[Feature Request] Add GGUF Model Support to WanVideo TextEncode Cached Node

[makisekurisu-jp]

The custom node WanVideo TextEncode Cached does not currently support loading GGUF models. I’d like to request compatibility with GGUF format in future updates.

https://huggingface.co/city96/umt5-xxl-encoder-gguf

[Pythonpa]

yelp, hope this can be achieved

[makisekurisu-jp]

@kijai

https://huggingface.co/Kijai/WanVideo_comfy/blob/main/umt5-xxl-enc-bf16.safetensors

Could you add support for GGUF quantization for the umt5-xxl model? The model you provided is 10.5GB in size, and regardless of whether I use the preset workflows such as InfiniteTalkh or FlashVSR in ComfyUI-WanVideoWrapper, my 12GB 4070 Ti GPU runs into OOM errors.

[makisekurisu-jp]

#80