Commercial use of this model is prohibited.

When you use this model to generate images or use the training set to generate models to generate images, you should indicate the source of the model and the the artist.

https://lit.link/en/QTONAGI

None of my model display pictures use controlnet, so you can try to reproduce the pictures. These pictures are randomly generated, I just picked the correct anatomy among them.

These 512*448 resolution pictures are just to show the difference between different base models and vae. (Blame my damn 1660s graphics card!) So don't judge my models by their performance.

Honestly, I do not really like pastel-waifu-diffusion VAE because it brings sharp edges.

The resolution of the training set is between 576 pixels and 1280 pixels. I recommend that you generate images at a resolution between 768px - 1024px. (Don't use 1024*1024, it will generate repeated structure.)

Recommend Base Model: CounterfeitV25_25, dalcefoPainting_3rd > Cetus-Mix

Recommend Vae：CounterfeitV25_25 > pastel-waifu-diffusion

Recommend Weight：0.7 - 1.0

2 Lora version. The newest version performers better but overfits a bit. It may not be able to meet your promt needs very well. However, I prefer the prettier one.

Pick according to your liking.

Enjoy it.