[Diffusion] Make QwenImageLayered resolution configurable

[xingsy97]

Motivation

QwenImageLayeredBeforeDenoisingStage hardcodes resolution = 640 instead of reading from the pipeline config.

# qwen_image_layered.py L425
resolution = 640  # TODO: support user-specified resolution

# qwen_image.py L503
resolution: int = 640  # TODO: allow user to set resolution

Modifications

Use server_args.pipeline_config.resolution in QwenImageLayeredBeforeDenoisingStage.forward() instead of hardcoded 640. Remove resolved TODO comments.

Default behavior unchanged (still 640). Users can now configure resolution via QwenImageLayeredPipelineConfig.

Accuracy Tests

Existing qwen_image_layered_i2i CI test covers this code path.

No model output changes, default resolution value remains 640.

Benchmarking and Profiling

N/A, no performance impact (config field read vs integer literal).

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.

[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 addresses a limitation where the QwenImageLayeredBeforeDenoisingStage component hardcoded the image resolution, preventing users from customizing it. The changes enable the resolution to be dynamically configured through the QwenImageLayeredPipelineConfig, providing greater flexibility for users while preserving the existing default behavior.

Highlights

• Configurable Resolution: Made the QwenImageLayered resolution configurable by reading from the pipeline configuration instead of using a hardcoded value.
• Code Cleanup: Removed outdated TODO comments that indicated the resolution should be user-specified, as this functionality is now implemented.
• Backward Compatibility: Ensured the default resolution remains 640, maintaining existing behavior while allowing for customization.

🧠 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/configs/pipeline_configs/qwen_image.py
  • Removed a TODO comment indicating that the resolution attribute should be user-settable, as this functionality is now implemented.
• python/sglang/multimodal_gen/runtime/pipelines_core/stages/model_specific_stages/qwen_image_layered.py
  • Replaced the hardcoded resolution = 640 with a dynamic value fetched from server_args.pipeline_config.resolution, allowing for configurable image resolution.

Activity

• No human activity has been recorded on 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 makes the resolution for QwenImageLayered configurable by replacing a hardcoded value with a parameter from the pipeline configuration. The associated TODO comments have also been removed. The changes are straightforward and achieve the intended goal, improving the flexibility of the pipeline while maintaining the existing default behavior.

[yhyang201]

/tag-and-rerun-ci

[mickqian]

user still can't specify (override) this from cli, am I understanding this correctly? if so, should we support it

[xingsy97]

user still can't specify (override) this from cli, am I understanding this correctly? if so, should we support it

@mickqian Yes. I just added the missing cli support with a UT.

[mickqian]

/tag-and-rerun-ci

[yhyang201]

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

— SGLDHelper bot

[mickqian]

well done