BRKN-KREA2-SOPHIA
This is a LoRA I created after members of my Telegram group voted for the character they wanted me to make.

She’s a fun, cute character who works at a radio station. Nothing overly complicated, ust an enjoyable character to experiment with and use in different situations.

The LoRA was trained entirely on clothed images, and it responds very well depending on how strongly you apply it and what kind of scene you want to create.

I recommend using it at around 0.5–0.6 strength, but you can safely push it all the way up to 1.0. Going above 1 may eventually introduce some image degradation, especially when using FP8 models. She was made with clothed and naked images in the dataset, She can get a little playful and mischievous too, just make sure you guide her with the right prompt.

ex: Small perky breasts, ....
Trigger: Sophia

There isn’t much more to explain, she’s simply a cute, playful character made for you to have fun with. Enjoy!