LoRA created from 100 images with art by 小野妹子:

Recommended weight 0.5 to 0.8 depending on your model and prompt.

No trigger word is needed, but you can use 'onono_imoko' if you want.

Due to artist influence realistic models may dismiss some prompts.