类型
LORA
发布时间
2023-05-08
基础模型
SD 1.5
触发词
a woman
复制


Delightfulhug (Nikki)
THE BEST WAY TO GET ACCURATE RESULTS
Make a regular 512x image and manually upscale it through image to image. If you need to use hires fix I used Uniscale balanced and UltraSharpX4 they looked okay but changed the face a bit.
trained on "a woman" I would recommend not using any photo jargon starting prompt with "a photo of" just start every prompt with a woman and go into details, but do whatever you want blud.
Photos are very prone to have slight watermark since every trained photo contained one
1. 转载模型仅供学习与交流分享,其版权及最终解释权归原作者。
2. 模型原作者如需认领模型,请通过官方渠道联系海艺AI工作人员进行认证。我们致力于保护每一位创作者的权益。 点击去认领
