Understands booru and e621 tags.
It's an EPS model but it seems to work fine without EPS sampling, but I'd still recommend you use the correct settings for whichever tool you use. (ComfyUI/A1111/etc.)
I do not use A1111, but for ComfyUI you have to use the node ModelSamplingDiscrete or an equivalent custom node and set it to EPS sampling.
In my testing the model gets the best results when using Euler CFG++ in ComfyUI. I'll attach the workflow I use when I am not lazy. You need SamplerCustomAdvanced and SamplerEulerCFG++, and set it to use the regular version.
For prompts you should first give subject tags, then describe the image you want to generate with words and add content + quality tags last. Use proper punctuation.
Positive Illustrious prompts:
(subject, {1boy/1girl/2boys/2girls/etc.}, artist/style). Describe image with natural language. {A portrait of ...}. (tags, quality tags).
Negative Illustrious prompts:
bad quality, worst quality, worst detail, sketch, censor.