Neon Illustration Diffusion

Originally posted to HuggingFace by ProGamerGov

This fine-tuned Stable Diffusion v1.5 model was trained for 8000 iterations on a selection of artistic works by Pachu M. Torres. Training was performed using ShivamShrirao/diffusers with full precision, prior-preservation loss, the train-text-encoder feature, and the new 1.5 MSE VAE from Stability AI. A total of 4120 regularization / class images were used. Regularization images were generated using the prompt "artwork style", 50 DDIM steps, and a CFG of 7.

Note that v1 has a proclivity towards small pieces of text, but they are easy to clean up with in-painting.

If you like this art style, then please consider supporting the artist by purchasing one of their art pieces from their store.