Text-to-Video
GGUF
English
hyvid
lora
gguf-comfy

How to connect HunyuanVideo Enhance A Video to gguf workflow?

#1
by Snepper - opened

Hi.
How to connect HunyuanVideo Enhance A Video to gguf workflow?
https://github.com/kijai/ComfyUI-HunyuanVideoWrapper
improves the generation picture

seems you should clone his repo to the custom_nodes folder; just like the gguf node settings; but his wrapper loader might not work with gguf file, since his nodes.py doesn't call any gguf tensor reader; you could try to connect the gguf unet loader to his sampler's model point; test it and see does it work or not

image.png
this is the connection that is posted in their examples Well, how to connect GGuf to it??

you see the top-left corner, the model with green dot; try to connect this green dot with the gguf unet loader; you should have these two custom nodes (both gguf node and the wrapper node which you wanna integrate to) in your pack first; check it out and test

If it were that simple I wouldn't be asking this question, this sampler doesn't accept gguf models, I tried to edit the sampler node code, but I couldn't make it work with the gguf loader

seems this particular sampler requires the hyvid_embeds and fed_args; those info might need contain in the gguf quantized file

Sign up or log in to comment