用戶LoRA
Very similar to v1, but this time there is no need for trigger words. Optionally however, you can include "black eyes", and "portrait"/"full body".