[model-gateway] add llama4 vision image processor

[slin1237]

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 @slin1237, 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 integrates the LLaMA 4 Vision model into the system by adding a dedicated image preprocessor. This new component handles the complex image processing requirements of LLaMA 4, such as dynamic tiling and normalization, which are crucial for accurate model inference. The changes also include the necessary infrastructure for generating and validating golden test data, ensuring the robustness and correctness of the implementation.

Highlights

• LLaMA 4 Vision Processor: Introduced a new image processor specifically for the LLaMA 4 Vision model, enabling its integration into the system.
• Tile-Based Processing: The new processor implements advanced tile-based image processing, including dynamic resolution selection, resizing, padding, and normalization, with a default tile size of 336x336.
• Golden Test Integration: Comprehensive golden tests have been added to validate the LLaMA 4 Vision processor's output against HuggingFace's reference implementation, ensuring accuracy across various image dimensions and aspect ratios.

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 introduces a new image processor for LLaMA 4 Vision models. The implementation in Rust is well-structured, thoroughly documented, and includes comprehensive unit and golden tests to ensure correctness. The changes are clean and follow the existing patterns in the codebase.

I've identified a couple of areas for improvement:

• A bug in the preprocess method where some configuration options (image_std, size) are incorrectly ignored.
• The golden tests for LLaMA 4 Vision have weak assertions for the output tensor shape, which can be made more robust.

My review includes specific suggestions to address these points. Overall, this is a solid contribution.