[Fix] unify benchmark tool by migrating to tilelang.profiler.do_bench

[RMLYC]

• Replace baseline_profile tool with tilelang.profiler.do_bench
• Align benchmarking interface with the existing profile module
• Ensure consistency across all performance measurement tools

[gemini-code-assist]

Summary of Changes

Hello @RMLYC, 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 streamlines the benchmarking infrastructure by migrating from a custom baseline_profile implementation to the standardized tilelang.profiler.do_bench utility. This change centralizes performance measurement logic, enhances consistency with existing profiling modules, and simplifies the code by removing manual timing and warmup procedures, ultimately leading to more reliable and maintainable benchmarks.

Highlights

• Benchmarking Tool Migration: Replaced the custom baseline_profile tool with tilelang.profiler.do_bench for standardized performance measurement.
• Benchmarking Consistency: Aligned the benchmarking interface with the existing profile module to ensure consistency across all performance measurement tools.
• Simplified Timing Logic: Removed manual CUDA event-based timing and warmup loops, leveraging the do_bench utility for more robust and accurate profiling.

🧠 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 successfully unifies the benchmarking tool by migrating to tilelang.profiler.do_bench and aligns the benchmarking interface, which is a positive step towards improving consistency and maintainability. The removal of manual CUDA event timing in benchmarks/benchmark.py simplifies the code and leverages the dedicated profiling utility. The refactoring of imports in benchmarks/flash_attn/mha.py is also a good practice. However, a potential issue exists with the removal of tuple handling in baseline_program within mha.py, which might lead to incorrect latency measurements if the underlying flash_attn_interface.flash_attn_func returns a tuple.

[xysmlx]

These changes looks good to me, however there may be other issues in the code format:

1. class naming format
2. type annotation for each function

[RMLYC]

These changes looks good to me, however there may be other issues in the code format:

1. class naming format
2. type annotation for each function

Done

[RMLYC]

all comment resolved and ci passed @xysmlx

[xiayuqing0622]

LGTM