Draw whatever you want, really, anything goes. I can do anything, just look at me. As long as it’s real content, I can basically meet your requirements, whether it’s light and shadow, vision, pets, architecture, etc. No problem.
There are no rules. Write whatever prompt you want. Just use whatever you like.
Recommended image size should not exceed 1024. You can increase the resolution through high-definition repair.
It is recommended to use the DPM series sampler with a sampling step of 30 or more.
The rest is up to you.
Originally, I wanted to use it as a bottom model for my own training model, but I found that its performance is quite good, so I released it for everyone to use.
It integrates many models from many experts, so the model has good performance.
随便出图，真的，随便出，想什么搞什么，我这什么都能搞，不信你看 只要真实系的内容，你想要的，我这基本都能满足你的要求，什么光影啊、视觉啊、宠物啊、建筑啊、都没问题 没有什么讲究，提示词想写什么写什么，就是随便用 — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — 推荐出图尺寸不要超过1024，可以通过高清修复提高分辨率 推荐使用DPM系列采样器，采样步数30以上 剩下的随便搞 — — — — — — — — — — — — — — — — — — — — — — — — — — — — — — 本来是想拿来做自己训练模型的底模的，发现其实表现力都还挺不错的，就发出来大家一起用 融了很多大佬的模型在里面，所以模型才会有很好的效果