Adding tea cache wan2.2s2v

[aviveise]

[gemini-code-assist[bot]]

Summary of Changes

Hello @aviveise, I'm Gemini Code Assist[^1]! 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 refactors core components of the wan2.2s2v model within the diffsynth framework. The primary goal is to introduce a 'tea cache' for performance optimization and to improve code clarity through more descriptive variable naming. Additionally, the WanVideoPipeline has been enhanced to support an offline_preprocessing mode, providing greater flexibility for different data processing scenarios. These changes collectively aim to boost the model's efficiency and make the codebase easier to understand and extend.

Highlights

• Refactoring of Variable Names: Generic tensor variable names like x and context have been systematically renamed to more descriptive latents/hidden_states and encoder_hidden_states across multiple model and distributed processing files, enhancing code readability and maintainability.
• Tea Cache Integration: A new tea_cache mechanism has been introduced and integrated into the WanS2VModel and its associated pipeline functions. This cache aims to optimize computation by storing and reusing intermediate states, potentially improving performance.
• Offline Preprocessing Support: The WanVideoPipeline now supports an offline_preprocessing flag, which allows for dynamic configuration of pipeline units. This enables specialized workflows with new units like WanVideoUnit_InputVideoEmbedderPassThrough and WanVideoUnit_ImageEmbedderFusingOnly.
• Unified Sequence Parallelism Updates: The WanS2VModel has been updated to better integrate with unified sequence parallelism, including logic for chunking and gathering hidden states. A debug print statement was also added to indicate USP activation.

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 :thumbsup: and :thumbsdown: 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.

[^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.

[Artiprocher]

@aviveise Thanks for your contribution. Due to the substantial amount of code changes required, we will rewrite your code, which will be completed within 1–2 weeks.