Models>
Character | Lucifer – ObeyMe
Quick Generate
Model Description by Creator
Recommended parameters:
推荐参数：
Clip skip: 2
weight:0.4~1
Sampler: DPM++ SDE Karras | DPM++ 2M Karras
Steps: 25
Size: 512x512|512x768|768x512
CFG Scale: 7
————
Instructions:
1.Fan art, please do not use for commercial purposes.
2.In addition to the default trigger "luci\(obeyme\)", it's recommended to add another prompt "red eyes".
3.Common 2D base models can be used, but the art style generated may not vary significantly (overfitting). For pastel style, it's suggested to increase the weight. So far, it hasn't been successful in generating a good-looking thickly painted style (perhaps controlnet could be used to achieve it).
It is not recommended to use 3D base models.
Realistic base models can be used and will default to an Asian face.
4.Proficient in drawing portraits and upper body, but not skilled at drawing full body (if you insist on trying, please be mentally prepared).
5.If you want more variation in the art style, it is recommended to use "lora block weight" to decrease the weight of the background and pose.
6.If you want a stable output of costumes, it is recommended to set the "lora" weight to 0.9-1, without using layered weights. Increase the weight of the corresponding costume trigger words in the prompt. The focus should be on the costume, not the pose or composition. Lower the resolution and generate multiple images in one run, as there will be hits eventually. You can also try describing the clothing in the prompt, but be cautious as the base model might be influenced.
The following table shows the trigger for costumes.
https://docs.google.com/document/d/1JQjzlG6iV7vu0amfOMsqxzPDDt6MShSH6XDoC29NnQg/edit?usp=sharing
说明：
1. 同人作品，请勿商用。
2. 触发词除了默认的luci\(obeyme\)，最好再加一个red eyes。
3. 常用的二次元底模都可以跑，但跑出来的画风不会差太多（过拟合了），粉彩类的建议提高权重，目前没成功跑出过好看的厚涂风（也许需要借助controlnet来实现）。
不建议使用3d类底模。
现实类底模可以用，会默认出亚洲脸。
4. 擅长画肖像和半身，完全不擅长画全身（非要尝试的话请先做好心理准备）。
5. 想要画风差异大一点，建议使用lora block weight降低背景、动作权重。
6. 如果想要服装稳定出，建议lora权重在0.9~1，不使用分层权重，prompt里提高对应服装触发词的权重，不强求动作和构图，分辨率调低一点一次多跑几张，总有命中的。也可以尝试描述衣服的prompt，但容易被底模污染。
以下表格是服装的触发词。
https://docs.google.com/document/d/1JQjzlG6iV7vu0amfOMsqxzPDDt6MShSH6XDoC29NnQg/edit?usp=sharing
————
If you have any ideas, suggestions, or questions during use, you can ask them on Discord ↓
使用中有任何想法&建议&问题，可以点上面的链接进Discord里提问，也可以在哔哩哔哩提问↓
https://space.bilibili.com/1328667069
Also, we made a tutorial called ‘AI Painting Tutorial-How to Create Stunning Images with Stable Diffusion’, which can help some creators who need to level up their image-making skills. You can find the link in the FAQ section on Discord.
此外，我们制作了一个教程“AI绘画教程-喂饭级提高出图质量全流程（Stable Diffusion）”，可以帮到一些需要提高跑图质量的创造者，链接在Discord的FAQ里就可以看到。或者直接点击下面的链接。（也有视频版，但也许文字版更清晰一点，可以先使用ctrl+f搜索关键字找找有没有你遇到的问题，希望可以帮到你~）
https://www.bilibili.com/read/cv23237755
————
Buy us a cup of coffee and we’ll spill the beans.
https://ko-fi.com/co_subcreation
Or Fanbox
https://co-subcreation.fanbox.cc/
Or爱发电
https://afdian.net/a/Co_Sub_creation
Your support is the fuel that powers our world-building. Your ideas are the bricks that shape our world-making!
您的支持将成为创造世界的动力，您的想法将成为构建世界的基石！
