[Misc][Tools][Benchmark] Add readme file for auto_tune script

[Chenyaaang]

No description provided.

[github-actions]

👋 Hi! Thank you for contributing to the vLLM project.

💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels.

Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run fastcheck CI which starts running only a small and essential subset of CI tests to quickly catch errors. You can run other CI tests on top of those by going to your fastcheck build on Buildkite UI (linked in the PR checks section) and unblock them. If you do not have permission to unblock, ping simon-mo or khluu to add you in our Buildkite org.

Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging.

To run CI, PR reviewers can either: Add ready label to the PR or enable auto-merge.

🚀

[gemini-code-assist]

Summary of Changes

Hello @Chenyaaang, 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 clarity and organization of the auto_tune benchmarking script. It introduces a dedicated README.md to serve as the primary source of documentation, while simultaneously refactoring the script's location and simplifying its internal comments for better maintainability.

Highlights

• Documentation Improvement: Added a comprehensive README.md file for the auto_tune script, providing detailed instructions on its purpose, prerequisites, configuration, usage examples, output interpretation, and how it internally functions.
• Code Organization: The auto_tune.sh script has been moved into a new dedicated directory (benchmarks/auto_tune/) to improve project structure and maintainability.
• Script Readability: Extensive inline commented instructions within auto_tune.sh have been removed and replaced with a concise reference to the newly added README.md, making the script itself cleaner and easier to read.

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 is currently in preview and 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 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. ↩

[Chenyaaang]

@bvrockwell

[gemini-code-assist]

Code Review

The code changes introduce a README.md file for the auto_tune script, enhancing its usability. The script's internal comments have been cleaned up, and it has been moved to a dedicated directory along with its new documentation. The review focuses on improving the clarity and correctness of the new README.md file.

[bvrockwell]

This is awesome @Chenyaaang, thank you for adding this!

Do you think it would be a good idea to add "most model length" env var referenced here? Although I don't think this is the only place we should add it, I certainly think it could be useful to have a short snippet explaining what it is here and when to use it.

[Chenyaaang]

This is awesome @Chenyaaang, thank you for adding this!

Do you think it would be a good idea to add "most model length" env var referenced here? Although I don't think this is the only place we should add it, I certainly think it could be useful to have a short snippet explaining what it is here and when to use it.

Do you mean, for example user's max-model-len 32k, most-model-len 2k, then we start the server with most-model-len config, and when we send the benchmark, we only use 2k? because the benchmark script only generates input with 1 length.

[bvrockwell]

This is awesome @Chenyaaang, thank you for adding this!
Do you think it would be a good idea to add "most model length" env var referenced here? Although I don't think this is the only place we should add it, I certainly think it could be useful to have a short snippet explaining what it is here and when to use it.

Do you mean, for example user's max-model-len 32k, most-model-len 2k, then we start the server with most-model-len config, and when we send the benchmark, we only use 2k? because the benchmark script only generates input with 1 length.

yes exactly! We actually don't have it documented anywhere at the moment, so it may be nice to start referencing it in a few places in the docs (it can lead to a big improvement in performance, but no one knows about it!).

[bvrockwell]

@yaochengji is this ready to merge? thanks!

[yaochengji]

Thanks for adding the doc! Left a nit.

[Chenyaaang]

This is awesome @Chenyaaang, thank you for adding this!
Do you think it would be a good idea to add "most model length" env var referenced here? Although I don't think this is the only place we should add it, I certainly think it could be useful to have a short snippet explaining what it is here and when to use it.

Do you mean, for example user's max-model-len 32k, most-model-len 2k, then we start the server with most-model-len config, and when we send the benchmark, we only use 2k? because the benchmark script only generates input with 1 length.

yes exactly! We actually don't have it documented anywhere at the moment, so it may be nice to start referencing it in a few places in the docs (it can lead to a big improvement in performance, but no one knows about it!).

So in this case to get the auto tune param, customers can just treat the input/output len for most-model-len as the INPUT_LEN and OUTPUT_LEN. What do you think I should add here?

[yaochengji]

LGTM!