This LoRA is designed to reproduce Sairenji Haruna with a single trigger word.

It maintains stable character identity across a wide range of facial expressions (neutral, serious, smiling, embarrassed, surprised, etc.).

Haruna’s original hair color is black, and this LoRA is trained with a black-hair base.

This LoRA performs best at weights around 0.70–0.75. (best:0.72)