajety, 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 introduces robust data type validation for key metadata arguments (qo_indptr, kv_indptr, and kv_len_arr) within the flashinfer/mla.py module. By enforcing that these arguments are torch.int32 tensors, the changes aim to prevent potential errors and ensure compatibility with underlying operations that rely on specific data type conventions. Additionally, the relevant docstrings have been updated to reflect these int32 requirements, enhancing code clarity for developers.

Highlights

• Data Type Validation: Added explicit ValueError checks in the plan function within flashinfer/mla.py to ensure that kv_len_arr, kv_indptr, and qo_indptr tensors are of torch.int32 dtype, preventing potential runtime issues from incorrect data types.
• Documentation Update: Updated the docstrings for kv_indptr, kv_indices, and kv_len_arr in flashinfer/mla.py to explicitly state torch.Tensor(int32) as their expected data type, improving code clarity and documentation.

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. 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 adds important validation checks for the data types of metadata arguments in the plan method, ensuring they are torch.int32. This is a good improvement for robustness.

My review includes a suggestion to refactor the validation logic to be more concise and to add a missing check for the kv_indices tensor, which was updated in the docstring but not validated in the code.

[yzh119]

Thanks @pavanimajety for bringing this up!

torch.Tensor(int32)

is invalid typing and I encourage using:

torch.IntTensor

instead (IntTensor in torch refers to int32 tensor, specifically).

[pavanimajety]

Addressed the feedback!

[pavanimajety]

On a side note, does it make sense to make the ptr arrays int64? And leave the kv_lens and kv_indices in int32?

[yzh119]

does it make sense to make the ptr arrays int64? And leave the kv_lens and kv_indices in int32?

Sure at kernel-wise it could be supported, but let's leave it for later PRs :)

[yzh119]

LGTM