[Kimi-K2.5] Fix missing quant_config in KimiK25

[mmangkad]

Motivation

Fix missing quant_config in KimiK25 causing nvidia/Kimi-K2.5-NVFP4 to use bf16 KV cache instead of fp8.

Modifications

Accuracy Tests

Benchmarking and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.

Review Process

1. Ping Merge Oncalls to start the PR flow. See the PR Merge Process.
2. Get approvals from CODEOWNERS and other reviewers.
3. Trigger CI tests with comments or contact authorized users to do so.
   • /tag-run-ci-label, /rerun-failed-ci, /tag-and-rerun-ci
4. After green CI and required approvals, ask Merge Oncalls to merge.

[gemini-code-assist]

Summary of Changes

Hello @mmangkad, 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 addresses a critical configuration oversight within the KimiK25 model's initialization process. By ensuring the quant_config is properly stored, it rectifies an unintended fallback to a less efficient KV cache quantization, thereby enabling the correct fp8 quantization for specific models like nvidia/Kimi-K2.5-NVFP4.

Highlights

• KimiK25 Initialization Fix: The KimiK25 class constructor now correctly assigns quant_config to self.quant_config. This resolves an issue where the nvidia/Kimi-K2.5-NVFP4 model was inadvertently using bf16 KV cache instead of the intended fp8.

🧠 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.

Changelog

• python/sglang/srt/models/kimi_k25.py
  • Added self.quant_config = quant_config to the KimiK25 class's __init__ method to ensure proper configuration handling.

Activity

• The pull request was opened by mmangkad to fix a missing quant_config assignment in the KimiK25 model, which was causing incorrect KV cache quantization.

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 addresses a bug where the quant_config was not being stored in the KimiK25ForConditionalGeneration model, leading to incorrect KV cache types for quantized models like nvidia/Kimi-K2.5-NVFP4. The fix correctly assigns the quant_config parameter to self.quant_config in the model's __init__ method. This change is correct and ensures that the quantization configuration is available for other components of the system, resolving the issue. The implementation is straightforward and I have no further comments.

[Qiaolin-Yu]

/tag-and-rerun-ci