tried to make a style from old sd1.5 image generations

illustrious

• 0.7-0.85 for soft details

• 1 for actual style, but its trained data could overwrite the background. Also some sd1.5 shenanigans apply

• Recommended to accompany an artist tag with soft strength like:

  • 1girl, solo, eezistyle, (artisthere:0.5), etc...