A collection of GGUF for Illustrious models just to help low memory people and do some experiences:
Animagine - https://civarchive.com/models/260267
Bismuth Illustrious Mix - https://civarchive.com/models/1359028
Bubble hentai - https://civarchive.com/models/1164420
CherryMIX_XL - https://civarchive.com/models/1079112
CyberRealistic - https://civarchive.com/models/1125067
GoldenHour - https://civarchive.com/models/2433139
GSBeyondReality - https://civarchive.com/models/2455942
Hassaku - https://civarchive.com/models/140272
Holly Mix - https://civarchive.com/models/959490
iLustMix - https://civarchive.com/models/1110783
Nova Anime - https://civarchive.com/models/376130
Nova Furry - https://civarchive.com/models/503815
Nova Orange - https://civarchive.com/models/967405
Nova Reality DMD2 - https://civarchive.com/models/453428?modelVersionId=2439777
One Obsession - https://civarchive.com/models/1318945
One obsession Branch - https://civarchive.com/models/1368727
PerfectDeliberate - https://civarchive.com/models/24350
Poly-Goth Gloss - https://civarchive.com/models/2506452
Prefect illustrious - https://civarchive.com/models/1224788
RagIllustriousMix - https://civarchive.com/models/1491577
RealPhotoIL - https://civarchive.com/models/2041366
Ri-Mix Omega - https://civarchive.com/models/996495
SmoothMix - https://civarchive.com/models/1695253
SmoothMix Ultimate - https://civarchive.com/models/2311905
TovaMix - https://civarchive.com/models/2284900
Unholy Desire Mix - https://civarchive.com/models/1307857
WAI-illustrious-SDXL - https://civarchive.com/models/827184
The rest of the GGUFs versions are in my huggingface - https://huggingface.co/BigDannyPt/Illustrious-GGUF-Collection, not going to upload since it takes a lot for it in CivitAI.
The hugginface link also includes the clips and vae for each model.
What I noticed is that Q6 or lower, will increase the s/it to double, not sure why...
If you have a specific model that you want the conversion, you can ask that I'll try to do it
Description
FAQ
Comments (19)
hello, i wanna ask, in comfy UI just add gguf loader? or need something to work?
You will need to add the GGUF loader, but then the ClipLoader and the VAE loader
@BigDannyPt what clip model you use?
@saytolol12 I have the clips from the normal SDXL
Which clip and vae you use?
The normal ones from SDXL.
But you can always extract them from a model and also do some tests
sorry for my stupid question, why the format is .gguf and not .safetensors?
can ConfyUI run .gguf?
because is a gguf conversion.
yes, with ComfyUI-GGUF node
@BigDannyPt i got problem, when i start queue. i got this error 'NoneType' object has no attribute 'device'
here is my workflow
Unet Loaer (GGUF)
KSampler
Empty Laten Image
Vae Decode
Load Vae
Safe Image
Dual Clip Loader (GGUF)
Clip Text Encode (Prompt) Positive and Negative
@Anying but the error is coming from which node?
also, you are also using gguf for clip?
these models are only for the Unet Loader GGUF, you then need to get the CLIP and the VAE and load them accordingly.
This is an example on what you should have - https://imgur.com/a/NnImn8F
The CLIPS can be something like this - https://civitai.com/models/1044804 just take noticed that you need CLIP-L and CLIP-G
The VAE can be this one https://civitai.com/models/296576/sdxl-vae
@BigDannyPt IT WORKS!!! it not showing any error again. generate 3 minute for 1024 dimension, but the result is blank black. not showing any image, maybe the VAE? im using another Vae for testing. gonna try vae from your link.
anyway, thanks for sharing...
Update: well, the result still Blank Black. I have download everything on your Link. Dunno what i miss
@Anying i think you are using the clip L twice, i got the same issue when tried to use those two CLIP.
here are the links for the ones that i've tried and worked correctly, they are the simulacrum ones from the same page
https://civitai.com/models/1044804?modelVersionId=1369648
https://civitai.com/models/1044804?modelVersionId=1208216
if you want to extract the VAE and CLIPs yourself, you can use the following workflow for it, it is fun to test with different CLIPs since it change the result
off course you will need the full model to extract the CLIP and VAE
@BigDannyPt im downloading 2 Clip simulacrum G and L. its 2GB for G, and 400mb for L. i tryed to switch it in case im in wrong order.
clip1 for G
clip2 for L
result = abstrack
i switch order
clip1 for L
clip2 for G
result = still abstrack
my PC is super potatoes, dunno if it can do extrackting Model
gtx 1650 4GB vram
with 8GB ram only.
i dont know much about ai, so i just using ComfyUI with simple node for SD1.
but i see this post, so it make me so exited to using Illustrious wich is better than SD1
@Anying can you share a screenshot of the workflow where it also shows the full name of the CLIP, VAE and model?
Error:
KSampler 'NoneType' object has no attribute 'shape'
--------
I assume you are speaking about this error.
¿¿¿¿ Anyone get to work this with KRITA AI-Diffusion ?????
@frikitin sorry, i delete all of them already. i just get back to SD1.5 TT__TT
¿¿¿¿ Anyone get to work this with KRITA AI-Diffusion ?????
The new versions "ALL" with included VAE worked fine in Krita AI ComfyUI. Thanks!!
All of them have a minor quality loss.
Using upscalers makes nothing (remacri, ESRGAN, etc).
The best way is to use DMD2 lora (get more photorealistic, losing artistic style).
The times gains about 15seg faster GUFF Vs Normal checkpoint.