The model is a mix of ((NovelAI+TrinArt2+SD1.5 weighted sum at 0.20) + Anythingv3 + Lorain's Mix add difference 100). It seems to mainly be talented as tasks of singular anime girls, highly detailed backgrounds, and many outfits and locations though lacks in other things like stylization with sci-fi and non-anime fantasy.
Adding the tag "NSFW" to prompt will also magically make images generate nice images and handle the themes its weak in better.
For best color results I use the VAE for SD found here: https://huggingface.co/stabilityai/sd-vae-ft-mse-original
Description
SafeTensors conversion
FAQ
Comments (2)
Is there a CKPT version of this model? I am using DiffusionBee on Mac. Thanks.
where did you create this image?