Workflow request

#1
by aamobp - opened

can you please release a workflow for this model with the special ComfyUI-GGUF node and one else for the normal way?

Yes! There are already ComfyUI official templates for the original one, but I will also put example workflows up. (I fell asleep yesterday before I finished testing and uploading everything)

Hi @aamobp - I have put up some workflows in https://huggingface.co/molbal/MiniMax-H3-GGUF/tree/main/workflows - and I will add some more as I test and verify them locally

and I will add some more as I test and verify them locally

hi, thanks a lot for sharing!
it would be great to combine 3 in 1 gguf workflow with bypass possibility, to be able to switch between t/i/ref to video modes

I prefer the simpler workflows myself, so I do not have what you look, however there are many on civitai, for example:

You can use any of these, if you want to them to work with GGUF models, all you need to do is replace the 'Load Diffusion Model' node with Unet loader (Dynamic VRAM) or Unet loader (GGUF)

image

I prefer the simpler workflows myself, so I do not have what you look, however there are many on civitai, for example:

You can use any of these, if you want to them to work with GGUF models, all you need to do is replace the 'Load Diffusion Model' node with Unet loader (Dynamic VRAM) or Unet loader (GGUF)

image

hi , it does not work with Unet loader (GGUF) , also gguf loader is not working , where did u get Unet loader (Dynamic VRAM) ?

Hi @inykokso - please use https://github.com/molbal/ComfyUI-GGUF instead of https://github.com/city96/ComfyUI-GGUF

thanx , it works now , deleted whole gguf custom node , and replaced with yours

Sign up or log in to comment