yunyouyou

详情

模型描述

the best character lora i have tried.

The character training set is 512 * 512, so theoretically anything larger than this resolution can be used, and the issue of landscape is taken into account during training, unlike most loras with limited materials, which have poor performance on landscape images.

Due to my 'method', certain specific words may cause collapse. If this happens, please add 'naked' or 'faceless' in the negative prompt words

此模型生成的图像