本次模型的主题是抖音网红美女：栗子熟了，温馨予，训练基于chilloutmix模型，训练素材都来自于抖音视频截图，主要以面部为主，所有照片来源均来自网络，每张照片均进行50次运算，最终得出LoRA模型。Tag可以使用lizishule，因为训练图大多为面部，而且是抖音视频截图，没有高清照片，可能产出图片显示人物清晰度不够，可以适当补充tag完善。
The theme of this model is Douyin (TikTok) popular beautiful female influencers: Lizi Shule, Wenxin Yu, trained based on the chilloutmix model. The training materials are all from screenshots of Douyin videos, mainly focusing on the face. All photos are obtained from the internet, and each photo undergoes 50 calculations to ultimately produce the LoRA model. The tag “lizishule” can be used because most of the training images are of faces, and they are screenshots from Douyin videos, so there may be insufficient clarity in the resulting images. Additional tags can be added to improve the searchability.