Support for WAN21 Model in TensorRT

[zhangkom]

Hi team,

I'm a user of ComfyUI with TensorRT integration and I'm excited about the performance gains it provides for inference. I was wondering if there's any timeline or plan to add support for the WAN21 model in future updates?

This model seems promising for [briefly mention why, e.g., "high-quality image generation with advanced features"], and TensorRT optimization would make it even more accessible on NVIDIA hardware. If it's already in the works or if there are any workarounds/prerequisites, I'd love to hear more.

Thanks for your hard work on this project!

Best,
zhangkom

[kanarch66]

Wan2.2 too