Hugging Face's logo

unsloth
/
Qwen-Image-Layered-GGUF

Image-Text-to-Image
GGUF
English
Chinese
ggml
quantized
unsloth
qwen
Model card Files
xet
 Community
3

Diffusers run gguf?

#2
by gabbo1995 - opened
Discussion
gabbo1995

Is there a way to run it with diffusers in python? Could you please provide an adapted example of how to load the DiT in gguf in diffusers? Thank you so much for your contribution!

danielhanchen
 Unsloth AI org

We're working on a guide, for now you'll need to use something like ComfyUI
gabbo1995

Thank you!!

marrrcin

@danielhanchen how is it going with this?
johndpope

https://github.com/search?q=+Qwen-Image-Layered-GGUF+&type=code

https://github.com/dseditor/ComfyuiWorkflows/blob/1fe44468779a3bfbc592cd5109bc760c41daec0b/Qwen/ImageEditQwen.json#L1007 < haven't tried it
ZKong
•
edited Jan 14

Diffusers need to compile gguf in every running , very bad. Can you make it work like ComfyUI-GGUF without compiling. If diffusers can have gguf with cpp speed, it will be great. For now ,diffusers can not match comfyui with gguf and adaptive vram . Comfyui can run LTX2 with just 1G vram and less than 32G ram.
danmadro
•
edited Mar 11

Is there a way now to run this gguf model with diffusers or llama.cpp?

Sign up or log in to comment