mix of many models, VAE is baked，good at NSFW
很多模型的混合，vae已经烘焙，擅长nsfw
setting:
Denoising strength: 0.55,
Clip skip: 2,
ENSD: 31337,
Hires upscale: 4,
Hires steps: 15,
Hires upscaler: R-ESRGAN 4x+,
Eta: 0.68
2023.3.19 update
添加了两个版本，0.8a更真实，0.8c更动漫画
Two additional versions were added, 0.8a being more realistic and 0.8c being more anime
2023.6.13 update
更新测试版本0.9，此版本非常不稳定，很考验玩家的负面关键词使用熟练度，达到示例图效果需要下载所需lora，必备面部修复插件和动态阈值插件，此版本不再烘焙VAE，但我个人建议还是使用vae-ft-mse-840000-ema-pruned,示例图全部由此VAE生成，经测试此版本容易出现手脚崩坏和服装单一问题，建议使用相关lora配合使用，所有示例图片可以直接拖入SD图像浏览器查看相关参数
Update test version 0.9, this version is very unstable, it is a test of the player's proficiency in using negative keywords, to achieve the effect of the example image, you need to download the required lora, necessary face repair plug-ins and dynamic threshold plug-ins, this version no longer bakes VAE, but Personally, I recommend using vae-ft-mse-840000-ema-pruned. All the example pictures are generated by this VAE. After testing, this version is prone to hand and foot collapse and clothing oversimulation. It is recommended to use related lora to cooperate with it. All picture parameters Can be directly dragged into the SD image browser to view related parameters
关于图像模糊问题
请使用controlnet进行分块放大修复，如果这还帮不到你你可以借助外援继续放大，电脑不太好的玩家建议使用Topaz和Upscayl进行放大，Upscayl目前是免费的而且速度很快，相关使用教程可以在YOUTUBE和BILIBILI上找到，有很多的视频博主讲解了contronlnet的使用方法，另外图像设置也尽量按照示例图参数不要随意改动，图像采样一般只用DPM++SDE，超级分辨率放大算法请使用4x-UltraSharp，当然其它的算法也可以自行尝试
About image blur
Please use controlnet to zoom in and repair the blocks. If this still doesn’t help you, you can continue to zoom in with the aid of foreign aid. Players with poor computers are recommended to use Topaz and Upscayl for zooming in. Upscayl is currently free and very fast. Related tutorials It can be found on YOUTUBE and BILIBILI. There are many video bloggers explaining how to use contronlnet. In addition, the image settings should also follow the example image parameters as much as possible and do not change at will. Generally, only DPM++SDE is used for image sampling. Super resolution zoom-in algorithm please Use 4x-UltraSharp, of course other algorithms can also be tried by yourself
写了这么长简介还不看，在评论区问的一律拖出去斩首
If you don’t read such a long introduction, anyone who asks in the comment area will be dragged out and beheaded
Description
more anime
FAQ
Comments (74)
do anyone know where the Hires upscale and Eta
settings are?
在face repair 勾勾旁边罢了
@elbert thz a lot...i'm so new to this stuff
@lamfong223648 Click "Hires. fix", it below "Sampling method"
@imax210 thz~
can you share to source to train this? what content did you use? I'm interested to know the original artists, thanks.
It may be difficult to answer this question. This is a fusion model, which incorporates many good models in my opinion. I can't remember exactly which ones, because I have integrated several versions, about 10 of which also include models that have been deleted by the original author, so it is difficult for you to restore them
@zhazhahui345 Hello, may I ask if anyone knows that this 3D image can be created using this model? I would like a 3D rendering like this one
why we say vae is (baked)? what does it mean? the "baked"
It means the VAE is included in the checkpoint. You don't need to download and use a separate file for the VAE
为什么生成的图总是不清晰呢？？求大佬解答，感谢！！
重回幅度开低了是不是
@chosen 这是哪个设置？我这是英文界面
Sampling steps 开高点？
@treaenc771 我都是开20，需要开多少呢？
@627371157117 开超分辨
很多時候是lora的影響，如果有用lora，先拿掉再試試
@zhangshuang 就是高分辨率重绘不是有个重绘幅度吗
參考一下這些關鍵字:
best quality, ultra high res, photorealistic,detailed beautiful girl,
8k, RAW photo, best quality, masterpiece,
realistic, photo-realistic,physically-based rendering,ultra-detailed,
模型不错就是太涩了不敢用
说实话，我对比了几十个类似风格的模型，这个综合性能排第一
能不能压缩一下这个模型，4g有点大
老哥，救救我。为啥我生成的全是怪物脸
没开超分
可以试一下FaceEditor的修脸插件
真是好东西，操作简便，生成的效果也巨赞
请问 vae 应该选自动还是无？
vae名字叫打死不看模型介绍
Why do my eyes look weird? not as refined as shown
画出来都有点糊，画300x 445看起稍微清晰点，是需要设置什么吗？求大佬解答，感谢！！
开高分辨率修复啊，简介里都说了，你们不看简介的吗
@zhazhahui345 感谢，后面看了，小显存吃不消，高清修复太消耗显存了。
@cnvista604 老哥，可以尝试一下MultiDiffusion+VAE，原理是把图切分为小块进行分辨率修复，对小显存比较友好，然后生成好之后再通过后期处理放大2倍，出来的效果就挺好的了，但是有的时候参数调不好会有毛边，可能需要你自己按照自己的机器配置进行一下调整
很强大的模型，和各种lora的协调性也很好，就是不太会画神兽（滑稽）
为什么我画手都是特别畸形
Does anyone know what the promt, model, lora, etc. is? of the cover image of this checkpoint. (the girl with the pink hair and rabbit ears).
@zhazhahui345 Can the author please provide details of the cover image?
does anyone know why my images is kinda blurry and bad resolution? any other model works fine except this
I have same issue, try other prompts or cancel your VAE, maybe it could help
@CapRogers im putting my VAE to automatic is that good?
@shinjong866834 Try using it and not using it. Automatic is not really helpful
请问0.8版本和0.8A，0.8c 都有什么区别呢？0.8包含了a和c吗？
果然人类本质都是不看简介的，我写这么短就是让你们能看完的，你们竟然还是不看，我服了，0.8a经测试更写实，0.8c经测试更动漫化，0.8是更综合
@zhazhahui345 没开翻译看不懂2333
@zhazhahui345 大佬也可以在原版加些触发词用来指向画风。
我明明把所有的敏感选项都开了 为什么还会出现“此创作者的任何图片均不符合您的内容偏好。”
是啊，很奇怪
同问
对的 大家整的可以看 就是作者整的看不了
哦吼 浓度太高了 限制级也接受不了的程度
很多P站作者用了大佬你的模型，而且很多韩风的，我还以为用的韩国模型，原来是中国的大佬，太酷啦！
p站用这个做视频？
@GentleNight PIXIV，不是PH。
拿来做lora的底模效果不大好，但是生成图的效果很棒！！！NB！！
good!
麻烦问一下，我的设置是
Hires steps 15 Denoising strength 0.55
生成高清图后 图片经常和我原图不一样 有的时候会多手 或者脚变成手 请问这种情况应该调整哪里
不开高清修复生成小图手部没有问题就直接放进control net，用tile模型固定构图
@1198954804205 好的谢谢
@1198954804273 我也是用了很久，遇到同样的问题，看了你的回复才知道tile是干这个用的，之前从没用过。我得进一步研究fill的用处了
This is Fantexi v0.8 but with Chinese name. making it impossible to search in English. requesting developers to kindly add "Fantexi" in the name. or any other English name so it can be searched. thanks
Generates breasts not naturally a bit sagging, but close to silicone implanted ones.
感谢作者大大，用了几个月后回来，这是全站最好的2.5D模型，不接受反驳，希望作者大大尽快更新新版本
大佬什么时候能推出新版本呢？这个模型真心用得比其他的好太多了，特别是脸
My results are using this model is blurry and not as sharp as some others, even though I'm using the same prompt and generating the same image, minus the blurriness.
大佬什么时候跟新？你这个模型很好用，脸部特别精细
這個真的強
Can you please provide details of the cover image?