Fix HybridAttnBackend forward for linear attention

[akhilg-nv]

Fixes #18983

Motivation

HybridAttnBackend lacked a forward() method, causing TypeError when RadixLinearAttention called attn_backend.forward() with mixed_qkv, a, b arguments. This occurs for hybrid attention models (e.g., Qwen3-Next) when we select different attention backends for prefill/decode.

Modifications

Added forward() method that accepts both regular attention (q, k, v) and linear attention (mixed_qkv, a, b) interfaces and routes to the appropriate HybridLinearAttnBackend.

Accuracy Tests

Validated that it runs, and briefly tested GPQA on a much smaller sample for sanity.

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 @akhilg-nv, 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 bug in the HybridAttnBackend that caused a TypeError when attempting to use linear attention. By introducing a comprehensive forward() method, the system can now seamlessly handle both standard and linear attention mechanisms, ensuring proper functionality for hybrid attention layers like those found in Qwen3-Next models.

Highlights

• HybridAttnBackend Functionality: Resolved a TypeError in HybridAttnBackend by implementing a forward() method, which was previously missing when RadixLinearAttention attempted to use it with linear attention arguments.
• Unified Attention Interface: The newly added forward() method now supports both traditional attention (q, k, v) and linear attention (mixed_qkv, a, b) interfaces, routing calls to the appropriate backend.

🧠 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/layers/attention/hybrid_attn_backend.py
  • Added a forward method to HybridAttnBackend to correctly dispatch calls for both regular and linear attention types.
  • Imported Tuple and Union from typing to support more flexible type hints for the new forward method's parameters.

Activity

• No specific activity (comments, reviews, progress) has been recorded for this pull request yet.

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 correctly addresses a TypeError by adding a forward method to HybridAttnBackend. This new method properly handles both standard and linear attention arguments by delegating the call to the appropriate backend based on the forward mode. The change is clean and directly solves the described issue. I've added a couple of comments regarding type hint consistency which could prevent potential runtime errors.