Brief
Lora model of Vivian Banshee from Zenless Zone Zero. Trained with 1024x1024 resolution for SD1.5 and SDXL, 1536 x 1536 for Pony and Illustrious, and enhanced with 3D model for best accuracy. (actually majority of the mataset is 3D model screenshots )The training dataset is categorized to further improve the flexibility, making my model easy to remove/change the costume.
Usage
Trigger prompt is: vivian banshee
Additional prompts for her default in-game maid costume: frilled hairband, guimpe, white shrug, ascot, white sleeveless shirt, layered skirt, print pantyhose, platform heels
Please note that including all prompts will have the complete set of costume, remove one of more will remove the associated part of costume in most cases, so you know what to do.
The model was well trained that the default costume can be removed/changed easily even with high lora weight, recommended weight is 0.8~1