类型
LORA
发布时间
2023-07-10
基础模型
SD 1.5
触发词
otokawa_saori
复制










Otokawa Saori from Discipline
Use green_eyes, blue hair tags and hair_ribbon if needed.
Trained using DAdaption @ 180 epochs (1 repeat) and seems to give fairly good results but struggles with blending clothing types together.
The AdamW8Bit version does a better job overall but loses a little clarity. I personally will probably switch back to this method of training and increase the amount of epochs to increase detail. It's all a learning curve for me.
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
