gguf's general.architecture is 'qwen' but not 'qwen_image'
#3
by
Jianqiao1
- opened
Hi, Comfyui-GGUF node need qwen-image model's general.architecture is 'qwen_image', but now these gguf general.architecture is 'qwen'. Would you like to fix them? Thanks.
Even if I transformed the architecture, I still couldn't achieve the acceleration of twinflow
Even if I transformed the architecture, I still couldn't achieve the acceleration of twinflow
Currently, your model cannot be loaded at all using the gguf node, which makes it completely useless.
However, after correcting the architecture parameters, it works perfectly.
Try this set of nodes: