This is a Locon. You probably need the extension in A1111.
Config file is in the "Text Encoder" download.
The base model I used is Illuminutty Diffusion
You need to download the negative embeddings to replicate my generations. https://files.catbox.moe/0gjduq.zip
You may want to look at the tags in the dataset I provided to figure out what prompts to use.
Trained on images of stained glass from League of Legends' Shurima Cinematic
Dataset is poorly tagged, if someone can help me manually tag the dataset that would be awesome. It would help me train a better model.