独自のSD2.1 768モデルの出力などを学習し、他のモデルとマージしました。
Learned the outputs of our own SD2.1 768 model, etc., and merged them with other models.
We need the license to match kohaku-XL_alpha7.
1) Do I have to write "kblueleaf" and "Aikimi" somewhere every time I publish an image using this model?
A. No.
2) Do I have to write "kblueleaf" and "Aikimi" every time I publish a new model merging this model?
A. Yes
Kohaku-XL
https://civarchive.com/models/136389?modelVersionId=150441
Other merging models
animaginexl,sdvn7Realartxl_beta,blue_pencil-XL-v0.3.1
Description
FAQ
Comments (5)
这几个模型什么关系，哪个最好
subarashii 2.5D,Best I've ever seen.Like other sdxl,Prefer to draw portraits,lack of imagination.Animated models seem to have lower pixels compared to real models.
I love the style, but I'm getting artifacts (white spots) in the dark parts of the images. They can also be seen on many of the gallery images below. Has anyone figured out a way to prevent this?
Same. I tried to fix it with LoRA and Embedding but no luck.
I had the same problem along with the colors were washed out, fixed by setting the VAE to sdxl_vae.safetensors.