diffusion: support zimage

[yhyang201]

Motivation

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 @yhyang201, 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 introduces initial support for the ZImage diffusion model into the SGLang multimodal generation system. It establishes the necessary configuration classes, model implementations, and pipeline stages to enable ZImage's functionality. The changes reflect an ongoing development effort, with debugging statements and placeholders indicating that the feature is still under active construction and marked as '[do not merge]'.

Highlights

• ZImage Model Integration: Comprehensive support for the ZImage diffusion model has been added, including its specific architecture, VAE, and text encoder configurations. This involves new configuration files and a dedicated pipeline.
• Generic Hugging Face Text Encoder: A new generic TransformersModel and HFModelArchConfig were introduced to load text encoders from Hugging Face, enhancing the flexibility of text encoding within the multimodal generation framework.
• Guidance Scale Validation Update: The validation logic for guidance_scale across several pipeline stages has been updated to allow non-negative values (including zero), providing more flexibility for model inference.
• Extensive Debugging Additions: Numerous print statements have been added throughout the codebase, particularly in pipeline stages and model loading, indicating active development and debugging efforts for the new ZImage integration.

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 support for the zimage diffusion model. The changes are extensive, including new model implementations, configurations, and pipeline adjustments. The work appears to be in progress, as indicated by the [do not merge] in the title and the presence of numerous debugging statements (print, torch.save). My review focuses on removing this debug code, fixing clear bugs like an incorrect import, and addressing hardcoded values that should be configurable. There are also some potential issues with unused parameters and overly restrictive assertions that should be addressed before merging.

[mickqian]

/tag-and-rerun-ci