[Diffusion] chore: ensure CFG Zero Star numerical stability for Helios model

[RuixiangMa]

Motivation

Add explicit .float() conversion in Helios's optimized_scale function, align with diffusers PR #13214 to ensure CFG Zero Star numerical stability in bf16

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]
• (https://docs.sglang.io/developer_guide/contribution_guide.html#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, 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 enhances the numerical stability of the CFG Zero Star calculation within the optimized_scale function. By ensuring that tensor operations are performed with full floating-point precision, it prevents potential issues that can arise when working with bf16 (bfloat16) data types, aligning the implementation with best practices for robust deep learning computations.

Highlights

• Numerical Stability: Explicitly converted positive_flat and negative_flat tensors to float() within the optimized_scale function to prevent numerical instability, particularly when using bf16 precision, aligning with a Diffusers PR.

🧠 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/multimodal_gen/runtime/pipelines_core/stages/model_specific_stages/helios_denoising.py
  • Added explicit .float() conversions for positive_flat and negative_flat tensors.

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 adds explicit .float() conversions to the optimized_scale function to improve numerical stability, particularly with bf16 precision. The change is correct and well-motivated. I've provided a suggestion to refine the implementation for better clarity and slightly improved memory efficiency by avoiding parameter shadowing and an unnecessary intermediate variable.

[RuixiangMa]

/tag-and-rerun-ci

[mickqian]

/tag-and-rerun-ci

[yhyang201]

/rerun-failed-ci

[yhyang201]

/rerun-failed-ci

[yhyang201]

@mickqian Nvidia CI passed and PR is approved, ready for merge

— SGLDHelper bot