feat(transformers): supplement glm4v processor

[wcrzlh]

What does this PR do?

This pr is proposed to add glm4v processor to avoid torchvision usage during transformers >= 4.53.0.

Fixes # (issue)
✅ LogitsProcessorList/StoppingCriteriaList is imported from mindone.transformers rather than transformers to avoid problem of tsf version incompatibility
✅ supplement chat_template.jinjacall in processing_utils

Adds # (feature)
✅ Glm4vImageProcessor/Glm4vProcessor
✅ Update ms2.7.0 performance

Example Usage & Performance
Please refer to usage & performance from Glm4v docs.

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline?
• Did you make sure to update the documentation with your changes? E.g. record bug fixes or new features in What's New. Here are the
  documentation guidelines
• Did you build and run the code without any errors?
• Did you report the running environment (NPU type/MS version) and performance in the doc? (better record it for data loading, model inference, or training tasks)
• Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

@xxx

[gemini-code-assist]

Summary of Changes

Hello @wcrzlh, 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 expands the multimodal processing capabilities for the GLM-4V model within the mindone.transformers library. It introduces dedicated image and general processors, ensuring seamless integration with the existing auto-loading infrastructure. Additionally, it addresses a specific compatibility issue with Flash Attention 2 and refines the system for loading chat templates, making the framework more robust and adaptable.

Highlights

• GLM-4V Processor and Image Processor: Introduced new Glm4vProcessor and Glm4vImageProcessor classes to handle multimodal inputs for the GLM-4V model, including dynamic image resizing and patching logic.
• Auto-mapping Integration: Integrated the new GLM-4V processors into the auto-mapping system, allowing them to be automatically loaded by AutoProcessor.
• Flash Attention Fix: Implemented a temporary workaround for Flash Attention 2 in GLM-4V modeling to ensure compatibility with attention mask handling in MindSpore by switching to eager attention when _attn_implementation is 'flash_attention_2'.
• Chat Template Loading Improvement: Enhanced the chat template loading mechanism to prioritize .jinja files, providing more flexibility for template definitions.

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 the Glm4vProcessor and Glm4vImageProcessor to support the GLM-4V model. The changes include adding these new processor classes, registering them with the Auto classes, and updating an example script to use them. Additionally, there are some related refactorings and fixes in other parts of the codebase, such as flash attention integration and processing utilities.

My review has identified a critical issue in the new Glm4vImageProcessor that would cause a runtime error, and a high-severity issue in Glm4vTextModel related to state management that could lead to incorrect behavior. Please see the detailed comments for suggestions on how to fix these issues.