This is a LoRa model that trained by 40+ img generated with NijiV5, and it's a SDXL version.

Thank you for your assistance during the training set collection process, 時ばちゃん .

This model can generate semi-transparent silhouettes under starlight, which can be used to enhance the native effects of the XL model. The starlight flickers inside the shadow, and the color can be controlled through color + theme.

weight 0.6-1.0 maybe generates better imgs,trigger is ladyshadow.