Not tested much at all on realism checkpoints. May take multiple tries to get a good seed, but some excellent results once you do!








Only tested on Anime and non photorealistic checkpoints. Trained on an Anime checkpoint, not base 1.5, so needs booru tagging support.
While browsing, I noticed that there were a few Rita LoRAs, but none for her original presentation, Umbral Rose. So that became the target of my next embed! I haven't played the game in years, but Umbral Rose had always been a favorite.
This one's pretty straightforward and works pretty well, as long as you're using a booru-like checkpoint and follow their tagging conventions.
It's an embed, not a lora, so no triggers besides the filename, and she's been trained to come out with the maid outfit most of the time; if it's missing, simply add
maid outfit, maid headdressto the prompt. Easynegative and bad-image negative also recommended for best results.
Enjoy!
Not tested much at all on realism checkpoints. May take multiple tries to get a good seed, but some excellent results once you do!
1. สิทธิ์ของโมเดลที่โพสต์ซ้ำเป็นของผู้สร้างต้นฉบับ
2. ผู้สร้างต้นฉบับที่ต้องการรับรองโมเดล โปรดติดต่อเจ้าหน้าที่ SeaArt AI ผ่านช่องทางทางการ คลิกเพื่อรับรอง
