Reference from Honkai StarRail Character, Black Swan.

Due to nature of SDXL Using lot of VRAM, it advise the user to adding quality prompt such as masterpiece, best quality and lot more. User may copy generate info from one of image that been posted instead.

If user still wish to utilize hires, it is advise to use with Tiled Diffusion & VAE.
For Great result, please use Animagine XL V3

Simplify about Model:

• This Model are train with Animagine XL V3 Base

• default weight is recommended

• The hair braid may not be consistent

How to use Lora:

• -download the LoRa Model

• -move this data into your Stable Diffusion -> Models -> Lora