Feature: Tiled VAE Decode Support for Wan Video Workflow

[NeuralFault]

Implemented Tiled VAE Module: Integrated TiledVAEDecode into the Wan Text-to-Video and Image-to-Video workflows, replacing the standard VAE decoder when enabled.

Enabled Addons System: Activated the addon system for WanSamplerCardViewModel, allowing Tiled VAE (and future modules) to be selected from the UI. Keeping UI/UX consistency with the Image-gen UI's.

Improved Module Application: Updated BuildPrompt logic in Wan ViewModels to correctly execute module actions (like VAE replacement) during the build pipeline.

Serialization Fixes: Added missing [JsonPropertyName] attributes to TiledVAEDecode builder to ensure proper parameter passing to ComfyUI.

UI Progress Fix: Patched InferenceGenerationViewModelBase to correctly display "Executing [Node]" status for long-running nodes (like VAE Decode) that occur after sampling is complete, fixing a bug where the UI remained stuck on the sampler progress bar.

[gemini-code-assist]

Summary of Changes

Hello @NeuralFault, 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 significant enhancements to the video generation workflows by integrating a Tiled VAE decoding module. This change not only optimizes memory usage for high-resolution video outputs but also improves the user experience through a new addon system in the UI and more accurate progress reporting for long-running operations. The underlying prompt building and serialization mechanisms have been refined to support these new capabilities seamlessly.

Highlights

• Tiled VAE Decode Integration: The TiledVAEDecode module has been integrated into both Wan Text-to-Video and Image-to-Video workflows, replacing the standard VAE decoder when enabled. This allows for more efficient processing of large video generations.
• Addons System Activation: The addon system has been activated for WanSamplerCardViewModel, enabling the selection of Tiled VAE (and future modules) directly from the UI, ensuring consistency with existing Image-gen UI/UX.
• Improved Module Application Logic: The BuildPrompt logic in Wan ViewModels has been updated to correctly execute module actions, such as VAE replacement, during the prompt building pipeline, ensuring proper application of selected modules.
• Serialization Fixes for ComfyUI: Missing [JsonPropertyName] attributes have been added to the TiledVAEDecode builder, ensuring that parameters are correctly serialized and passed to ComfyUI.
• UI Progress Display Fix: A bug in InferenceGenerationViewModelBase has been patched to correctly display the 'Executing [Node]' status for long-running nodes (like VAE Decode) that occur after sampling is complete, preventing the UI from appearing stuck.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

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 Tiled VAE Decode support for video workflows, which is a great feature addition. The implementation includes enabling an addon system, fixing serialization issues with JsonPropertyName attributes, and patching a UI progress display bug. The code changes are logical and well-executed. My main feedback is around code duplication in the BuildPrompt methods of the InferenceWanTextToVideoViewModel and InferenceWanImageToVideoViewModel classes. Refactoring this would improve long-term maintainability. Overall, this is a solid contribution.

[mohnjiles]

Thank you!