This is just quantized (Q8_0, Q5_0) versions of the model Fluxmania III that excels in artistic photography. All credit goes to the original authors, I'll be happy to delete this post once (if) official GGUF version(s) are published.
There is (obviously) some quality loss, so if your hardware allows it, use fp8/fp16 versions.
Q5_0: VAE and clip are included.
Q8_0: Base model only, use Flux VAE and clip.
Tested in SwarmUI (just set Architecture in the model metadata to Flux.1 DEV) and ComfyUI with 12GB VRAM. Let me know if I should add Q4_0 for 8GB cards.
Samples have been created using DPM++ 2M sampler and beta or normal scheduler.
Sorry in advance for any shortcomings, this is my first share here.
Description
Without VAE and Clip. Significantly closer to the FP16 quality.
FAQ
Comments (8)
Hey! Thanks for making .gguf versions of the model.
It would be great if you could also add a k4 version? I'm running on 8gb and feel like I'm missing out big time on this model.
Thanks!
I'm using RTX 2060 6GB VRAM with Forge, and Q8 can be run on it. Got around 1:45 minutes for generation.
please kindly add Q4_0 for 8 gb cards <3
I get beautiful results but sometimes I get banding, across part of the image or the whole image. How do I fix that. Thank you.
Could you do this for the latest fluxmania model?
Can you please add Q2 version?
Also, can you make latest Legacy version in Q2 version?
Thanks!
Please add Q4 for 8GB cards. Thanks!
The download link has become invalid.