[bugfix] Add 'disaggregation_mode' parameter to warmup function when compile deep_gemm manually #8618
Conversation
…eep_gemm manually
There was a problem hiding this comment.
Summary of Changes
Hello @lbh2001, 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 implements a crucial bug fix to enhance the robustness of the DeepGEMM compilation warmup process. It introduces a new parameter to the warmup function, enabling it to dynamically adjust request parameters based on the disaggregation mode, thereby ensuring proper functionality in distributed and disaggregated environments.
Highlights
- Bug Fix for DeepGEMM Compilation: This pull request addresses a bug (issue #8617) where the DeepGEMM compilation warmup process did not correctly handle disaggregation modes, leading to potential issues in distributed environments.
- Parameter Addition to Warmup Function: The
warm_up_compileasynchronous function inpython/sglang/compile_deep_gemm.pyhas been updated to accept a newdisaggregation_mode: strparameter. - Conditional Bootstrap Configuration: Within
warm_up_compile, if thedisaggregation_modeis not 'null', thegenerate_req_inputobject'sbootstrap_roomis set to0andbootstrap_hostis set toFAKE_BOOTSTRAP_HOST. This ensures the warmup request behaves correctly in disaggregated setups.
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
-
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. ↩
There was a problem hiding this comment.
Code Review
This pull request adds the disaggregation_mode parameter to the warm_up_compile function to correctly configure warmup requests for DeepGEMM compilation, especially when disaggregation is enabled. The changes are logical and address the intended bugfix. I've added a couple of suggestions to improve maintainability by using an enum instead of a magic string for the disaggregation mode check. Otherwise, the changes look good.
|
@merrymercy @Ying1123 @zhyncs @hnyls2002 hello, can you help review this pr? |
…compile deep_gemm manually (#8618)
…compile deep_gemm manually (#8618)
…compile deep_gemm manually (#8618)
…compile deep_gemm manually (sgl-project#8618)
…compile deep_gemm manually (sgl-project#8618)
Motivation
fix #8617
Modifications
Accuracy Test
Benchmark & Profiling
Checklist