[FlashInfer] Bump FlashInfer version from 0.6.3 to 0.6.4

[mmangkad]

Motivation

• CuteDSL FP4 MoE for DeepSeek-R1 Performance (#2398)
• TRTLLM-Gen MxFP8 MoE Integration (#2505)
• GDN Decode CuteDSL Kernel (#2498)
• TRTLLM-Gen Skip-Softmax Attention (#2477, #2547)
• Critical Bug Fixes
  • FP4 one-shot launch config instability fix (#2557)
  • W4A8 autotune crash fix (#2564)
  • CUDA Graph fix for Sampling (#2432)
  • SM121 (DGX Spark) DeepSeek prefill support (#2559)

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 updates the FlashInfer library version from 0.6.3 to 0.6.4 across the project. This update integrates the latest features and critical bug fixes from the FlashInfer project, ensuring the system leverages the most recent performance enhancements and stability improvements.

Highlights

• New Features: Incorporated CuteDSL FP4 MoE for DeepSeek-R1 Performance, TRTLLM-Gen MxFP8 MoE Integration, GDN Decode CuteDSL Kernel, and TRTLLM-Gen Skip-Softmax Attention.
• Critical Bug Fixes: Addressed issues including FP4 one-shot launch config instability, W4A8 autotune crash, CUDA Graph fix for Sampling, and SM121 (DGX Spark) DeepSeek prefill support.

🧠 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

• docker/Dockerfile
  • Updated the FLASHINFER_VERSION argument from 0.6.3 to 0.6.4.
• python/pyproject.toml
  • Updated the flashinfer_python and flashinfer_cubin dependencies from version 0.6.3 to 0.6.4.
• python/sglang/srt/entrypoints/engine.py
  • Updated the required flashinfer_python version in the assert_pkg_version check from 0.6.3 to 0.6.4.
• python/sglang/srt/server_args.py
  • Updated the minimum flashinfer-python version check for enabling flashinfer_trtllm from 0.6.3 to 0.6.4.
• python/sglang/srt/utils/common.py
  • Updated the example min_version in the docstring for check_pkg_version_at_least from '0.6.3' to '0.6.4'.
• scripts/ci/cuda/ci_install_dependency.sh
  • Updated the FLASHINFER_VERSION environment variable from 0.6.3 to 0.6.4.

Ignored Files

• Ignored by pattern: .github/workflows/** (1)
  • .github/workflows/release-docker-cu13-framework.yml

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 bumps the FlashInfer version from 0.6.3 to 0.6.4. The changes are applied consistently across the Dockerfile, Python dependencies (pyproject.toml), version assertions in the engine and server arguments, and the CI installation script. This update is necessary to incorporate the latest features and bug fixes from FlashInfer. The changes are correct and well-contained. I approve this version bump.

[mmangkad]

/tag-and-rerun-ci