Remove RoPE fallback vllm custom ops

[jinyouzhi]

Motivation

Try to remove vllm custom ops dependency.
#13328 (comment) cc @merrymercy

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.
• Work with maintainers to merge your PR. See the PR Merge Process

[gemini-code-assist]

Summary of Changes

Hello @jinyouzhi, 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 simplifies the Rotary Positional Embedding (RoPE) implementation by completely removing the reliance on VLLM custom operations. By eliminating the conditional fallback to vllm._custom_ops.rotary_embedding, the codebase becomes more consistent and easier to maintain, ensuring all RoPE applications utilize a single, dedicated kernel.

Highlights

• Dependency Removal: Eliminated the dependency on vllm._custom_ops.rotary_embedding by removing the fallback mechanism for Rotary Positional Embedding (RoPE) operations.
• Code Simplification: Streamlined the RotaryEmbedding class by removing conditional logic for use_fallback_kernel and unifying the forward_cuda method to always use apply_rope_with_cos_sin_cache_inplace.

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 removes the dependency on vllm._custom_ops for the RoPE fallback implementation. The change simplifies the code by always using the sgl_kernel's apply_rope_with_cos_sin_cache_inplace function in forward_cuda. While this is a good simplification, the fallback was previously used for CUDA devices with head sizes not in [64, 128, 256, 512]. The existing tests do not seem to cover these cases. I've added a comment to suggest adding test cases for these head sizes to prevent potential regressions.