Accurate LORA for Ryōna from Senran Kagura, may need to use a lower weight to use different styles.<s>I suggest adding "blonde hair, heterochromia" to the prompt to generate her more accurately.</s> Fixed in V2. You do not have to tag her features, only outfit trigger wordsoutfits:<ul><li>shinobioutfit, <s>wings dress</s> Built into v2</li></ul><ul><li>heartbikini</li><li>student uniform</li></ul>

Ryōna (Senran Kagura) [Illustrious/Anima/Pony]

Same dataset, a little bit trickier to prompt

Files with SHA256: 203b45bc05654f88532ea12fcb1941f5be1a697eff0f03ea0eeebed7b8a9bc86