Need help using this GGUF model in ComfyUI
I’m new to ComfyUI and trying to use the Qwen-Image-Edit-Rapid-AIO-GGUF model. Could someone share a simple workflow that shows how to load the GGUF UNet, the required CLIP/text-encoder, and the VAE?
I installed the GGUF addon but not sure which nodes to use or where each file should go. A basic example workflow would really help. Thanks!
DOWNLOAD THIS ANIME IMAGE, and drag it directly to any place on comfyui canvas and you will get this model workflow
DOWNLOAD THIS ANIME IMAGE, and drag it directly to any place on comfyui canvas and you will get this model workflow
Thanks for the workflow! I noticed it requires adding an extra text-encoder and VAE, but the original AIO model (https://huggingface.co/Phr00t/Qwen-Image-Edit-Rapid-AIO) doesn’t need any external CLIP or VAE — everything is already packed inside the checkpoint.
So just to confirm:
Since this GGUF version is a quantized version of the AIO checkpoint, shouldn’t it also include the built-in text-encoder and VAE, instead of needing separate ones?
Just want to be sure if this GGUF behaves differently or if I’m missing something.
GGUF only has the unet. So, we need to load TE and VAE separately.
I was lazy to remove AIO in their names. Sorry for that.
I have not come across any gguf quant with inbuilt TE and VAE
Thanks for the clarification! Got it now — GGUF includes only the UNet, so TE and VAE must be loaded separately. Appreciate the explanation.