support Sail-VL2 models

[hjh0119]

No description provided.

[gemini-code-assist]

Summary of Changes

Hello @hjh0119, 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 significantly enhances the system's capabilities by introducing full support for the SAIL-VL2 series of multimodal models. It involves registering these models and their associated templates, implementing specific generation logic to ensure compatibility, and updating user-facing documentation. A new template class has been developed to streamline the multimodal input encoding, ensuring these vision-language models function seamlessly within the existing framework.

Highlights

• New Model Support: Adds comprehensive support for the SAIL-VL2 family of multimodal models, including SAIL-VL2-2B, SAIL-VL2-2B-Thinking, SAIL-VL2-8B, and SAIL-VL2-8B-Thinking variants.
• Model Integration: Integrates SAIL-VL2 models by registering them in the system's model and template constants, ensuring they are discoverable and usable within the framework.
• Custom Generation Logic: Implements a custom generate method patch for SAIL-VL2 models to correctly handle inputs_embeds during the generation process, adapting to their specific requirements.
• Multimodal Template: Introduces a new SailVLTemplate class to manage image token processing and embedding for SAIL-VL2 models, facilitating proper multimodal input handling.
• Documentation Update: Updates both the Chinese and English documentation files to list the newly supported SAIL-VL2 models, providing clear information for users.
• Unit Tests: Adds a dedicated unit test (test_sailvl2) to verify the correct functionality and integration of SAIL-VL2 models within the system.

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

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 Sail-VL2 models, including model and template type definitions, model registration, a new template class, documentation updates, and corresponding tests. My review focuses on improving the documentation, correcting a type hint, fixing an incorrect assertion, adding a missing tag for the new model, and simplifying the test code for better clarity and correctness. Overall, the changes are well-structured, but a few refinements are needed.

[hjh0119]

/gemini review

[hjh0119]

#5898

[gemini-code-assist]

Code Review

This pull request adds support for the Sail-VL2 model family. The changes include adding model definitions, a new template implementation, and corresponding documentation and tests. The implementation reuses the internvl model architecture, which is a good practice. My review focuses on improving code correctness, maintainability, and clarity. I've identified a critical bug in an assertion, a potential device placement issue, and several opportunities to improve code clarity with better naming and comments, especially regarding a specific dependency version constraint.