A model trained on 74 images of Hyoon, or @aikuros. I've found it works best around 1.0 to 1.2, but YMMV. My first LORA, so don't expect it to be perfect. Hopefully someone finds this useful!