我们需要真正理解女性年龄的模型而不是只能生成萝莉！
We need a model that truly understands female age, not one that only generates "loli"!
所以这是一个旨在生成成熟女性并且试图理解多样的女性年龄，而不是只能生成小孩、女孩、萝莉的模型。
Therefore, this is a model designed to generate mature female and attempts to understand the diverse ages of women, rather than only generating "children, girls, or loli."
我的推特:@ekll01
My Twitter: @ekll01
版本说明请查看About this version
Please see "About this version" for the version description.
Please refer to the workflow that I have disclosed.
目前支持的年龄tag
女性：lxlx、mesugaki、bishoujo、lady、mature lady
男性：shxtx、bishounen、gentleman
请您用以上tag代替1girl、1boy。
Currently Supported Age Tags
For females: lxlx, mesugaki, bishoujo, lady, mature lady
For males: shxtx, bishounen, gentleman
Please use these tags instead of 1girl or 1boy.
Recommended Settings（general）
CFG: 5
Steps: 40
Sampling: Euler a
Scheduler: simple
Resolution: 832*1216 above
希望喜欢这个模型的朋友能够量力为模型进行一下Bid投资。这不是请求，但是如果有人能够帮助模型进行Bid，我会十分感谢！
I hope that friends who like this model can consider making a "Bid" investment in the model according to their ability. This is not a request, but I would be very grateful if someone could help the model with a "Bid."
Description
ONEPRO
The ONEPRO version is a fix on the ONE version to the problems pointed out by the user "byoobthusiast" and some of the issues I found myself in a few days of use.
There is no room for more changes in matureritualONEPRO in my power at the moment, and then unless there is a breakthrough, matureritualONEPRO will likely be the final iteration.
ONEPRO版本是在ONE版本上修改由用户“byoobthusiast”指出的问题和我自己在几天的使用中发现的一些问题。
目前在我的能力内matureritualONEPRO已经没有更多修改的空间了，之后除非有突破，不然matureritualONEPRO版本就是最后的版本了。
generation guide
生成设置
CFG 6
Step 40
Sampler采样器: Euler ancestral
negatives 负面提示词：
worst quality,bad quality,lowres, bad proportions,bad anatomy,bad reflection, logo,text,watermark,username,date,signature,artist name,jpeg artifacts
FAQ
Comments (36)
ONEPRO came fast! Thanks for the quick turnaround. Do you feel a difference between Euler A at 30 steps vs 40 steps? I know it makes a difference for Euler but I've never really tried Euler A that high. I'm totally willing to if there's even a small improvement.
In my own tests, the anatomy was good at both 30 and 40 steps, with stronger hand detail at 40 than at 30
感謝你的付出, 我會大量使用這工具
谢谢，也希望帮忙宣传一下
今年的工作伴随着onepro版本的发布已经结束了，我要休息几天来玩三国志14，之后会投入到lora制作中。
希望大家能帮忙多多宣传一下，感谢各位使用者。
This year's work has concluded with the successful release of the OnePro version. I will take a few days off to play Romance of the Three Kingdoms XIV, after which I will fully dedicate myself to the development of LoRA.
Hope you can help more publicity. Thank you to all our users for your support.
Ver.２、３とすべて使っていましたがここにきてONEでブレイクスルーした印象。
私の大好きなMILFキャラが出し放題状態で下半身が乾いている暇がないです。
楽しんで
MILF最高
“我们需要真正的女性而不是女孩！”
这就是我点赞的理由！
这个模型非常好，它能生成像御姐的角色，但好像没什么随机性，我使用同一组的提示词，随机种子随机，生成的图像都差不多没啥改变，这是优点还是缺点呢，是我使用方法有问题吗？
我用的是onepro
@460672106622 不是你的方法有问题，是目前存在这个缺点，暂时年龄和多样性不可调和（以我的个人能力来说） 现在正在尝试一个年龄稍微年轻一点但是更多样性的模型
@EKLL 谢谢回答，祝愿你能找到更好的方法去训练模型，这个熟女的模型很好，生成的女性非常有韵味
@460672106622 谢谢 我正在努力尝试解决这个问题，无奈本身熟女的资源就太少了，谁让现在是萝莉的天下
@EKLL 对啊，我不太懂这些模型训练，其它模型就是使用成熟女性的提示词，也不会生成这个模型的熟女那样的，都有点萝莉的感觉，也有点像动漫里女高中生的感觉
@EKLL 问一下，那种用手使劲抓胸，让胸凹陷下去很多的，或者衣服勒的胸很紧，让胸变形的，总结就是，让胸受力变形的提示词，我用反推的插件，推出来个skindentation,我使用这个提示词，生成的图像并没有太大变化
@460672106622 可以试试添加grabbing_another's_breast、breasts_squeezed_together
@460672106622 或者直接用自然语言描述，模型是可以一定程度上理解自然语言的
@460672106622 Use the 'deep skin' prompt instead of 'skindentation' and you'll get the result you're looking for.The idea behind neuron activation is data. On the Danbooru website, there are 115K images with skindentation tags and only 12K images with deep skin tags. The former is an order of magnitude higher than the latter.The 'Skindentation' effect is so common in anime drawing that typing 'Skindentation' as a prompt doesn't work. The anime model has already internalised the 'Skindentation' patterns.
I do like this Model alot as this is one of the few (or the only one) which actually is able to generate mature people properly. Nevertheless this model seems to have problems with skin tone variations. Please confirm whether anyone else cannot generate black/dark skin etc.
Thank you for your use. I have just tested the question you raised and can generate a picture of black skin. Your Prompt can be provided at your convenience.
@EKLL Thx for the feedback. Guess I need to keep trying.
@houseconspiracy535 euler_a ays\cfg5.0\Step 20
negative(worst quality,bad quality,lowres,
bad proportions,bad anatomy,bad reflection,
logo,text,watermark,username,date,signature,artist name,jpeg artifacts,)
This is my setup.
@EKLL This actually helped, many thx. Another question: is there any special trigger for furry or robots? I am trying Lola Bunny for example with "furry", "female furry" or "body fur" but the generations always end up human. Same with robots like EDI from Mass Effect or Cortana from Halo.
@houseconspiracy535 With regard to the furry, I utilized the following prompt:
furry female with body fur and fluffy fur, red eyes, long blonde hair with pink highlights (1.2), fox ears and tail, standing pose, fingerless armlets, a lace-trimmed red shirt revealing cleavage, black miniskirt, thigh-high stockings, open-toe footwear, heart pendant, choker, paws with small claws, large breasts, upper body close-up, looking directly at the viewer from the front, smiling with an open mouth displaying fangs.
This prompt can be successfully generated.
For the robot, I employed the following prompt:
looking directly at the viewer against a simple white background, red eyes, full-body view, equipped with a weapon such as a sword, pink eyes, no human elements, glowing effect, robotic design with claws, mecha style, science fiction theme, floating posture, open hands, glowing eye, one-eyed or extra eyes, fusion elements, thrusters, super robot, redesign focusing on mecha features.
This prompt can also be successfully generated.
Excellent model! Excellent work 👍 I really like the style you gave it. Now it will probably be my main model on par with Oblivious Mix (I still need to test it, but so far everything is great, it seems to be working fine), I didn't think that anyone could compare with Goofy_Ai and his models, they are very good, several heads above their analogues. But you did it! This model definitely deserves to be at the very top of the top❤️
Thank you very much
熟女熟男控最爽的一集
我好了，谢谢茄子
你好，我使用的是秋叶的SD-WeiUI，8gb显存的4070lap，但我遇见了一些问题。我在使用别的6-7GB大小的checkpoint时，就算我把模型全部加载进显存里，也可以直接生成1700*1700大小的图片。但使用这个checkpoints时，先是提示我【NansException: A tensor0 with NaNs was produced in Unet. This could be either because there's not enough precision to representthe picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer 2 to float32" option inSettings > Stable Difusion or using the -no-half commandline argument to fix this. Use -disable-nan-check commandline arqument todisable this check.】，我只能靠在launch.py文件里加入【sys.argv.append("--no-half")】来解决。之后如果全部加载进显存里就只能生成74*74大小的图片，再大就会提示显存不足。我只能在秋叶的启动界面把模式从【（SDXL）8GB以上显存】更改为【低显存（4GB以下）】来解决，这样我的显存占用会一直维持在4GB以下，但我的内存和SSD却会一直维持在几乎100%的状态。
Details
Files
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.