You can download my LoRAs that haven't been released here a few days earlier.
Chaeyoung
Please consider to support me by buying me some coffee so I can make a new LoRA faster. I do open commissions there too if you want to make your own custom LoRA: My BMAC page
Trigger word: egyiren
Best checkpoints/models to use:
Realistic: VisionGen, ChillOutMix, ArtEros, Snapdd00, Henrealmix
Fanart/painting style: Dreamshaper, LucidMix, SunshineMix, Majicmix
Join KPOP Diffusion Discord: https://discord.gg/NrBDMX79Da
egyirenYou can download my LoRAs that haven't been released here a few days earlier.
Saerom
Championing the Beauty of Handsome Korean Men
1. lots of anime characters and women, but not much about men.
2. and all the male models are so ugly!
So I'm distributing this model for the value of handsome Korean men (even though I'm a man).
*Use the 'man' prompt, not the 'boy' prompt.
*no NSFW
iamthezenon@gmail.com
-
Stable diffusion is a realm where imagination meets art, and the ordinary blends seamlessly with the extraordinary. Its vast universe is populated with a myriad of characters – fantastic beauty, fearless heroines, celebrities, and so much more. But in this colorful and diverse realm, one might notice a peculiar imbalance: a conspicuous lack of appealing male characters.
Why is the charm and allure of male characters often overlooked, replaced instead by exaggerated features that often border on caricature? Why do their female counterparts, in stark contrast, get to bask in the limelight of beauty and grace? It seems that the narrative of generative images has been skewed, subtly implying that aesthetic appeal is the exclusive domain of women, while men are left with the short end of the stick.
And I, for one, respectfully disagree.
Even though I'm a man, I feel compelled to champion the cause of handsome Korean men. Not merely as a reaction to the prevailing narrative, but as a testament to the fact that men, too, can embody beauty, charm, and appeal. It is a disservice to the art form itself to limit our interpretation of 'man' to the 'boy' prompt. It is high time we allowed the 'man' to exist in all his glory, celebrating the spectrum of masculine beauty just as we do the feminine.
In light of this, I've decided to introduce a model that serves to bridge this gap. This model is a tribute to the handsome Korean man, defying stereotypes, and championing the cause of aesthetic appeal in male characters. This character is not just a representation, but a statement, a challenge to the status quo, and a call for diversity.
But this initiative is not about undermining the representation of women or female characters. Quite the contrary. It's about balance, about showcasing the beauty of both genders, about creating a universe that is truly reflective of our own.
So let's reimagine the narrative, let's rewrite the story. Let's celebrate the value of a handsome Korean man, not just in the name of diversity, but for the sake of equality, representation, and authenticity. Let's step beyond the confines of the 'boy' prompt and embrace the 'man' prompt wholeheartedly, asserting that men too have a rightful place in the world of beauty and appeal.
Because Generative-AI, like art, should be a mirror to society, reflecting all its vibrant colors, not just a selected few. And it's high time we acknowledged the handsome Korean man's rightful place in this beautiful realm.
man
Generates Necronomicon pages, trained off 67 images found using Bing image search for 'Necronomicon pages'
Include Necronomicon Sketch in the prompt for best results.
Images were made using this with the JIM EIDOMODE checkpoint
Necronomicon Sketch
CFG Scale：4
lora'strenth：0.5-0.8
Clothes prompts：white sleeveless jacket、denim shorts、twin tails、black long hair
trigger word：amano_hina
width and height: 512*768
Denoising strength: 0.45
Upscaler: SwinIR 4x
Upscale by 1.8
刚刷完天气之子，太感动了，所以把阳菜酱做成了模型
CFG Scale推荐4
权重：0.5-0.8
服装tag：white sleeveless jacket、denim shorts、twin tails、black long hair、blue eyes
触发词：amano_hina
一定要用高分重绘生成图片，参考设置如下：
分辨率: 512*768
重绘幅度: 0.45
放大算法: SwinIR 4x
放大倍数：1.8
amano_hina
Here's my favorite armor set from FFXI!
Trained on 22 green screen captures from Altana Viewer. I've had success around 0.75 weight.
About half of the training images also had him holding a Temple Knight Army Shield
and wielding a generic single-handed sword
Already taking notes for the next version. Thinking of including more training images with the helmet removed, and including all the races from FFXI.
Please note: This model was only trained on one of the male Hume avatars.
===If you like my work, stop by to hear my music or watch my videos!===
PLD_IM
It makes everything like vector art.
Or just more like a cool poster.
Created as described in the tutorial of konyconi but with some own tweakings: https://civitai.com/models/52697/tutorial-konyconi-style-lora
If you want, support me here:
https://displate.com/dervlex?art=5d91f0dbbcb29
https://www.patreon.com/user?u=90661954
in the style of 0mib
Sloth from Fullmetal alchemist
trigger word: sloth
sloth
Ganyuganyu (genshin impact)
Lan Fan from Fullmetal alchemist
trigger word : lan fan
lan fan
Please consider to support me by buying me some coffee so I can make a new LoRA faster. I do open commissions there too if you want to make your own custom LoRA: My BMAC page
Trigger word: bpjisoo
Best checkpoints/models to use:
Realistic: VisionGen, ChillOutMix, ArtEros, Snapdd00, Henrealmix
Fanart/painting style: Dreamshaper, LucidMix, SunshineMix, Majicmix
Join KPOP Diffusion Discord: https://discord.gg/NrBDMX79Da
bpjisoo
Making models can be expensive. Do you like what I do? Consider supporting me on Patreon 🅿️ or feel free to buy me a coffee ☕
This was requested many times, but I decided to bump it up in priority after it got requested again yesterday.
Should work fine at around weight 1 on AnyLoRA ckpt. Use zeronis to increase fidelity.
How to use LoRA's in auto1111:
Update webui (use git pull like here or redownload it)
Copy the file to stable-diffusion-webui/models/lora
Select your LoRA like in this video
Make sure to change the weight (by default it's :1 which is usually too high)
zeronis
可以使用这个lora，生成一些类似于手绘风格的图像
一定要加入geryscale和watercolor两组词，并避免输入其他颜色词
玩的开心!
You can use this lora to generate some hand-drawn style images
Be sure to include the words geryscale and watercolor and avoid entering other colour words
Have fun!
monochrome
Arknights Penance Ambience Sythesthesia 2023 outfit.
First debut in Ambience Sythesthesia 2023 PV. However, this is not a in-game outfit that you can buy. So much for a petty game company(Ofc I'm joking, or am I?).
Recommend weight between 0.7 and 0.9. Usually this model can reproduce her hair decoration and asymmetric thighhighs very well. If you want to add the sunglasses, just add eyewear on head, tinted eyewear to your prompts. Likewise, add single leg thighhigh, single leg pantyhose to your prompts if you find her legwear missing.
明日方舟斥罪音律联觉2023服装，辑速狂飙。
仅在音律联觉PV中登场，游戏里的皮肤别想了，画了不卖。手机里有一款心胸狭隘的游戏.jpg
推荐0.7到0.9的权重。基本上对于头部细节和非对称的丝袜还原得比较好。炼的时候没有把太阳镜提示词融进去，想还原太阳镜的可以加 eyewear on head, tinted eyewear。同样地，如果发现丝袜细节不对，也可以加上 single leg thighhigh, single leg pantyhose.
penanceeyewear on head, tinted eyewear
Lora trained on Androssi Zahard from Tower of God
small dataset (28 images)
tags: androssi zahard, 1girl
optional: single horn, short dress, cropped vest, fang, orange eyes
models used:
anime: CoffeeMix, Counterfeit-V3.0
2,5D: MeinaUnreal - V3, majicMIX, PsyGon Animerge - v1.1
negative embeddings: embedding:bad-hands-5, embedding:ng_deepnegative_v1_75t, embedding:bad-artist, embedding:EasyNegative
Recommended settings:
Make sure you are aware on the usage instructions of LORA
VAE: vae-ft-mse-840000-ema-pruned, kl-f8-anime2, blessed2.vae (for anime models)
Sampler: DPM++ SDE Karras or DPM++ 2M Karras
Steps: 20
Clipskip: 2
Upscaler : img2img in comfyui using 4x-UltraMix_Smooth
CFG Scale : 5 to 9 (mostly 7 and 8)
LORA weight for txt2img: 0.6-1
Denoise strength for img2img: 0.4 to 0.7
If you want the exact same settings, just drag and drop the picture into ComfyUI
for the pictures i used comfyui
androssi zahard
Hello everyone, this is the sister project of AnimeFus called Kyo Mix v0.3. We trained this project using anime characters from KyoAni into a single model. In addition, we have created a new project called AnimeChar, which aims to remake the characters with higher quality using Lora based on our dataset.
We are excited to introduce our new trained Lora model, Kurokawa Akane (黒川 あかね), a major supporting character in the Oshi no Ko manga series.
Trigger words: kurokawa_akane, [Optional: blue_hair, blu_eyes]
You can support our work by buying us a drink through the following link: https://www.buymeacoffee.com/ryalandai.
You can also follow our Instagram account for AI painting and model updates: https://www.instagram.com/rya.land.ai.
If you like our model, we would appreciate it if you could give us a 5-star rating.
Please refer to the 'About this version' section for more details about the model version and included characters. The model base we used is Anything v4.0, and we have provided sample images generated by at least two base models.
-- Please also support --
AnimeChar - Railgun Mix | Misaka Mikoto (御坂 美琴)
AnimeChar - Oshi Mix | Kotōbuki Minami (寿 みなみ)
AnimeChar - Hyouka Mix | Chitanda Eru (千反田 える)
AnimeChar - Oshi Mix | Arima Kana (有馬 かな)
AnimeChar - Oshi Mix | Hoshino Rubi (星野瑠美衣) [v2.0]
AnimeChar - Oshi Mix | Hoshino Ai (星野 アイ)
AnimeChar - Liz Bird Mix | Kasaki Nozomi (傘木 希美)
AnimeChar - Tamako Mix | Kitashirakawa Tamako (北白川 たまこ) & Asagiri Shiori (朝霧 史織)
kurokawa_akane
Trained on fine arts for hours, good at person, fantansy, animal, art style.
768*768 based on rev ckpt. can use "fusionart" enchance the style, weight 0.85-1 is ok.
Caption by LLM model , can recognize different tags.
Use simple negative prompt. You can try different settings and different models.
==================>
Welcome to use free zipzap chatgpt extension
https://www.google.com/search?q=zipzap+ai ,
ZipZap.AI - Free ChatGPT Powered AI Assistant
=================
Feedback is important to me ,Thanks for your likes and follow.
You can try it on ai-creator.net and ai-creator.net/arts with tags,based on different model or styles.
1987 Action Figure Playset Packaging
Trigger: [ videldbspinkdress, 1girl, solo, black hair, dress, blue eyes, hairband, short hair, pink dress, flower, rose, red hairband ]
This one was kind hard to recreate the samples, but it was a successful sampler trip XD.
videldbspinkdress1girl, solo, black hair, dress, blue eyes, hairband, short hair, pink dress, flower, rose, red hairband
中文 | English
仿ももこ 的作画风格的LoRA模型。非常精致的画风。
权重建议值：1 (or 0.6~1.4)
基于模型：Anything-5RE
训练信息：共计562张原画（含裁剪图），迭代2248步，多分辨率迭代训练。
其他的一些建议：这位画师大大的画风很充实，建议多加些装饰物词条，以体现细节。
LoRA model with a painting style imitating MomoKo. The art style is extremely delicate.
Recommended weight value: 1 (or ranging from 0.6 to 1.4).
Based on the model: Anything-5RE.
Training information: A total of 562 original paintings (including cropped images) were used for training, with 2,248 iterations using multi-resolution iteration training.
Other suggestions: This artist's painting style is very detailed and rich. It is recommended to add more decorative tags to reflect the details.
1girl
Just a try for fun. I wish there will be someone else to make a better one.
娱乐性尝试。望能抛砖引玉。
In mose cases it is not compatible with normal clothes like white dress……that is to say, "nude" or "naked_xx" is better. However, this lora might still bring you some shader effects.
大多数情况下它与白裙子这种日常服饰并不兼容……这意味着“nude”或者“naked_xx”效果更好。然而，这个lora还是能带给你些许光影效果。
Some tags that can bring interesting results: white skin, blue skin, blue theme, mechanical parts, bone, spine
一些能带来有趣结果的tag：white skin, blue skin, blue theme, mechanical parts, bone, spine
translucent skin
Get early access to build and test build, be able to try all epochs and test them by yourself on Patreon or contact me for support on Discord
To support me you can buy me a coffe at https://ko-fi.com/antonioriolo
Recommended settings:
Make sure you are aware on the usage instructions of LORA
Sampler: DPM++ SDE Karras (Recommended for best quality, you may try other samplers)
CFG Scale : 5 to 10
LORA weight for txt2img: reccomended 0.3-0.7. (I used 0.5)
Keys: wearing latin_dance_dress, dance floor
Very simple to use, just use the keywords, i advice to try with 1.3 weight in all keywords.
As i did in my generated images you can prompt somthing like "red dress" to have a colored one, it's not necessary a weight on it usually.
dance floor keyword is not necessary but can give a little context to the image.
Some of the images are generated using controlnet poses pack that you can find in my page.
Works also with anime.
Enjoy!
If you'd like to support my efforts and have access to early and bleeding edge builds, please think about joining my Patreon: https://www.patreon.com/betterCuteAsians
wearing latin_dance_dressdance floor
Wes, the protagonist of Pokemon Colosseum. Activation phrase is "WesPC".
WesPC
Elodie Chérie is a French former porn actress (1992-2004), born in 1966 in Saint Etienne, France.
This model is trained with SFW sources only. If you get NSFW content, it's on you.
Please provide constructive criticism.
You can support me here to be able to render better models (https://ko-fi.com/djo226). Supporters may request model.
Prompt : realistic, natural lighting, EC, post-processing, ultra realistic, highly detailed, very detailed, cinematic, hyper realistic, intricate scenery, sharp, photo by Annie Leibovitz, film, studio lighting, detailed skin, ultra realistic, bokeh, sharp features <lora:EC:0.9>
Negative prompt: (deformed iris, deformed pupils, semi-realistic, cgi, 3d, render, sketch, cartoon, drawing, anime, mutated hands and fingers:1.4), (deformed, distorted, disfigured:1.3), poorly drawn, bad anatomy, wrong anatomy, extra limb, missing limb, floating limbs, disconnected limbs, mutation, mutated, ugly, disgusting, amputation
Steps: 80, Sampler: DPM++ 2M Karras, CFG scale: 12, Seed: -1, Face restoration: GFPGAN, Size: 768x512, Model hash: 81086e2b3f, Model: realisticVisionV20_v20
Face restoration is optional, may provide better result if not activated.
Use EC as trigger. Add (chubby:0.5) to the prompt to match her during her carrer.
Important:
This Lora model is a virtual representation of a natural person and it is important to emphasize that its use must not be illegal or degrading. I want to remind you that any infringement in this regard is the responsibility of the user and not that of the designer of this model.
It is crucial to respect laws and ethical standards regarding the use of this technology. Any abusive or illegal use of this model is strictly prohibited and could result in legal consequences for the user.
As the designer of this model, I take the responsibility of ensuring that it is used in an ethical and responsible manner very seriously. I rely on each user to do the same and to act in compliance with the laws and ethical standards in force.
EC
枪神纪英雄角色‘桃丽丝’的人物模型。
推荐<lora:桃丽丝v1:0.6>或者<lora:桃丽丝v1:0.7>或者<lora:桃丽丝v1:0.8>
mxmtwintails, blonde hair, purple skirt, 
Hi, this is my first model and so I don't think it works perfectly.
So far, the settings I'd recommend are as follow:
Model: Chilloutmix, or PerfectWrold seem to work nicely.
Weight for the LoRa seems to work well between 0.5 and 0.8, but I think you'll need to experiment
DPM++ 2M Karras
Again, I think you'll need to experiment for yourself. I'll attach a couple of examples I made so you have an idea of how it looks.
Just a heads up, because of the photos I used to train this LoRa, she has a tendency to be bare chest or to have nips slipping out of her bra.
If the character ressemble a former JAV actress, it's only pure coincidence.
(As Requested)
I've been meaning to make this model for a while now. I know there are a few other models of her on this site, but probably my model is a bit different. have fun and try to compare between them.
Thank you for downloading. Please leave a rating and comment if you like my models. also leave a comment on what model would you like to see next. Please consider to leave a tip if you like my work 😌☕️ enjoy
https://ko-fi.com/meantweetanthony
want to get up to 10 epochs for your custom LoRA model requests + 50 generated images
https://ko-fi.com/meantweetanthony/commissions
Trained on 50 photos with ChillOutMix, work best on DPM++ SDE Karras sampling steps 20-30 CFG Scale 6. weight 0.8-1.0 is recommended for better results.
lightweight LoRA, works on any generator (easydiffusion,stablediffusion etc). share your results.
Disclaimer : Usage of my LoRA models is 100% under your own responsibilities. I bear no responsibilities on however the models are used and it is beyond by jurisdiction.
1girl
We were not blessed with images of too many minor characters, so we had to recreate them four times. We compromised.
It stabilizes at about 0.8 influence, but the pose is inflexible, as if it has been over-learned.
It seems the only way is to force the control net to change the pose.
I've only uploaded unrelated pictures, but I'd appreciate it if you could follow me on Twitter.
blue hairlipstickhair over eyes1990s (style)retro artstyle makeupuniformepaulettesnecktielong sleeves
This is LoRA for the Leaf character from Black souls. I tried very hard to make it, but there are still some flaws, such as weird nsfw generation and weird generation of clothes other than the main one. But otherwise, LoRA turned out very well and I hope for your positive feedback.
Triggers: maryann, origcostume, officialartstyle(only works to improve recognizability, does not convey the original style).
Also tags fairy wings, green leotard, pointy ears can help
Trained on anything v4.5
maryannofficialartstyle
For many Dm it is important that the background is white, remember that this must be configured manually.
For the rest, the amount of space given to the image, the dementions change the plate created quite a lot, so it is important to play with that, it is not advisable to always shoot the 512-512 in this Lora
Ultra complex food generator comes with style activated, the same can be deactivated by lowering the intensity of the Lora.
General use is to generate food for RPG bells.
The most advisable thing is to choose well, the type of background and plate where the food is placed.
This little parrot is incredibly generous in terms of the range of options, so you should be able to create whatever you want (as long as it's food or drinks - potions).
Lora to give women tails
Trained on Realistic Vision
Recommend adding "multiple tails" to negative prompt
tailmorph
I am the Reindeer Czar, I love Reindeer, but I also love Wolves. I am a Slav, and grew up with the cartoon Ну, погоди! I love the character Wolf. He is so fun. I like the new 3D series started last year.
Я царь оленей, я люблю северных оленей, но я также люблю волков. Я славянин и вырос на мультике Ну, погоди! Мне нравится персонаж Волк. Он такой веселый. Мне нравится новая серия 3D, запущенная в прошлом году.
I wanted to make him for SD, so here he is!
Я хотел сделать его для SD, и вот он!
Weigh the LoRA like 0.7 for best result.
Взвесьте LoRA как 0,7 для лучшего результата.
Tags to help:
Теги в помощь:
(tank top, watch, shoes, tail)
Later I will do Hare, and maybe also Badger and Gazelle.
Позже сделаю Зайца, а может быть еще Барсука и Газель.
Have fun! Post some cool images if you like!
Веселиться! Разместите несколько крутых изображений, если хотите!
NuPogodiNewWolf
Get early access to build and test build, be able to try all epochs and test them by yourself on Patreon or contact me for support on Discord
To support me you can buy me a coffe at https://ko-fi.com/antonioriolo
Recommended settings:
Make sure you are aware on the usage instructions of LORA
Sampler: DPM++ SDE Karras (Recommended for best quality, you may try other samplers)
CFG Scale : 5 to 10
LORA weight for txt2img: reccomended 0.3-0.7. (I used 0.5)
Keys: wearing soccer_uniform, back view, inside soccer_field, kicking the ball
Very simple to use, just use the keywords, i advice to try with 1.3 weight in all keywords.
As i did in my generated images you can prompt somthing like "red outfit" to have a colored one, it's not necessary a weight on it usually.
As you can see while using "kicking the ball" keyword, anything can happen, you can get amazing images of soccer action, but you can also see balls coming from nowhere with wrong sizes sometime. Just wait for a lucky seed.
Works also with anime.
Enjoy!
If you'd like to support my efforts and have access to early and bleeding edge builds, please think about joining my Patreon: https://www.patreon.com/betterCuteAsians
wearing soccer_uniforminside soccer_fieldkicking the ballback view
Suzuki Sonoko LoRA
Use 'forehead' prompt for on/off her bangs
Please consider to support me by buying me some coffee so I can make a new LoRA faster. I do open commissions there too if you want to make your own custom LoRA: My BMAC page
Trigger word: twicemomo
Best checkpoints/models to use:
Realistic: VisionGen, ChillOutMix, ArtEros, Snapdd00, Henrealmix
Fanart/painting style: Dreamshaper, LucidMix, SunshineMix, Majicmix
Join KPOP Diffusion Discord: https://discord.gg/NrBDMX79Da
twicemomo
中文 | English
仿Gomzi的作画风格的LoRA模型。
权重建议值：0.8 (or 0.6~1.2)
基于模型：Anything-5RE
训练信息：共计620张原画（含裁剪图），迭代1240步，多分辨率迭代训练。
其他的一些建议：当权重为1时，头发颜色好像有点过拟合了，必要时可以着重强调头发颜色。
LoRA model with a painting style imitating Gomzi.
Recommended weight value: 0.8 (or ranging from 0.6 to 1.2).
Based on the model: Anything-5RE.
Training information: A total of 620 original paintings (including cropped images) were used for training, with 1,240 iterations using multi-resolution iteration training.
Other suggestions: When the weight value is set to 1, the hair color seems to be overfitted. When necessary, it is recommended to emphasize the hair color.
1girl
pornstar Teagan Presley was born on July 24, 1985
🔧Model base: AngrA_AnimFlex_v01 , AngrA_RealFlex_v01
📝🌟 Trigger: quiron_TeaganPresley_v3, <lora:quiron_TeaganPresley_v3:1>
📝 Variation: (quiron_TeaganPresley_v3:X), X=0.8 to 1.4
📝 Variation: <lora:quiron_TeaganPresley_v3:Y>, Y=0.6 to 1.4
📝 Denoising strength: 0.5 to 0.65
🗒️💡Sample: quiron_TeaganPresley_v3, blonde hair, realistic lighting, trending on Artstation, medium breasts, <lora:quiron_TeaganPresley_v3:1.1>
🚀 This LORA likes: (best quality:1.1), (masterpiece:1.2), (realistic:1.2), (detailed:1.1),
👇Negative Prompt: paintings, sketches, (worst quality:2), (low quality:2), (normal quality:2), lowres, normal quality,
Or / And
👇Negative embbeding: bad_prompt, bad-artist, bad-hands-5,
Hire. fix: ok
Restore faces: ok
quiron_TeaganPresley_v3 quiron_TeaganPresley_v3, 
This is LoRA, in which I try to depict Kirio Hikifune from Bleach TYBW anime.
I had a small training data, I'll retrain it after her more showcase in anime (if Pierott will show her fight). Also you can add good results which I'll use in the next training.
Please, write all pros and cons of the LoRA, your comments and reviews allow me to remake the LoRA (cause I have no possibility to check it properly) and to make my future loras better.
japanese clotheskimonohaoricleavagelipstickweaponmakeup
license:creativeml-openrail-m
https://huggingface.co/spaces/CompVis/stable-diffusion-license
gold skin girl
(generation advice below) 👇
also known by her professional name Milena D, was born on October 18, 1987, in the Czech Republic. Milena began her modeling career in 2007, when she was discovered by a photographer who spotted her in a shopping mall. Her striking beauty and natural charisma caught the eye of many photographers, and she quickly became a sought-after model in the industry. Milena gained widespread recognition as an online model. Her alluring and sensual style, combined with her stunningly beautiful features, captivated audiences worldwide.
Aside from her modeling career, Milena has a passion for fitness and healthy living. She is an avid runner and practices yoga regularly, emphasizing the importance of balance and well-being in her life. Unfortunately, Milena's modeling career was cut short due to health issues. In 2017, she announced on her social media accounts that she was retiring from modeling due to a chronic illness. Her fans were saddened by the news, but showed their support and admiration for her courage and grace in handling the situation.
Today, Milena is still active on social media, where she continues to share her fitness journey and inspire others with her positive outlook on life. She remains a beloved figure in the modeling world and a role model for many.
Most images have been generated with LORA weighted @ .9 - 1.
Most samples done with sampling steps ~40.
Sometimes I just send it to Ultimate SD upscale after a quick render and run 2x at a scale of two.
Ultimate SD upscale script used @ ~.15 - .4 denoising using 8x NMKD Faces or 4X UltraSharp redraw upscalers.
The prompts continue to evolve. CyberRealistic is a reasonably flexible model for being realism focused. I still think they have trumped the SD 1.5 game. 👑
Leave a review if you liked using this.
Leave a comment if you are having issues getting good output, I'll try to respond.
MilenaAngel
This is a Rem inspired look with an Xmas theme, it's not meant to look exactly like Rem, simply in the style of her, which can be modified with different Loras and Models.
Steps 20-40
I find 640x960 resolution works well
High res fix x2 of Step 20 and denoise of 0.3-0.5.
Model trained with majicMIX realistic V4
Use these terms for best results:- rem, rem (re:zero), blue hair, Christmas theme, Long hair, Royal crown, hair ornament, hair over one eye
Easier to get NSFW results using these terms:- uncensored, no panties, skirt lift, pussy, clitoris,
Another model trained for AOM3 using the collab/guide by holostrawberry [Guide] Make your own Loras, easy and free - colabs | Stable Diffusion Other | Civitai
Another very nsfw model because we don't have enough of those on civitai. The artist makes a lot of furry and anthro content which can work but I haven't tested extensively. One of the results is that this lora has no problems with green/yellow skin but will anthropomorphize the character in the process.
A weight of 1 works when used by itself. For most character lora I used 0.8.
whispering_for_nothing
中文 | English
某位画师曾说过：“AI不可能模仿得了我的画风。”（请勿跳脸）
仿Mignon的作画风格的LoRA模型！虽说不可能有原画那么高的质量，从测试的效果来看还是很不错的。
权重建议值：1 (or 0.6~1.4)
基于模型：Anything-5RE
训练信息：共计567张原画（含视频截图），迭代1134步，多分辨率迭代训练。
其他的一些建议：必要时请加入词条：uncensored。
LoRA model with a painting style imitating Mignon! Although it is impossible to achieve the same high quality as the original paintings, the test results are still very good.
Recommended weight value: 1 (or ranging from 0.6 to 1.4).
Based on the model: Anything-5RE.
Training information: A total of 567 original paintings (including video screenshots) were used for training, with 1,134 iterations using multi-resolution iteration training.
Other suggestions: When necessary, please add the tag "uncensored".
1girllooking_at_vieweruncensored
I made a video on how to achieve the same quality or just repeat my art
P.S. If you want to support me, then you can subscribe to my Patreon and get builds earlier than they come out here (versions on patreon will always be newer by one than here) This will help me a lot in updating the video card over time to make models even better and faster (the main thing is better)
Eta noise seed delta 31337
LORA weight 0.5-0.6
Use Abyssorangemix3aom3_aom3a3.safetensors With these negative prompts:
(worst quality, low quality, extra digits:1.4) or (worst quality, low quality:1.4)
Trigger Words: 1980s \(style\)
Optional Trigger Words: retro artstyle
just use hires fix as much as your video card is capable of, so as not to give an error, with 8gb of video memory it is 512x768 with hires fix 2-2.15, upscaler Latent (nearest-exact), the picture will be much more detailed. If you want a style close to the original but with more detail, don't use Denoising strength above 0.6, but to avoid artifacts, do not use too low a value, optimally 0.5
CFG Scale - 5.5-7 (for some reason hands are a little better at 6)
The training took place at a resolution of 512x768, so it makes no sense to use 512x512
Combined with my other LORA
How to use LoRA's in auto1111:
Update webui (use git pull like here or redownload it)
Copy the file to stable-diffusion-webui/models/lora
Select your LoRA like in this video
Make sure to change the weight (by default it's :1 which is usually too high)
*Information taken from Lykon
Please post your work with or without comments, it will help me improve. Thanks!
If you like my work, click on the heart above, I will be pleased :3
1980s \(style\)
Lora for sticking a cock in a sock, specifically rubbing penis against leg inside thighhigh. Also works with futa flaccid penis in own thighhigh using "own penis in own thighhigh" (maybe also traps/etc). Futa can be triggered unprompted, this is because of disembodied penis being close to crotch and being converted to futa by SD, not because of futa images in dataset (tried training without them and it didn't help at all), you can add futanari in negative with high (1.4~) weight to prevent most of it.
Since this deals with legs, expect a lot of the usual leg nonsense, especially since SD is particularly bad with legs going up. The Lora already increases consistency with legs up compared to normal, not much more can be done. There also tends to be extra hands holding up the legs; you can try to prevent this by adding stuff in negative prompt like "holding legs" and "hands on own legs", it'll prevent some but not all.
If you want pov, lower the weight of the pov tag, otherwise it has too much influence towards vaginal sex. Most images will be pov-ish even without the tag in prompt anyway, so I would recommend not adding it, it doesn't mesh well with the lora.
penis in thighhighcum in thighhigh
<<<Revised version of removed sexy jk uniform>>>
Recommended aspect ratio 1:2 (example 250:500)→Hires.fix Upscale by 2(example 500:1000)
Trigger Words: seifuku cosplay,seifuku cosplayer
Recommended weights: 0.6-0.9
Recommended negative prompt: paintings, sketches, (worst quality:2), (low quality:2), (normal quality:2), ((monochrome)), ((grayscale)),((sofa,bed,carpet,rug))
suggested Lora combination: JapaneseDollLikeness,chinesedolllikeness,taiwanDollLikeness,KoreanDollLikeness
seifuku cosplay,seifuku cosplayer
My first lora.
Any feedback is apreciated since i'm still learning the process.
Can create her consistently but isn't the best at making the iron cross or the computation jewel. NSFW capable. Can easily create other outfits, works well with most anime models (from what I've tried).
Works with other Loras from what I've tried but results may vary.
Best at strenght 1 and (serebryakov:1.2) on promt
(serebryakov:1.2), 1girl,bangs, blue eyes, brown hair,ahoge, long sleeves, long hair, jacket, upper body, uniform, military, military uniform, cross, (serebryakov:1.2), 1girl, breasts,blue eyes, brown hair,ahoge, white shirt, long hair, boots,green pants, suspenders,black boots,cross, iron cross,
girl, solo, anime
红边黑丝，谁会不爱？觉得好看，请点红心
第二版，减少了背景和服装的污染，增强了和其他Lora模型混用的效果
1）基础模型为ChilloutMix，我喜欢用Henmixreal
2）建议高权重使用，0.5以下有可能变成黑边黑丝
3）最优触发词为red edge，建议使用(rededge:1.5)增强，若触发词不管用，请尝试rededge，rededge black pantyhose，rededge pantyhose
4）建议搭配高跟鞋、小皮鞋，训练素材中不包含光脚图片
Red edge black silk, who will not love? Think it looks good, please click the red heart
The second version, reducing the pollution of the background and clothing, enhanced and other Lora models mixed with the effect
1) the base model for ChilloutMix, I like to use Henmixreal
(2) recommend the use of high weights, 0.5 below the risk of becoming black edge black silk
(3) the optimal trigger word for red edge, it is recommended to use (rededge:1.5) to enhance, if the trigger word does not work, please try rededge, rededge black pantyhose, rededge pantyhose
(4) recommended with high heels, small leather shoes, training materials do not contain barefoot pictures
黒いシルクに赤い縁取りが嫌いな人はいないでしょう。 もし似合うと思ったら、赤いハートをクリックしてください。
バージョン2、背景や衣装の汚染を減らし、他のLoraモデルとのミキシングを強化しました
1) ベースモデルはChilloutMix、私はHenmixrealを使うのが好きです
2) 高加重を推奨、0.5以下だと黒縁黒絹になる危険性あり
3) 最適なトリガーワードはレッドエッジです、(rededge:1.5)で強化することをお勧めします、トリガーワードがうまくいかない場合は、レッドエッジ、レッドエッジ黒パンスト、レッドエッジパンストなどをお試しください。
4）ハイヒール、小さな革靴を合わせることをお勧めします、トレーニング材料は裸足の写真を含まない
빨간색 트리밍이 있는 블랙 실크를 좋아하지 않을 사람이 있을까요? 예쁘다고 생각되면 빨간 하트를 클릭해 주세요.
배경과 의상 오염이 적고 다른 로라 모델과의 믹싱이 강화된 버전 2
1) 기본 모델은 ChilloutMix, 저는 Henmixreal을 사용하는 것을 좋아합니다.
2) 높은 가중치를 권장하며 0.5 미만은 검은 가장자리 검은 실크가 될 위험이 있습니다.
3) 최적의 트리거 단어는 레드 엣지이며, (레드 엣지 : 1.5)를 사용하여 강화하는 것이 좋습니다. 트리거 단어가 작동하지 않으면 레드 엣지, 레드 엣지 블랙 팬티 스타킹, 레드 엣지 팬티 스타킹을 시도하십시오.
4) 하이힐, 작은 가죽 신발, 트레이닝 자료에는 맨발 사진이 포함되어 있지 않은 것이 좋습니다.
rededge,red edgerededge black pantyhoserededge pantyhose
I am the Reindeer Czar, and I like the game Deer Avenger. Bambo is the hero and he is very cool.
The games are from the late 90s so the image size is not HD. Despite this, I did my best to get the good training data. The outlines are a bit not crisp, but I think overall it is pretty good. It represents the cool character of Bambo.
Helping Tags:
(ripped sleeves, green open vest, chest tuft, antlers, deer ears, snout, black eyes)
Helping Negative Tag:
(closed vest)
Weight around 0.65 to 0.7 for best result, but feel free to play around.
If you like it, post some images. :) I would like to see them.
BamboCzar
The LoRA model was additionally learned from paintings of girls by painters. It can output a painting that looks like a masterpiece in a museum. It is additionally trained on ChilloutMix basis and is recommended for use with ChilloutMix. This is my first LoRA model and will be updated in the future.
画家たちによる少女の絵画から追加学習したLoRAモデルです。美術館にある名画のような絵を出力できます。ChilloutMixベースで追加学習しており、ChilloutMixでの使用をお勧めします。これは私が初めて作ったLoRAモデルで、今後アップデートします。
sls
战双诺克提模型，默认使用权重1，推荐权重0.8，目前该模型还在优化中，可以收藏一下等待后续优化更新版本，请勿商用！推荐使用主模型：
https://civitai.com/models/44036/boxmix-25dmale-enhancement
Zhanshuang Nuo Keti model, the default weight is 1, and the recommended weight is 0.8. At present, the model is still being optimized. You can save it and wait for the subsequent optimization and update version. It is recommended to use the main model:
战双帕弥什 诺克提 モデル、デフォルトの重みは 1、推奨重みは 0.8 です。現在、モデルはまだ最適化中です。保存して、その後の最適化とバージョンの更新を待つことができます。メイン モデルを使用することをお勧めします。 :
nkttattofinger gloves
A LoRA model of Sekhmet of Death from Last Origin. The trigger tag is 'sekhmet of death'. The key features are 'egyptian, egyptian clothes, dark skin, dark-skinned female, jackal ears, pointy ears, dark elf'. The 784mb VAEs (NovelAI, Anythingv3, Orangemix, Counterfeit) are recommended. 0.6 ~ 0.8 weights are good. The preview images are generated using 1.6 version (original) YesMix.
Some preview images are generated using 'Hires Fix' and upscaled using 'SD Upscale'. Thus, if you wanna generate the preview images, pls download the original preview images and analyze them using 'PNG INFO' in webui to get all generating information.
The maximum of generating resolution should not exceed 1024 theoretically, otherwise extra/deformed head/limbs will be waiting for you :). Try to use 'hires fix' to prevent this problem and help to generate higher-resolution images.
For 784mb VAEs (NovelAI, Anythingv3, Orangemix, Counterfeit) users, make sure add '--no-half-vae' in command line of webui to prevent generating black images. See voldy's log for the details of edit method. For high-performance GPU, make sure 'medvram' and 'lowvram' mode are disabled.
sekhmet of deathegyptianegyptian clothesdark skindark-skinned femaledark elfpointy earsjackal earshime cut
The LoRA model was additionally learned from paintings of girls by painters. It can output a painting that looks like a masterpiece in a museum. It is additionally trained on Anything v5.0 basis and is recommended for use with Anything v5.0. This is my first LoRA model and will be updated in the future.
画家たちによる少女の絵画から追加学習したLoRAモデルです。美術館にある名画のような絵を出力できます。Anything v5.0ベースで追加学習しており、Anything v5.0での使用をお勧めします。これは私が初めて作ったLoRAモデルで、今後アップデートします。
slsThis is not a request, it's just a tribute to one of my favorite artist. My favorite songs are One in A Billion (with Wake Up Girls) and Lion (from Macross Frontier). Which song of her you like most ?
Thank you for downloading. Please leave a rating and comment if you like my models. also leave a comment on what model would you like to see next. Please consider to leave a tip if you like my work 😌☕️ enjoy
https://ko-fi.com/meantweetanthony
want to get up to 10 epochs for your custom LoRA model requests + 50 generated images
https://ko-fi.com/meantweetanthony/commissions
Trained on 50 photos with ChillOutMix, work best on DPM++ SDE Karras sampling steps 20-30 CFG Scale 6. weight 0.8-1.0 is recommended for better results.
lightweight LoRA, works on any generator (easydiffusion,stablediffusion etc). share your results.
Disclaimer : Usage of my LoRA models is 100% under your own responsibilities. I bear no responsibilities on however the models are used and it is beyond by jurisdiction.
1girl
Form for requests.
I also sometimes do polls on my Patreon for the next LoRA (no payment required).
A weight of ~1.0 is recommended.
Feel free to experiment with other weights and share your results if you want to!
The main prompt is "shrug \(clothing\)"
Always add either "long sleeves" or "short sleeves" depending on what you want.
The LoRA really likes midriffs, so you might need to use negative prompts for some outfits.
It should work on most anime models. I don't know about realistic ones.
View the sample images for some prompting examples.
Let me know if you have any issues.
Some preview images use some of my character LoRAs. They should be auto tagged by civitai.
Used embeddings for sample pictures:
If you have any feedback or questions just write in the comments or hit me up on reddit or discord.
shrug \(clothing\), long sleevesshrug \(clothing\), short sleeves
你可以通过这个非常cooooooooooooooool的lora
生成只属于你自己的带有克味的女孩子，并带有一些艺术气息
甚至不是女孩子（bushi
一定要增加girl的权重，例如（girl:1.3）,不然你的老婆可能就变成克苏鲁了（那不是更棒吗）
建议使用权重为0.8，搭配图生图，
如果一定要用文生图模式，请千万不要勾选高分辨率修复，除非你真的很喜欢古神
建议加入触发词
With this very coooooooooooooooool lora you can
Generate your own girl with a touch of gram and a touch of art
Not even a girl (laugh)
Be sure to increase the weight of the girl, for example (girl:1.3), otherwise your wife might become a Cthulhu (wouldn't that be even better？)
It is recommended to use a weight of 0.8, with a picture-generated picture.
If you have to use the text-to-picture mode, please don't check the high-resolution fix unless you really like Cthulhu.
Suggest adding trigger words
关于新版本：
使用了金字塔噪声进行了重新训练，强烈建议搭配ControlNet的Tile使用
如果你喜欢我的工作，请给个五星好评，这是我坚持下去的最大动力
About the new version:
Retrained using pyramidal noise, highly recommended for use with ControlNet's Tile
If you like my work, please give a 5-star review, it's the biggest motivation for me to keep going
glass,plants
This is not a necessary, but If you like my Lora, please consider to donate or even Comissioning your own requested Lora on My Ko-fi
https://ko-fi.com/warrensky/commissions
Please give 5 star reviews by sharing your picture creation if you like it,
Use any checkpoint based on SD1.5
its 768,768 image
turn on Hires.fix (or not, please experiment with it yourself)
Hires steps: 10-17, denoising:04-65 ish.. just try it yourself
CGE: 4-7
Use ERSGAN 4X+Anime 6B, DPM++ 2M Karras or DPM++ SDE Karras
Use 20-50 sampling steps
Strength: 0.4-1
I suggest turn off the restore faces check mark for better results of this Lora
She might look like someone from real life, but she is not what you think. She is a totally fictional character.
It is prohibited to use this model for commercial purposes and any scenarios of illegal acts and purposes.
本人原则上不支持使用此模型生成任何成人内容。用户在使用本模型时，请务必遵守当地法律法规，勿侵犯他人的名誉权、隐私权或肖像权。
I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.
私は一般的に、このモデルを使用して成人コンテンツを生成することをサポートしません。 このモデルを使用する場合は、ユーザーは当地の法律や規制を守り、他人の名誉権、プライバシー権または肖像権を侵害しないようにしなければなりません。
This LoRa try to mimic the simple illustration style from kids book.
This is not a necessary, but If you like my Lora, please consider to donate or even Comissioning your own requested Lora on My Ko-fi
https://ko-fi.com/warrensky/commissions
Please give 5 star reviews by sharing your picture creation if you like it,
Use any checkpoint based on SD1.5
its 768,768 image
turn on Hires.fix (or not, please experiment with it yourself)
Hires steps: 10-17, denoising:04-65 ish.. just try it yourself
CGE: 4-7
Use ERSGAN 4X+Anime 6B, DPM++ 2M Karras or DPM++ SDE Karras
Use 20-50 sampling steps
Strength: 0.4-1
I suggest turn off the restore faces check mark for better results of this Lora
She might look like someone from real life, but she is not what you think. She is a totally fictional character.
It is prohibited to use this model for commercial purposes and any scenarios of illegal acts and purposes.
本人原则上不支持使用此模型生成任何成人内容。用户在使用本模型时，请务必遵守当地法律法规，勿侵犯他人的名誉权、隐私权或肖像权。
I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.
私は一般的に、このモデルを使用して成人コンテンツを生成することをサポートしません。 このモデルを使用する場合は、ユーザーは当地の法律や規制を守り、他人の名誉権、プライバシー権または肖像権を侵害しないようにしなければなりません。
📌Training model：AnythingV5V3_v5PrtRE.safetensors
📌LoRA set：0.6~0.8
📌Prompt：Roxy Migurdia, witch hat, blue eyes, blue hair, ahoge|呆毛, twin braid, twintails, brown coat, cape|斗篷, robe|長袍, braided ponytail, knee boots, witch
📢I release my models for free, if you like my works, you can support me and buy me a drink.☕☕☕https://www.buymeacoffee.com/hansom283
Roxy Migurdia
weight strength: About 0.75
Sampling steps: 15
No Trigger
Use Model: chilloutmix_NiPrunedFp32Fix
実写系の開口器です。
主に上半身で機能します。正面や少し斜めの角度で機能します。
舌が出ます。
私が別にアップロードしているNose HookのLoraとは相性があまり良くありません。
併用する場合はどちらも0.7あたりに合わせてみてください。
AI初心者が作った物なので生成に偏りや不具合があると思います。
良い画像ができたらアップロードしてくれると嬉しいです。
実在する人物の名誉を棄損する使い方はしないでください。
Is this guy heating a meal that became cold long time ago?
这货是不是在炒冷饭啊？
Anyway, with this lora you can let your girls put on a sexy maid costume with underboob and crotchless pantyhose just like Indomitable chan.
总之，用这个lora你能让你的姑娘像不挠酱一样穿着一套性感的有南半球和开档连裤袜的女仆装。
In addition, if you just want Indomitable, you can input "black hair, very long hair".
此外，如果你只想要不挠的话，可以输入"black hair, very long hair".
PS. her pantyhose's ass_cutout may not be easy to see, as AI is fond of completion.
PS. 她连裤袜的屁屁开口可能并不容易看到，AI喜欢画完。
indomitable maidmaid headdress
使用建议:
配合一些Lora, TI, VAE等使用文件
如果你觉得不错，就给我打个五星评价吧
如果您有任何问题，请通过xsarchitecturalstudio@gmail.com与我联系
Suggestions for use:
Lora, TI, VAE and other documents for use
If you think it's good, give me a five-star review
If you have any questions, please contact me at xsarchitecturalstudio@gmail.com
XS轻量模型系列在库编号100+轻量模型将有新的工作流和使用方式，大致分为主体轻量模型与辅助轻量模型（LoRa、Loha、Perfusion、TI、VAE等）
主体轻量模型：XSArchi_1***V1，V1将作为主体轻量模型编号进行规整命名，在工作流中起到主导图像生成的工作，复合功能将决定于轻量模型功能。使用权重在0.5-1.3左右。
辅助轻量模型：XSArchi_1***V1.1，V1.1（即此Lora第二版）将作为辅助轻量模型编号进行规整命名，在工作流中起到辅助图像生成的工作，无复合功能，分为视角、风格、应用场景等20余种功能轻量模型，详细功能将在轻量模型介绍中标出。使用权重在0.1-0.5左右。
V2+轻量模型将被更新或迭代。
新轻量模型将分为两种不同的训练、融合、提取等方式进行制作，在库编号相同但版本不同的轻量模型，在训练集的处理上大致无异。在库编号不同但版本相同的轻量模型，功能不同外，训练集及训练方式也各不相同。
新版本将加入中英文命名的方式，解决大家对于模型标识、分类出现的困难。
其余版本介绍：
在库编号100-的轻量模型为早期试用模型，将被迭代。
在库编号100+的轻量模型为正式模型，分为商用及非商用模型，将会持续更新。
在库编号百位数以复数形式出现的为独立授权模型，针对不同平台、企业、组织、个人等进行授权、商用、非商用流通。
在库含有编号的轻量模型被认作为垂类模型库在库模型，将配合XS工作流进行使用。
XS Lightweight model series in the library number 100+ lightweight model will have a new workflow and use methods, roughly divided into the principal lightweight model and auxiliary lightweight model (LoRa, Loha, TI, VAE, etc.).
Body lightweight model: XSArchi_1***V1, V1 will be structured and named as the body lightweight model number, and play a leading role in image generation in workflow. Composite function will be determined by the function of the lightweight model. The right to use weight is around 0.5-1.3.
Auxiliary lightweight model: XSArchi_1***V1.1, V1.1 (i.e., the second version of Lora) will be structured and named as the auxiliary lightweight model number, and play an auxiliary role in image generation in workflow. There is no compound function, and it is divided into more than 20 functional lightweight models such as perspective, style and application scene. The right to use is about 0.1 to 0.5.
V2+ lightweight model will be updated or iterated.
The new lightweight model will be divided into two different ways of training, fusion, extraction and so on. The lightweight model with the same library number but different versions will be roughly the same in the processing of the training set. In addition to the different functions of lightweight models with different library numbers but the same version, the training sets and training methods are also different.
The new version will add the way of naming in Chinese and English to solve the difficulties of model identification and classification.
Introduction to other versions:
The lightweight model in library number 100- is an early trial model that will be iterated.
The lightweight models in the library number 100+ are official models, divided into commercial and non-commercial models, and will be updated continuously.
The number of hundreds in the database appears in the plural form as an independent authorization model, which carries out authorization, commercial and non-commercial circulation for different platforms, enterprises, organizations, individuals, etc.
The numbered lightweight models in the library are considered as vertical model models and will be used in conjunction with the XS workflow.
Create characters with a good physique, but not exaggerating, actively looking for their physiques to enter the range of what is anatomically correct, for example "David".
While the bara creates characters similar to those in jojo bizarre adventures
In any case, be careful with the checkpoints you use or your characters will be very limited, try to use a good checkpoint for men (they are not necessarily the most downloaded)
You can create furry characters, anatomically correct is only limited to muscles
mepios (50) nipples (39) abs (39) short hair (36) navel (32) thighs (32) large pectorals (29) looking at viewer (29) erection (27) penis (26) thick thighs (25) testicles (22) stomach (21) uncensored (21) topless male (20) underwear (17) smile (17) bulge (16) male underwear (16) male pubic hair (16) pubic hair (15) animal ears (14) feet out of frame (13) nude (13) black hair (12) furry (12) sideburns (11) ass (11) blue eyes (11) furry male (11) dark skin (10) blonde hair (10) pants (10) blush (10) spiked hair (9) armpits (9) tail (9) veins (9) dark-skinned male (8) lying (8) underwear only (7) from behind (7) back (7) tattoo (7) brown hair (7) thick eyebrows (6) sweat (6) jewelry (6) on back (6) spread legs (6) veiny penis (6) completely nude (6) white male underwear (6) simple background (5) headband (5) navel hair (5) sitting (5) precum (5) gloves (5) clothes lift (5) bare pectorals (5) full body (5) hat (5) briefs (5) shorts (4) erection under clothes (4) white background (4) looking back (4) black eyes (4) outdoors (4) day (4) facial hair (4) mature male (4) white hair (4) arms behind head (4) green hair (4) clothes pull (4) pants pull (4) one eye closed (4) cum (4) cowboy shot (4) ass focus (4) shirt lift (4) earrings (4) open clothes (4) collar (3) wet (3) boxers (3) male swimwear (3) english text (3) mask (3) forehead protector (3) armpit hair (3) open mouth (3) sidepec (3) sweatdrop (3) masturbation (3) large penis (3) male masturbation (3) arm tattoo (3) monochrome (3) mouth hold (3) tank top (3) biceps (3) flexing (3) bodysuit (3) beard (3) jockstrap (3) tongue (3) green eyes (3) pillow (3) on bed (3) horns (3) artist name (3) sky (3) holding (3) cat ears (3) flaccid (3) see-through (2) wet clothes (2) medium hair (2) boxer briefs (2) bed sheet (2) american flag (2) beach (2) stubble (2) swim trunks (2) covered nipples (2) fangs (2) dog ears (2) dog tail (2) dog boy (2) shoes (2) necklace (2) sneakers (2) closed mouth (2) tan (2) chain (2) undressing (2) greyscale (2) harness (2) black shorts (2) lifted by self (2) black tank top (2) brown fur (2) orange fur (2) standing (2) clothing aside (2) tiger boy (2) bow (2) bowtie (2) spikes (2) furrification (2) tongue out (2) water (2) arm behind head (2) wolf boy (2) shirt (2) bed (2) japanese clothes (2) closed eyes (2) censored (2) heart (2) pointy ears (2) scar (2) black pants (2) toned (2) monster boy (2) purple eyes (2) multiple boys (2) 2boys (2) chibi (2) grey background (2) long hair (2) belt (2) cloud (2) selfie (2) chest hair (2) bottomless (2) body fur (2) petals (2) condom (2) condom wrapper (2) hood (2) v (2) foreskin (2) baseball cap (2) black male underwear (2) cat tail (2)
Please use in 0.7 strength.
Recommended prompt:
doala, mascot ,baseball cap,baseball uniform,open mouth
If you want to generate 2d illustration,(megami magazine:1.4) on prompt.
doala
Japan Airlines / JAL / 日本航空 Stewardess / Cabin Attendant Uniform
Trained on 35 photos with ChillOutMix, work best on DPM++ SDE Karras sampling steps 20-30 CFG Scale 6. weight 0.7 is recommended for better results. try to experiment with different weights that works for you.
also, it works together with all of my other LoRA models. have fun trying to mix & match
example setting :
HashimotoKanna_JP_Idol : weight 1.0
JALStewardessUniform : weight 0.7
sample prompts :
a woman in a uniform
a woman in a uniform and red scarf
a woman in a blue dress and red scarf
a woman in a black dress and red scarf
a woman in a black suit
black dress / blue dress / black suit/ neck scarf / red scarf
lightweight LoRA, works on any generator (easydiffusion,stablediffusion etc). share your results.
Thank you for downloading.please leave a rating and comment if you like my models. also leave a comment on what model would you like to see next. Please consider to leave a tip if you like my work 😌☕️ enjoy
jal uniform
-This model can be used at 1 weight and with many anime style models (can also be used with other lora like style or concepts)
-in view of the comments on my starfire model, I recommend using the lora with this keywords aside to the main one: aqua hair, aqua eyes, green jacket, hairclip, short shorts, leotard, short hair, sidelocks
Sggo
I feel like its a bit weak, maybe I'll retrain it for longer
Reed: Well first of all I want to tell you that the lora works well with certain checkpoints, (not all checkpoints they have good support for male characters) so it is normal that they end up creating jobs with problems. Try to use a checkpoint with good support and variety when it comes to male characters.The last option is directly another training method so it gives super different results, look at the examples. In any case, remember the important thing: use a good checkpoint, almost the entire community creates checkpoints for waifus. Try to choose one, not necessarily nfsw, totally the lora takes care of that, but if it has good support in male characters, layers some model for rpg gamestumos (111) pectorals (94) thighs (94) thick thighs (91) short hair (90) large pectorals (87) nipples (72) abs (68) mature male (66) facial hair (64) looking at viewer (59) penis (58) navel (57) stomach (53) erection (50) testicles (48) beard (45) bulge (45) feet out of frame (40) male underwear (38) sideburns (36) underwear (35) male pubic hair (35) smile (35) thick eyebrows (34) ass (34) blush (33) topless male (33) nude (32) navel hair (32) pubic hair (31) black hair (29) veins (27) chest hair (25) completely nude (25) shirt (22) cum (22) from behind (22) ass focus (21) brown hair (21) spread legs (20) arm hair (20) full body (19) bare pectorals (19) looking back (19) back (19) pants (18) hairy (18) stubble (18) leg hair (18) veiny penis (18) open clothes (15) flaccid (15) male masturbation (15) dark skin (14) dark-skinned male (14) masturbation (14) holding (13) ejaculation (13) projectile cum (13) blue eyes (13) underwear only (12) jewelry (12) hat (12) anus (12) biceps (12) outdoors (11) sky (11) gloves (11) armpits (11) sweat (11) sitting (10) weapon (10) scar (10) male swimwear (10) closed eyes (10) colored skin (10) jacket (9) sunglasses (9) red hair (9) furry (9) furry male (9) monster boy (9) lying (9) blonde hair (9) large penis (9) multiple boys (9) white hair (8) water (8) horns (8) swim briefs (8) goatee (8) harness (8) shorts (8) green skin (8) mustache (8) blue hair (8) headband (8) armpit hair (8) spiked hair (8) beach (7) holding weapon (7) briefs (7) foreskin (7) tail (7) open shirt (7) animal ears (7) artist name (7) presenting (7) tight (7) 2boys (7) sword (6) earrings (6) wet (6) chest harness (6) white shirt (6) flexing (6) tattoo (6) monochrome (6) greyscale (6) precum (6) bandages (6) tank top (6) long sideburns (6) thick arms (6) covered nipples (6) multicolored hair (5) open jacket (5) pectoral cleavage (5) black male underwear (5) pointy ears (5) piercing (5) orc (5) tusks (5) white male underwear (5) cape (5) cowboy shot (5) belt (5) long hair (5) pulled by self (5) one eye closed (5) jockstrap (5) collared shirt (5) yaoi (5) feet (5) teeth (5) median furrow (5) closed mouth (4) day (4) cloud (4) old man (4) socks (4) fingerless gloves (4) kneeling (4) clothes pull (4) censored (4) bottomless (4) fur trim (4) holding sword (4) nipple piercing (4) heart (4) overalls (4) alternate pectoral size (4) grey hair (4) on back (4) arms behind head (4) sharp teeth (4) soles (4) green eyes (4) white background (4) ass grab (4) patreon username (4) eyewear on head (3) sand (3) undercut (3) bound (3) brown pants (3) old (3) mask (3) shoes (3) clothing aside (3) spikes (3) penis peek (3) green hair (3) body hair (3) baseball cap (3) red cape (3) christmas (3) santa costume (3) pants pull (3) medium hair (3) seductive smile (3) forehead protector (3) nipple rings (3) after anal (3) twitter username (3) naked overalls (3) no pants (3) cropped legs (3) on bed (3)
Vi, model for a famous online photographer. This is a WIP.
I use this LORA to experiment with new techniques.
Use at your own risk.
Vi
this is my first Lora i was doing it cus i was h0rny so don't blame me if it didn't work
1girl, long hair, brown eyes, solo, purple hair
Trigger：lum
Default Outfit：orange eyes, long hair, green hair, aqua hair, multicolored hair, tiger print, strapless bikini, yellow boots
weight：0.7~0.9（Weight varies from model to model, so try different weights.）
Please support me to produce continuously. patreon
lumorange eyes, long hair, green hair, aqua hair, multicolored hair, tiger print, strapless bikini, yellow boots, yellow horns
Hi all, this is a sister project of AnimeFus - Kyo Mix v0.3, which trained with anime characters from KyoAni into a single model. On the other hand, AnimeChar aims to remake the character with higher quality using Lora based on our dataset.
In Railgun Mix, we would like to introduce our new trained Lora model: Misaka Mikoto (御坂 美琴), my favourite character in anime >///< . Let's bili bili together~~
You may buy me a drink to support my work ^v^ : https://www.buymeacoffee.com/ryalandai
My instagram for ai painting and model updates: https://www.instagram.com/rya.land.ai
Also, you welcome to give 5 stars to the model.
Please check "About this version" to know more details about the model version and included characters. The model base is Anything v4.0.
-- Please also support --
AnimeChar - Oshi Mix | Kotōbuki Minami (寿 みなみ)
AnimeChar - Hyouka Mix | Chitanda Eru (千反田 える)
AnimeChar - Oshi Mix | Arima Kana (有馬 かな)
AnimeChar - Oshi Mix | Hoshino Rubi (星野瑠美衣) [v2.0]
AnimeChar - Oshi Mix | Hoshino Ai (星野 アイ)
AnimeChar - Liz Bird Mix | Kasaki Nozomi (傘木 希美)
AnimeChar - Tamako Mix | Kitashirakawa Tamako (北白川 たまこ) & Asagiri Shiori (朝霧 史織)
misaka_mikoto
Rikka Sumeragi 皇六花 | Seika Girls Academy High School 聖華女学院公認竿おじさん
https://www.pixiv.net/users/86166786
<lora:aiueOkaStyle_v10:0.1>
https://civitai.com/models/22458/art-styleaiue-oka-style
<lora:saiminSeishidou:0.1>
https://civitai.com/models/45869/saimin-seishidou-style
<lora:odaNonStyleLORA_v2:0.2>, oda Non,
https://civitai.com/models/25758/art-styleoda-non-style
Prompt:
地点：(Night:1.7), Cyberpunk_CityView, Before Window,
动作：sitting at attention,sitting on couch,
服饰：red jacket,black pants, jewelry,collarbone, earrings, pantyhose, thighhighs,glasses,cleavage,pencil_skirt,garter_belt, garter_straps,
角色：<lora:Rikka_Sumeragi-V1:0.7>,<lora:aiueOkaStyle_v10:0.1>, <lora:saiminSeishidou:0.1>,<lora:odaNonStyleLORA_v2:0.2>, oda Non,
头发：Black_white_hair,long_hair,hair_pulled_back,purple_eyes, High ponytail,
角色鎖定：
1 girl, Manga,20yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,
眼神表情：
looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),
畫質:
official art,extremely detailed CG unity 8k wallpaper, perfect lighting,Colorful, Bright_Front_face_Lighting,
(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,
photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2)
身材:
(Beautiful,large_Breasts:1.4), (beautiful_face:1.5),(narrow_waist),
Negative prompt:
EasyNegative, bad-artist-anime, bad-hands-5, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry, out of focus, censorship, Missing vagina, Blurry faces, Blank faces, bad face, Ugly, extra ear, amputee, missing hands, missing arms, missing legs, Extra fingers, 6 fingers, Extra feet, Missing nipples, ghost, futanari, Extra legs, Extra hands, panties,pants, (painting by bad-artist-anime:0.9), (painting by bad-artist:0.9), text, error, blurry, jpeg artifacts, cropped, normal quality, artist name, (worst quality, low quality:1.4),
Steps: 27, Sampler: Euler a, CFG scale: 7,
Seed: 1058981049, Size: 512x768,
Model hash: 3e9211917c, Model: yesmix_v16Original,
Denoising strength: 0.5, Hires upscale: 2,
Hires steps: 18, Hires upscaler: Latent
Update dataset. Adjust trigger tag to 'fashi-girl'.
Update dataset. Slightly optimize body shape. Try to make the face more alluring. v5.4 version is conventional LoRA model.
Upload Lycoris version (v5.3). Remove hyper-realistic images in dataset that may cause bad results. Try to make face more natural and beautiful.
Due to the instability of Lycoris caused by sensitivity to learning rate settings, the updated model (5.2 version) is still conventional LoRA model. Use images published by some users as dataset to train the model. Fix the compatibility problem of non-NAI-based checkpoints. Now arbitrary anime model with NAI's VAE or kl-f8-anime2 VAE can also generate good results using this LoRA, theoretically. This version is a bit overfitted that will be fixed next time.
New model is trained using Lycoris method (all versions before 5.1 version are conventional LoRA model), pls install this extension to use this new model in webui. Adjust trigger tag to 'fashi-g'. If some problem occur during generating pls use old version.
Update dataset and model. Decide to upload 5-epoch version whose color saturation is a bit better than higher-epoch version after careful consideration. Adjust rank to 128 to keep perfect face. The trigger tag is unchanged. The preview images are generated using latest version of YesMix and EasyNegative embedding.
Reduce loss to ~0.11 using Lion optimizer. Btw, if you use Hires. fix, 0.55~0.6 weights of Denoising should be enough, otherwise the face may be slightly deformed. Btw, the preview images are generated using latest version of YesMix and EasyNegative embedding.
Update dataset and model. Add trigger tag 'fashi-girl' to improve generating probability. Adjust rank to 32 to balance file size and quality.
Prune tags to make the description more accurate.
Merge previous LORA model that is easy to generate nice images to slightly improve color saturation and optimize face. Adjust rank to 32 to balance file size and quality. The preview images of new version are generated using latest version of YesMix.
Adjust dataset to fix bad philtrum. Reduce training epochs to fix deformed limbs. The preview images of new version are generated using latest version of YesMix.
Update and balance dataset to remain the images of mature female that fit personal aesthetics. New version of LORA model is trained using Lion optimizer discovered by Google Brain that is purportedly better than Adam(w). The preview images of new version are generated using latest version of YesMix.
Update dataset and model. The preview images of new version are generated using latest version of YesMix.
Clean the dataset to remove the images of unattractive characters. Use offset noise to improve ability to generate lighter/darker images. The preview images are generated using latest version of YesMix.
Dataset is updated, so model is updated. Reduce the model file size to 1MB. The generating quality is unchanged.
This is an experimental LORA model trained with about 100 images of fashionable girls that fit personal aesthetics. Therefore, this model is named as "Fashion Girl". Due to the small-scale dataset that are composed of realistic/photorealistic images, some output images will remain anime style. The 784mb VAEs (NAI, Orangemix, Anything, Counterfeit) are recommended. 0.6~0.8 weights should be enough. This model will be continuously updated as the dataset is updated.
fashi-girlred lipsmature femalemakeup
The training was not very satisfactory, so we have to do a new version
1girl, 
你知道她没穿，但她看上去又穿了
我就喜欢这种反差感
tattoo这个tag有奇效
玩的开心
如果你喜欢我的工作，请给个五星，这是我坚持下去的动力！谢谢
You know she's not wearing it, but she looks like she's wearing it again
I love the contrast
The tattoo tag works wonders
Have fun
If you like what I do, please give it a 5 star, it's what keeps me going! Thank you!
bodypaint
这是一个关于立方体视角的中国式建筑的风格化Lora模型
在调用触发词后能够展现可爱的立方体建筑模型
如果再稍加点提示词，可以变成古建筑风景图
This is a stylized Lora model of a isometric view of a Chinese building
After calling the trigger word, it can show the lovely isometric architecture model
If you add a few more words, it can be turned into an ancient architectural landscape
isometric chinese style architecture
Please consider to support me by buying me some coffee so I can make a new LoRA faster. I do open commissions there too if you want to make your own custom LoRA: My BMAC page
Trigger word: twicesana
Best checkpoints/models to use:
Realistic: VisionGen, ChillOutMix, ArtEros, Snapdd00, Henrealmix
Fanart/painting style: Dreamshaper, LucidMix, SunshineMix, Majicmix
Join KPOP Diffusion Discord: https://discord.gg/NrBDMX79Da
twicesana
Actress
the training was on a model - CyberRealistic 2.0.
I'm Render on - EndlessReality
if you want to get speed, drive a sports car
if you want to drive off-road, take an SUV
if you want to hammer a nail, take a hammer
it's the same here - if you want to get results, read the recommendations and follow them
personally, I most often perform generation with parameters: DPM ++ 2M Karras, Sampling Steps 35, Denoising strength 0.35, CFG Scale 3.5-4.5
In the depths of a forgotten library, amidst ancient tomes and whispered secrets, I discovered the Instrumentum insaniae—an infernal device capable of summoning Lovecraftian nightmares. Crafted from melded metals, its intricate gears and pulsating crystalline orb exuded an eerie luminescence, hinting at the forbidden knowledge it held.
With caution, I approached the artifact noticing that each dial and lever seemed poised to unlock the gateway to madness, filling me with a chilling anticipation of the cosmic horrors it could unleash. Guided by forgotten incantations, I dared to traverse the boundaries of mortal perception.
As the Instrumentum insaniae sprang to life, its eldritch radiance cast macabre shadows, transforming ordinary objects into abominable forms. Statues twisted into tentacled monstrosities, while the architecture itself shifted in alignment with the chaotic madness of the cosmos. Through its unholy power, I became a conduit, beholding the terrifying beauty of Lovecraft's cosmic pantheon.
Yet, a whispered caution echoed within my mind, warning of the perils that accompany such forbidden knowledge. With a heavy heart, I chose to part ways with the Instrumentum insaniae, leaving it behind in the hopes that others would take up its mantle, for the delicate balance that separates our world from the abyss must be preserved, and the artifact's dormant secrets may hold the key to understanding the mysteries that lie beyond.
CthulhuTECH prompt notes and fun facts
You do NOT need to use the mythostech keyword, the lora is just that powerful, we're talking elder gods up in your GPU, some concepts will works a lot better without using it
Check out my samples and prompts (there’s like 500 something sample images, or will be once they’re uploaded minus anything the filter rejects
Yes, you can make porn with it, yes, some of it is confusingly hot
If you are upscaling, keep your denoise low, you don't want to correct away some coolthulhu, With REV 1.22 and coll remacri upscaler i use 0.32 as my baseline
Generally works fine at a strength of 1, you may want to adjust with some models
Purechaos edition is a lot more random, if you want to make a lovecraftian nightmare and don't care a lot about control, go forth, forsake the elder sign, invoke nyarlathotep… and adjust the strength of the lora or keyword at your peril. I’ve only included a few purechaos examples because you really don’t need them.
Its ok to put 'human' in the negative, helps a lot with hybrids and deepones and if you want to do some crazy mutants
Glowingrunes
glowingrunes is OP, i suggest putting it in a negative 1.0 then calling positive glowingrunes when you want them, depending on what you’re prompting a glowingrunes of 0.2 will still show over a negative 1.0 glowingrunes, especially if you’re also using cyberpunk.
So much work
Trained on just over 3300 AI generated lovecraftian images, Generated nearly 2600 image examples curated down to just over 500.
Coffee Maker included because this started as a konyconi style LORA, but it isn't, not anymore :D. Still they make excellent stuff and you should check em out.
All sample images are purely this lora on top of REV 1.22, no hidden loras, TIs, tricksy editing, everything is raw prompt output.
Known Issues
There are still a couple of weak areas that will me more cartoony, industrial equipment for example, books, and for some reason mobile phones.
Crouching people often have 3 feet or random shoes.... I don't know why, i suspect its maybe bleed from the crouching people in robes confusing the data
Bioluminscent doesn’t work well on anything, more of a general glow…. A lot of the training images were beautiful so i think maybe I need to train this as a general concept and not tie it to specific keywords
Leave a review
If you like it, please review it
For the love of Ithaqua buy me a coffee
https://ko-fi.com/theartificialanalyst
Prompt words
The best reference is this spreadsheet which has all the major prompts and illustrates how they fit into the hierarchy. You don’t need to use the whole hierarchy at all, but it may help to understand how things fit together (especially when working with humans, deep ones or artificial lifeforms)
The spreadsheet is currently a little incomplete because of an error I made but hopefully it’ll be completed in the next 24 hours… For now you can always get the complete list of every keyword ever by looking at the metadata with a text editor, or clicking the little ‘i’ symbol on the lora in additional networks (or via many extensions and tools)
Quickstart - if you just wanna jump in:
Human is the basic people keyword, I use man and woman for gender (you can kinda prompt futanari and easily prompt androgyny … whats the equivalent of futa for girl parts on dudes, thats like ezmode)
Mutant,Cyborg,Cultist,Demonic,Tattoo,Mask,Sorceror,Soldier,Manyeye,Dieselpunk,Fashion,Encountersuit,Cyberpunk,Scifi,Flamethrower
Monsters use Monster as their basic keyword and unless you want to be truly random I suggest actually putting in a type of monster (ghoul,chimera,manticore,unicorn etc)
Furry,Winged,Bioluminescent,Scaly,Octopoid,Humanoid,Dragon,cthulhu
Building creates buildings, I suggest being more specific e.g. a house, an apartment block, etc. You can get interiors but theyre a bit inconsistent except what ive specifically trained (e.g. cafe)
Industrial,Dieselpunk,Cyberpunk,Tendrils,Skyscraper,scaly
Technology is a catch-all for tech, its a bit mixed and needs work but if you prompt a particular thing (engine, server, air compressor) it’ll generally work ok.
Containment:Sphere,Hexahedral,pyramid - these make different shapes of containment for power plants, mystic thingoes or looking like cool AI hubs (adding eyes is cool)
Coffeemaker,pot,Mobilephone,Screen,Server,computer
Landscape c’mon ;)
Obelisks,Temple,Ruins,Pyramid,Statues,maesoleum make structures, you may want to negative them as necessary
Cyberpunk - particularly for urban or structures, dieselpunk will work a bit but it’s not trained
Jungle,Desert,bushland,Urban,Volcanic,hotspring,lava,Arctic,Underwater,abyss,forest,Park,mountain,Swamp,Hills,beach,fleshland
Hovercraft you can try some modifiers like large, small, transport but it’s inconsistent.
Cyberpunk, dieselpunk
Animal - same as monster, best to prompt a specific animal but you don’t have to.
Aquatic,Bioluminescent,Insect,serpent,
Balloon - makes zeppelins, can’t prompt much else
Cyberpunk,dieselpunk
Artifact - makes statues and small objects and things to put them in, definitely use some of the other prompts to guide it.
Statue,Obelisk,Relic,Curiosity,glowingrunes,Container,Statuette,token
Artificial - robots n such
Woman,Man,Cyborg,Android,Cyborg,Robot,Cyberpunk
Mech - Makes mechs or all types depending on how you prompt it, quads and crawlers are significantly harder to prompt
bipedal,Quad,Crawler,Biomechanical,Behemoth,Dieselpunk,cyberpunk
Deepone - lovecrafts fish-people. Seriously, check my prompts, these are hard.
Hybrid,Woman,cyberfish,Abomination,Man,pureblood,mutant,Lanternfish,goldfish,Withspawn,Angler,Bioluminescent,shark,cyborg,Encountersuit
Industrial - machinery n junk
Machine,Mining,Generator,Vehicle
Portal - y’know, portals
Doorway,rift,Primitive,Square,Circle,Hex,watcher,Tri,Scifi
Vehicle - from sports cars, to sports cars with guns, to tanks
Battlecar,Warmachine,battletank,Armed,Spikes,Spines,legs,cyberpunk,Closecombat,dieselpunk
Spaceship - sigh you understand. Can be a bit inconsistent across types of ship, prone to white background
Leviathan,Naval,Ufo,angular,Cyberpunk,Dieselpunk
Aircraft - … suggest fighter, modern fighter, bomber, etc
Antique,Cyberpunk,dieselpunk,drone
Book - On its own makes pretty cool books, doesn’t play nice with making different kinds of books.
Future plans
Assuming people like it enough, review it, etc. I will release future versions will will include some or all of the following (if people dont like it, i'll just keep it to myself :D ) I have currently put approximately a full week of work days into this project, that's a lot of time. Making some of these changes will take considerably MORE time.
Separate the major keywords into full concepts to remove some of the mixing (e.g. deepone would become a keyword not just a prompt) This is a major project in its own as i'd need to move a lot of things around (e.g. compiling all the dieselpunk and cyberpunk stuff into a keyword where it's currently in a sub sub folders
Make separate Loras for those keywords
Even if i don't do full concepts I will be totally reorganising the data as the way I have done it is super flawed and makes you need to prompt/negative hard if you want a very specific result
Add more training data to some weak areas (such as industrial equipment and adding more types of books to books, aircraft, major deepone overhaul,etc,etc)
Make a 'lite' version that is more about style than being comprehensive
Make a 'cute' version
Train a model (people would need to Looooooove this for me to train a model as it takes forever, at least I can run dreambooth on runpod even though that costs a few bucks)
I have a YouTube channel where I neglect to review Loras, checkpoints etc.
https://www.youtube.com/channel/UCjXlRo9kik1YJYONdjz4kiw
But i do plan to start doing it again now this ridiculous project is released.
p.s. If you 'cant get my results with my prompts that's a you problem and I can't help you.
p.p.s Cthulhu loves you anyway, dream sweetly of it.
cthulhutech
Spicy, Salty & Sweet
심장을 파고들어 넌
=============
I'm been thinking of charging you guys.
And the price is a like and a 5-star review. And Leave the masterpieces that you generate.
:)
=============
Weight: 1
Denoising strength: 0.3
Model: chilloutmix, Perfect World, GuoFeng3, etc.
Resolution: 512x768, 512x1024,512x512, etc.
Lora trained on the art of Stephen Mckay
stephenmckay
First of a series of DR brand cars LoRA models.
All training suggestions are welcome.
Trained on a 38 images dataset and with 3800 reg images, model SD 2.1 768
drc city car
Nadja from "Tomorrow's Nadja/Ashita no Nadja"
Trained on 60 images.
Her and the Dandelion Troupe travel the world performing acts. Her talent is dancing and being adorable.
Original Manga Artist: Yui Ayumi
Studio: Toei Animation
Weight: 0.85 and lower.
Might need heavier prompt guidance for some of her outfits.
Original: Porkpie hat, red hat bow, blonde hair, long hair, blue eyes, dress, red dress, white apron, frilled apron, long sleeves, brooch, heart brooch, bowtie, ribbon, argyle, frills, ribbon, umbrella, parasol, socks, shoes, bow
Ballerina: Tutu, blue tutu, short sleeves, waist bow, heart brooch, jewelry, bracelet, earrings, short hair, single hair bun, blue hair bow, ribbon, closed eyes, ballet shoes
Flamenco Dancer: Red dress, bare shoulders, backless dress, backless outfit, black choker, brooch, hair ornament, short hair, single hair bun, frills, black trim/black skirt trim (last one is optional), red shoes, red high heels, rose, red rose, flower
Princess: Pink dress, frilled dress, layered dress, frills, frilled collar, pink bow, long hair, hair bow, white gloves, elbow gloves, jewelry, earrings, high heels, pink heels, pink rose, ballroom
Italy: Orange hairband, long hair, shirt, white shirt, frilled cuffs, long sleeves, puffy sleeves, vertical stripes, vertical striped pantyhose, striped, pants, brown footwear, boots, Italy, Italian flag, school uniform (I would change the last prompt here to something like shirt, vest, Italian vest, shoulder pads, orange tunic)
Switzerland: Dress, white apron, puffy short sleeves, frills, braid, twin braids, hair rings, curtsey, shoes, (optional: blue skirt, red skirt trim, black clogs, dirndl).
Negative Prompts
White background/Black background/Simple background
Cropped/Border
Porkpie Hat
Blurry
Chibi
Parasol
Umbrella
Any color that isn't blonde (if you want more character accuracy)
Argyle
Holding
Headwear
Anime coloring
Multiple girls/2girls/3girls
Parody/Style parody/Official Style
Monochrome/Greyscale
nadja
I tried to get this result without any LoRA but it seems almost impossible, this helps a lot even if the results are not always crystalline and flawless... Better than nothing :D
twistcolu 
泼墨是中国画的一种画法，将墨挥洒在纸或绢上。墨如泼出，画面气势奔放。
"Pomo" is a Chinese word that refers to a painting technique in traditional Chinese painting, in which ink is splashed or splattered onto paper or silk. The ink creates a dynamic and expressive effect, capturing the essence of the subject in a bold and powerful way.
使用时请尽量使用较低的CFG以达到黑白画面的效果
Please use a lower CFG to achieve a black and white image effect
ink splash
このLoRAモデルを使用して画像を生成し、インターネット上で公開する際に、他の人が簡単に見つけられるように、原作者とモデルのウェブサイトを明記していただけると、非常に感謝いたします
当您使用此LoRA模型生成图片并发布在网上，希望可以注明一下该模型的原网址和作者方便他人查找，谢谢
When you use this LoRA model to generate images and publish them online, I would greatly appreciate it if you could indicate the original website and author of the model to make it easier for others to find.
-
LoRA of Hassan(Serenity)
静谧哈桑的lora
-
works around 0.7-0.9, support 4 sets of official outfit, and 2 sets of cosplay
推荐权重0.7-0.9 支持4套官方外观和2套变装
-
Default默认=Default, SerenityOutfit
Prototype Outfit 苍银 便服= Prototype, OriginalDress
Poisonous Flower Blooming at Night 灵衣-毒之花=PFBN, ExtraOutfit, see-through
Merry Sheep欢悦绵羊=MerrySheep, XmasCostume
dangerous beast危险野兽=DBcos, BeastCosplay
royal icing皇家糖霜=RoyalIce, LancerAlterClothing, sleeveless, dark-skin female
-
Feel free to post any idea or character model request in the Discussion
有模型的需求可以B站私信或者在这里留言给我
id：不吃青蛙的阿龙
有空就会做
If you like my work, feel free to support me on patreon.com/RiuKiMK1
or RiuKi_MK1 is Creating AI artworks, LoRa models (buymeacoffee.com)
This LoRA can work well with ChillOutMix and NeverEndingDream (because I already test)
Steps: 30
Sampler: Euler a
CFG scale: 8
Normal using at 1 but 0.8 is good enough
This is a LoRa with the style of じゅん
The base model is Suzumehachi.
If you are the artist whose style is conveyed in part or fully by this LoRa, I suggest that you introduce this tool into your workflow. That way you can please your fans with even more art.
jun
Please consider to support me by buying me some coffee so I can make a new LoRA faster. I do open commissions there too if you want to make your own custom LoRA: My BMAC page
Trigger word: twicemina
Best checkpoints/models to use:
Realistic: VisionGen, ChillOutMix, ArtEros, Snapdd00, Henrealmix
Fanart/painting style: Dreamshaper, LucidMix, SunshineMix, Majicmix
Join KPOP Diffusion Discord: https://discord.gg/NrBDMX79Da
twicemina
Jun Gotou (old one, it's not so bad)
Sakura Toriumi
Yuzuha Aigae
Masayoshi Sawatari
crossed out - planned, if the character you need from this anime is not in the list, write in comment please.
//why i change format:
it's bad if your LoRa's don't sort in automatic1111 lora list by title(title need to be first);
it's bad if your don't see trigger word in name of model(and civitai crop's long names, but for version it have additional characters);
it's not so good to made 20 LoRa's for each title on Civitai, just look like spam.
multicharacter LoRas harder to control->it's bad.
// result is format <title name>_<character trigger word> // model can contain additional TW, like "Pajamas", but this is not necessary//
LoRa trained on this model https://civitai.com/models/49608/randomizer89merge
ogi sayuri
Model: chilloutmix_NiPrunedFp32Fix
Res: 512 x 832
DO NOT check "Restore faces"
The first version has time to optimize later
Every model available tries to mimic the style of Leonardo, but the training images weren't accurately chosen or tagged, and usually the result is very inaccurate. So I tried to fix that with this LoRA.
wjqleonardo
karin,blond hair,twintails,drill hair,blue eyes,hair ornament,puple short dress,puple miniskirt, sleeveless,wide sleeves,white legwear,boots,
This LoRa is trained based on anything 4.5. If you want to join a community of AI drawing, and to learn more about LoRa training tricks, you can scan the QR code below to join our WeCaht group chat.
这个LoRa模型是基于anything 4.5 训练的。 如果你想加入ai绘画团体，并且对LoRa训练小技巧感兴趣，你可以扫描下方的二维码加入我们的“C站炼丹术分享讨论群”，欢迎大家加入并且交流讨论！！！
yuli \(spy x family)\
If you like my work then Consider supporting me on Patreon
Also my new merge model is available https://civitai.com/models/53761/goofball-mix
Lora weight of 0.7-1
highres fix or img2img at higher resolution is highly recommended.
drop tour review. it is much appreciated
maid bikinifrills
崩坏3 q版画风模型。数据集来源自之前的爱丽的一次活动。很不错的画风 所以用来练了一个lora。
Honkai Impact 3 Q version art style model. The data set comes from a previous event of
Elysia. It’s a very good art style, so I practiced Lora with it."
The recommended value is 0.8-1
ver:1.1介绍：
使用金字塔噪声技术优化后的版本。效果有一定改善。
建议取值0.8-1。
同时更加建议搭配LoRA Block Weight (LoRA 区块权重配置)插件来使用。有插件情况下的话。
ALL权重的效果很不错。可以参考预览图。
Optimized version using pyramid noise technology. The effect has been somewhat improved.
Suggest a value range of 0.8-1.
It is also highly recommended to use the LoRA Block Weight plugin. The effect of the ALL weight is very good. You can refer to the preview
chibi
Please consider to support me by buying me some coffee so I can make a new LoRA faster. I do open commissions there too if you want to make your own custom LoRA: My BMAC page
Trigger word: leejieun
Best checkpoints/models to use:
Realistic: VisionGen, ChillOutMix, ArtEros, Snapdd00, Henrealmix
Fanart/painting style: Dreamshaper, LucidMix, SunshineMix, Majicmix
Join KPOP Diffusion Discord: https://discord.gg/NrBDMX79Da
leejieun
甜甜的lolita小裙子
希望您能尝试并分享生成的图片和反馈
A sweet lolita fashion dress
Hope you can try it and share the image you generated. Eager for your feedback~ Thanks~
Recommend config:
LORA weight for txt2img: reccomended 0.7
Keys: sweet_lolita
sweet_lolita
If you enjoy my models and want to help support my releases, I always need a coffee :)
I've decided to release this to gauge interest in future releases of styles and artists I've been working on. Please don't rip this off as your own, or do anything crazy with it. I am not releasing it for professional or commercial use, this is just for personal use and non-commercial research purposes. Please respect Yuki and his legacy, RIP. And please share your results!
Yukihiro Takahashi (高橋 幸宏 Takahashi Yukihiro, June 6, 1952 – January 11, 2023) was a Japanese musician, singer, record producer, and actor, who was best known internationally as the drummer and lead vocalist of Yellow Magic Orchestra.
None of the YMO members come out correctly in the base model. I'll definitely be doing all 3 of them as LoRAS.
Use "yukihiro takahashi" as your keyword
This LoRA stacks well on almost all models I've tried, from photorealistic to anime
Works best between 0.7-1.0
Please reference my posted generations for prompts, none have been retouched or inpainted, but I ran most of them through ControlNet Tile + Tiled Diffusion, which corrects his face almost perfectly
SD wants him to have green/hazel eyes, counteract it with "brown eyes" in prompt or "green eyes" in negative prompt
yukihiro takahashi
(It's been a long-term exploration to enhance similarity and so on.)
Suggested weighted around 0.55. Notice "About this version" info.💁♀️
Refer to sample images and notice their unique negative prompts and sampling method of each checkpoint. e.g. It differs between KanPiro Mix and HenmixReal.
(Welcome to feedback, review & give me more stars and red hearts!)~~
Be kind to these lovely.
It is prohibited to use this model for commercial purposes and any scenarios of illegal acts and purposes.
Ver2.0
=====================================
这是一个中世纪风格重型铠甲套装的Lora
推荐权重：0.8
希望在这里能看到大家分享制作的盔甲图
This is a medieval-style heavy armor suit for Lora.
Recommended weight: 0.8
I hope to see everyone share armor designs here.
medieval armor
TRIGGER WORD FULLY OPTIONAL ITS : Horridthreed
personally i like it better without
well well well... im back with another LoRA yet again and this time a realism / 3d booster!!! inspired by someone on reddit looking for this particular style... this is a 2 in one since ill be uploading a custom merge i made with this lora ^^
(generation advice below) 👇
was born on July 14, 1998, in Seattle, Washington. Growing up, Charlotte was an outgoing and adventurous girl who enjoyed exploring the outdoors and engaging in various sports and activities. In her early adult years, Charlotte decided to pursue a career in the adult entertainment industry. She began her career in 2017, and her natural beauty, captivating charm, and authentic performances quickly garnered her a dedicated following of fans.
During her career, Charlotte received significant recognition for her work, immediately making an impact with her unique style and genuine on-screen presence. Charlotte had the opportunity to collaborate with well-known industry professionals and other popular performers, further solidifying her reputation as a talented and captivating adult model.
Outside of her work in the adult entertainment industry, Charlotte has a wide variety of hobbies and interests. She is a passionate traveler, often sharing her experiences and adventures with her fans through social media. Her love for the outdoors has never waned; she continues to enjoy hiking, camping, and exploring nature. Additionally, Charlotte is an animal lover and has been known to volunteer at local animal shelters in her spare time.
She is currently exploring opportunities in mainstream modeling, acting, and photography, leveraging her experience and natural talent in front of the camera.
Most images have been generated with LORA weighted @ 0.7 - 1.
Most samples done with CyberRealistic model.
Sometimes I just send it to Ultimate SD upscale after a quick render and run 2x at a scale of two.
Ultimate SD upscale script used @ ~.15 - .4 denoising using 8x NMKD Faces or 4X UltraSharp redraw upscalers.
The prompts continue to evolve. CyberRealistic and EdgeOfRealism are reasonably flexible models.
Leave a review if you liked using this.
Leave a comment if you are having issues getting good output, I'll try to respond.
CharlotteCarmen
未完成品、です
erebus_chan
权重0.6~1,低权重可能会导致黑影无法覆盖人身,但高权重可能会导致画面变得糟糕.目前来看就算是高权重也没办法稳定的将黑影覆盖到人身上,不知道是不是训练集不够的原因.似乎使用人群相关的tag人身被黑影覆盖的概率会提高.
触发词:
Weirdcore,
如果想要画面出现眼睛可以尝试使用:black eyes, 想要画面出现文字可以尝试使用:english text,
你也可以跟我的另一个lora"strange_head"一起组合尝试出不同"有趣"的画面.( •̀ ω •́ )y
Weirdcore,
📌Training model：AnythingV5V3_v5PrtRE.safetensors
📌LoRA set：0.8
📌Prompt：Nagisa Kubo, purple hair, purple eyes, black sailor collar, school uniform, serafuku, white shirt, red neckerchief, black pleated skirt, (brown cardigan:1.1), open cardigan, long sleeves
📢I release my models for free, if you like my works, you can support me and buy me a drink.☕☕☕https://www.buymeacoffee.com/hansom283
Nagisa Kubo
This is a LoRA trained with images of Izuna Ajishiro from the anime Dead Mount Death Play.
LoRA also contains multiple versions of the same character with different clothes and positions, so you only need to recall his or her name to already have a few versions.
If you are interested in any of my other LoRA
To use the LoRa properly you will need NED.
Triggers with izuna, ((izuna)), 1girl
It also works well with weight 1 however for I always recommend a weight of 0.8 or 0.9
NED: Never Ending Dream (v1.0).
How to use LoRA's in auto1111:
Update webui (use git pull like here or redownload it)
Copy the file to stable-diffusion-webui/models/lora
Select your LoRA like in this video
Make sure to change the weight (by default it's :1 which is usually too high)
izuna1girl
未完成品，unfinished product （too few pictures） (too little training data)
terror_chan
his LORA mode is for Buddhists or those who are interested in Buddha statues, please do not use it in other pornographic fields, thank you。And please use it with respect. The resulting Buddha statue should try not to have incomplete limbs, or only the head of the Buddha remains. Show your whole body as much as possible. Be as dignified as possible. Amitabha.
這個LORA 模式是為了給佛教徒或是相關對於佛像有興趣的人使用，請勿用於其他色情的場域，並且請以恭敬的心來使用。產生的佛像，盡量不要出現有肢體不全，或是只剩下佛頭。盡量顯示全身。盡可能保持莊嚴。阿彌陀佛。感謝您
建议使用权重为 0.7 左右
目前使用效果比较好的大模型有 ReV Animated
建议生成了草图之后，使用图生图功能，会有较好的效果
可选择搭配lora SXZ D.U.C.K. for Game Assets 使用，权重建议0.3左右
Suggest using a weight of around 0.7
Currently, the most effective large models used include ReV Animated
It is recommended to use the image generation function after generating a sketch, which will have better results
You can choose to use it with Lora SXZ D.U.C.K. for Game Assets , with a recommended weight of around 0.3
hason_chest_v1.0
Yuuko Kiryuu 桐生夕子 | RapeLay 電車之狼 レイプレイ
https://www.pixiv.net/users/86166786
Prompt:
地点：(Night:1.7), Cyberpunk_CityView, Before Window,
动作：sitting at attention,sitting on couch,
服饰：green shirt,sleeveless, bracelet,beige skirt,tank_top,
角色：<lora:Yuuko_Kiryuu_RapeLay-V1:0.8>,<lora:aiueOkaStyle_v10:0.1>, <lora:saiminSeishidou:0.1>,
头发：Brown_short_hair,Bangs,Brown_Eyes,
角色鎖定：
1 girl, Manga,25yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,
眼神表情：
looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),
畫質:
official art,extremely detailed CG unity 8k wallpaper, perfect lighting,Colorful, Bright_Front_face_Lighting,
(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,
photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2)
身材:
(Beautiful,large_Breasts:1.1), (beautiful_face:1.5),(narrow_waist),
Negative prompt:
EasyNegative, bad-artist-anime, bad-hands-5, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry, out of focus, censorship, Missing vagina, Blurry faces, Blank faces, bad face, Ugly, extra ear, amputee, missing hands, missing arms, missing legs, Extra fingers, 6 fingers, Extra feet, Missing nipples, ghost, futanari, Extra legs, Extra hands, panties,pants, (painting by bad-artist-anime:0.9), (painting by bad-artist:0.9), text, error, blurry, jpeg artifacts, cropped, normal quality, artist name, (worst quality, low quality:1.4),
Steps: 27, Sampler: Euler a, CFG scale: 7,
Seed: 2082525009, Size: 512x768,
Model hash: 3e9211917c, Model: yesmix_v16Original,
Denoising strength: 0.5, Hires upscale: 2,
Hires steps: 18,
Hires upscaler: R-ESRGAN 4x+ Anime6B
A LoRA intended to make better breast expansion images.
Activate with the token "BEPlus", and include whatever other tags you normally use for BE
BEPlus
Himemushi Momoyo from touhou 18 Unconnected Marketeers.
※ Trigger words:
Himemushi Momoyo,
patterned clothing,
black shirt,
green ring,
orange ribbon,
orange bow,
black skirt
※ I also recommend adding holding pickaxe, holding shovel to a negative, to prevent yellow objects from appearing unexpectedly.
It can also be used as a trigger, but won't work that well.
※ Recommended weight: 0.8~0.9, CFG Scale: 5~8
(Not always the best, better try adjusting from this and find the best condition for your model.)
※ Clip skip: 2
※ Trained with animefull
Himemushi Momoyopatterned clothingblack shirtgreen ringorange ribbonorange bowblack skirt
This LoRA tries to make ASCII art which is practically impossible with SD but the results turned out pretty interesting, so thought of sharing it. It is subjective if you will like it or not.
This LoRA is trained on RevAnimated_v122, so try to use the same model.
The base txt2img image might be blurry and have less details, I recommend using high res fix if you can.
If you can't use high res fix, then send the base image to img2img tab and then upscale the resolution using a low denoising strength (0.3-0.45), this may do the same job as high res fix.
Then finally you can upscale the image to have even more texture. I recommend using 4x-UltraSharp or 4x_foolhardy_Remacri but you can also experiment if you feel like so :)
You can also experiment using restore faces wherever possible.
Instance Prompt: ascii_art
Recommended weights: 0.7, 0.8 (more recommended), 0.9
Add "white background" in prompt if you want black texture and white background
Add "black background" in prompt if you want white texture and black background
If you want colorful results, don't add white background or black background in prompt and use 0.8 LoRA weight.
If you want to convert a normal image with img2img, use 0.6-0.7 denoising strength, it can give you mixed results though, so don't expect too much. Also use higher weights.
Adding "pink theme", "blue theme" etc in prompt makes it look interesting.
Try avoiding "portrait" in your prompt if you don't want any humans.
ascii_art
An Exo Stranger Lora model trained on a little over 50 images. I've found it works pretty well with Dreamlike-Photoreal, RevAnimated, and Lyriel, however I'm sure other models will work just as well if not better. Like most models it takes a lot of trial and error to get crisp images with little to no errors, but I've only messed with it for a few hours.
I figure I should also mention that since there aren't a ton of quality images depicting her in different poses and clothes, so the image selection is pretty tight. Not going to lie a good number of the training images are thanks to StrangerSFM's work, https://twitter.com/strangersfm (NSFW!). Therefore, I wouldn't be surprised if a distorted watermark vaguely looking like that name shows up in some of the images.
It would also be good to know that I referred to her as 'Exo Stranger' first in the training image captions before her actual name. So if you name her in your prompt, I'd use Exo Stranger. Anyway, thanks for giving it a look and I hope it works good for you.
Exo StrangerElsie BrayElizabeth Bray
(It's been a long-term exploration to enhance similarity and so on.)
Training images are old inevitably. They are processed by Topaz to get higher definition and rid of mosaic. Yet then LoRA training steps must be restrained in case of blur after hires.fix.
~~
Both version suggested.
Suggested weighted around 0.66 on 020. Notice "About this version" info.💁♀️
Refer to sample images and notice their unique negative prompts and sampling method of each checkpoint.
(Welcome to feedback, review & give me more stars and red hearts!)~~
An actress active in 1980~1990s.
Be kind to these lovely.
It is prohibited to use this model for commercial purposes and any scenarios of illegal acts and purposes.
In my defense, someone had to do it.
This is a LoRA trained on Minecraft 3D NSFW art. It can be difficult to get good results, as the style is so different from the base model.
Trained on 386 images for 3597 steps, on AOM3A1.
mcnsfw
captain planetcolored skingreen hair
You can download my LoRAs that haven't been released here a few days earlier.
Minji
This LoRa is trained based on anything 4.5. If you want to join a community of AI drawing, and to learn more about LoRa training tricks, you can scan the QR code below to join our WeCaht group chat.
这个LoRa模型是基于anything 4.5 训练的。 如果你想加入ai绘画团体，并且对LoRa训练小技巧感兴趣，你可以扫描下方的二维码加入我们的“C站炼丹术分享讨论群”，欢迎大家加入并且交流讨论！！！
loid \(spy x family)\
Luna ルナ 露娜 | epic7 エピックセブン 第七史詩
https://www.pixiv.net/users/86166786
Prompt:
地点：(day:1.7),cloud, cloudy_sky, ocean, outdoors,a picture in front of a sunset background,
动作：standing in front of a blue sky with clouds,
服饰：purple_white_demon_outfit,shoulder_armor,armored boots, armor,tail_armor,
角色：<lora:Luna_epic7-V1:0.7>,elf,pointy_ears,purple belt,covered_navel,futuristic outfit,
头发：silver_white_hair,very_long_hair,horns,red_eyes,thighhighs,thighhighs,
角色鎖定：
1 girl, Manga,20yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,
眼神表情：
looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),
畫質:
official art,extremely detailed CG unity 8k wallpaper,
(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,
photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2)
身材:
(Beautiful,large_Breasts:1.4), (beautiful_face:1.5),(narrow_waist),
Negative prompt: EasyNegative, bad-artist-anime, bad-hands-5, lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, jpeg artifacts, signature, watermark, username, blurry, out of focus, censorship, Missing vagina, Blurry faces, Blank faces, bad face, Ugly, extra ear, amputee, missing hands, missing arms, missing legs, Extra fingers, 6 fingers, Extra feet, Missing nipples, ghost, futanari, Extra legs, Extra hands, panties,pants, (painting by bad-artist-anime:0.9), (painting by bad-artist:0.9), text, error, blurry, jpeg artifacts, cropped, normal quality, artist name, (worst quality, low quality:1.4),
Steps: 27, Sampler: Euler a, CFG scale: 7,
Seed: 3585642292, Size: 512x768,
Model hash: 3e9211917c, Model: yesmix_v16Original,
Denoising strength: 0.5, Hires upscale: 2,
Hires steps: 18, Hires upscaler: Latent
From a Reddit request, trained on photos of Egirl Missparaskeva/yourmommypasha
LORA was trained with publicly available photos.
missparaskeva
我らが主人公、ランス君
rance,blown hair,short hair,green armar,
目的：更好的增加触手play的概念
效果：权重1生效，可能需要的tag：tentacles,solo, breasts,habit, open mouth,clothing aside, tentacle sex, pussy,vaginal,spread legs,cum,
缺点：因为训练集的原因，对蓝色头发，宝石发饰拟合严重，如果需要蓝色以外颜色的头发请提高颜色权重
是否更新：会继续更新（时间未定）
PS：ღ( ´･ᴗ･` )比心
Objective: To better increase the tentacle play concept
Effects: Weight 1 takes effect, may need tag: tentacles,solo, breasts,habit, open mouth,clothing aside, tentacle sex, pussy,vaginal,spread legs,cum,
Cons: Because of the training set, sapphire hair accessories fit badly for blue hair, please increase the color weight if you need hair colors other than blue
Update or not: Will continue to update (time undetermined)
tentaclesbreastshabitopen mouthclothing asidetentacle sex
效果与V1.0有很大差异
根据我的测试：
CFG Scale: 4~7
lora权重:0.35~0.7(推荐0.4~0.6)
一定要用高分重绘生成图片，参考设置如下：
分辨率: 512*768
重绘幅度: 0.45
放大算法: SwinIR 4x
放大倍数：1.8
采样算法应该都行
当你给Lora一个较高的权重时，建议将户外环境prompt（例如outdoors、cityscape等）同步调高（高于1）
我还上传了一个欧美真人模型，如果喜欢可以下载—Euro-AmericanDollLikeness
V2.0 is almost totally different from V1.0
This is a lora which can generate beautiful chinese girls/women photos
According to my test:
CFG Scale: 4~7
lora's strenth: 0.35~0.7 (0.4~0.6 is the best I think)
width and height: 512*768
Denoising strength: 0.45
Upscaler: SwinIR 4x
Upscale by 1.8
Recommend setting the outdoor environment prompt (e.g. outdoors, cityscape, etc.) higher as well (above 1) when you give the lora a high strenth.
Welcome to download another "likeness" model of mine—Euro-AmericanDollLikeness
生成した画像を投稿したり、プロンプトについて話し合ったり、助けを求めたりできる Discord サーバーを作成しました。https://discord.gg/TJrR2T6uzt
anime
阳光明媚的一天，初音未来妹妹叠了最厚的甲找宋公明哥哥打野球拳，妹妹屡战屡败衣服越来越少，但是无论如何公明哥哥也到达不了这近在咫尺却遥不可及的本垒……因为这只是个正经的脱衣游戏。
—— 魔法哥哥 宋公明 制作
On a sunny day, Hatsune Miku put on her thickest armor to play rock-paper-scissors with Song Gongming, but Miku lost repeatedly and ended up with fewer and fewer clothes on. However, no matter how hard Gongming tried, he couldn't reach the home base that was close at hand but out of reach... because it was just a regular strip game.
—— By Magical Brother, Gongming Song
晴れた日、初音ミクは最も厚い鎧を身につけ、宋公明と野球拳をすることになったが、ミクは何度も負けて、だんだん服を脱いでいった。しかし、公明もがんばっても、手の届かない目の前のホームベースには到達できなかった...なぜなら、それはただの脱衣ゲームだったからです。
—— 魔法の兄、宋公明によって製作されました
화창한 날, 하츠네 미쿠는 가장 두꺼운 갑옷을 입고 송공명 형님과 야구권을 하게 되었지만, 미쿠는 계속 패배하여 점점 옷이 적어졌다. 그러나 공명 형님도 열심히 노력해도, 손길이 닿지 않는 바로 앞의 홈베이스에는 도달할 수 없었다... 왜냐하면 그것은 그냥 정말로 재미있는 적당한 스트립 게임이었기 때문이다.
—— 마법의 형, 송공명이 제작했습니다
This LoRa is trained based on anything 4.5. If you want to join a community of AI drawing, and to learn more about LoRa training tricks, you can scan the QR code below to join our WeCaht group chat.
这个LoRa模型是基于anything 4.5 训练的。 如果你想加入ai绘画团体，并且对LoRa训练小技巧感兴趣，你可以扫描下方的二维码加入我们的“C站炼丹术分享讨论群”，欢迎大家加入并且交流讨论！！！
lixiaoyao \(swrod and fairy\)
Tayuya from Naruto LoRA
Prompts for casual: tayuyacasual, casual, hat
Prompts for curse: tayuyacurse, curse, dark skin
Weight 0.7 - 0.9 for best generation.
tayuyatayuyacasualcasual, hattayuyacursecurse, dark skin flute
是的，我开启了新的姿势系列：No.5xx
Yes, I started a new pose series: No.5xx
请在any5或novel ai或其他动漫模型中使用No.505。
anyv5出图，强度0.5-0.7，注意负面tag我使用了EasyNegative：
EasyNegative - EasyNegative | Stable Diffusion Textual Inversion | Civitai
AnyV5 output image, intensity 0.5-0.7, note that my negative label uses EasyNegative:
EasyNegative - EasyNegative | Stable Diffusion Textual Inversion | Civitai
它在 640x960 或 512x768 尺寸下效果最佳。
It works best in 640x960 or 512x768 dimensions.
系列说明：
No.1xx 是我不太熟悉的人
No.2xx 是我知道的明星或博主
No.4xx 是一些实验性风格概念等。
Series description:
No.1xx is someone I am not familiar with
No.2xx is a star or blogger I know
No.4xx are some experimental style concepts etc.
我是dawn6666，我之前的账号被封禁，今后的内容在这个账号或者我的频道中发布（https://t.me/dawn66666666）。
I am dawn6666, my previous account was banned, and future content will be posted on this account or my channel（https://t.me/dawn66666666）.
fallen down
给我的小喵太做个lora，素材都是买的盒子截图的XD。
想要画正常的选择lite版，想要无脑立绘就选原版！
目前这个lora还是有一些问题包括：
背景比较容易变白，有点过拟合——暂时解决方案就是多写点细节tag，比如衣服，头发颜色、类型之类的。
和部分模型不太契合：推荐anything and everthing、pastelboy
但是意外的服装的融合方式还挺可以，花纹也挺精细。
有时间会多截点图丰富下衣服，不定时更新
没想到吧，这么快就更新了：
我用自己另一个风格化lora丰富了背景的同时，减小了dim。
现在需要触发词1boy，并且适当提高权重，不然衣服会偏向底模。
(1boy:1.5)jw3
It is probably a LoRA with a mixed photo style of beautiful women, and welfare images were selected from the high-definition set of more eye-catching photo models for training.If my other 'Pure Innocent girl' LoRA is a cute girl, then this' Oriental Beauty'LoRA can be considered a lady of temperament「お姉さん/御姉」.Perhaps because this is a mixture of multiple faces, the unique features are not very obvious. The weight ranges from 0.3 to 0.8, suitable for semi realistic and realistic effects.It is recommended to use 0.5 alone. If there are abnormalities when using it with other models, Lora, or some samplers, it is recommended to randomly adjust the weight, with a maximum of 0.9.
It can be used to generate nsfw, but it is prohibited to be used for commercial activities.If you like this LoRA, please like it 💛 Or five-star ☆ evaluation ⭐ Supported!
大概是一個美女混合的寫真風格的LoRA，從比較養眼的寫真糢特的高清套圖中篩選了福利圖用來訓練。如果說我的另一個“純真女孩”LoRA是可愛少女，那麼這個“東方佳麗”LoRA算是氣質禦姐。可能因為這是多張人臉的混合體，獨有的特徵並不太明顯。權重在0.3～0.8之間，適合半寫實和逼真效果。單獨使用建議0.5，如果搭配其它模型、Lora或者某些採樣器使用發生了异常，建議隨機調整權重，最高不要超過0.9。
可以用於生成澀圖，但禁止用於商業活動。如果你喜歡這個LoRA，請点赞💛或五星☆评价⭐️支持！
No1mm
第一次炼丹，人物是 八神可可，我最喜欢的游戏《Ever17》中的人物。
炼丹没啥经验，效果好坏我也不知道咋判断。
My first LoRA model for YagamiCoco (八神ココ) ,one character of my favorite game — Ever17.
yagami_coco
Do you guys have any suggestions for next one?
underwaterstyle
Actress
the training was on a model - CyberRealistic 2.0.
I'm Render on - EndlessReality
if you want to get speed, drive a sports car
if you want to drive off-road, take an SUV
if you want to hammer a nail, take a hammer
it's the same here - if you want to get results, read the recommendations and follow them
personally, I most often perform generation with parameters: DPM ++ 2M Karras, Sampling Steps 35, Denoising strength 0.35, CFG Scale 3.5-4.5
chilloutmix出图，强度0.6-0.8，注意负面tag我使用了EasyNegative：
EasyNegative - EasyNegative | Stable Diffusion Textual Inversion | Civitai
Chilloutmix output image, intensity 0.6-0.8, note that my negative label uses EasyNegative:
EasyNegative - EasyNegative | Stable Diffusion Textual Inversion | Civitai
系列说明：
No.1xx 是我不太熟悉的人
No.2xx 是我知道的明星或博主
No.4xx 是一些实验性风格概念等。
Series description:
No.1xx is someone I am not familiar with
No.2xx is a star or blogger I know
No.4xx are some experimental style concepts etc.
我是dawn6666，我之前的账号被封禁，今后的内容在这个账号或者我的频道中发布（https://t.me/dawn66666666）。
I am dawn6666, my previous account was banned, and future content will be posted on this account or my channel（https://t.me/dawn66666666）.
close-upupper body
I trained this LoRa on MeinaMixv9 and I generated the sample images in the same model.
Recommend weight for this LoRA is 0.5 to 0.7
Recommended prompts if you want to more close to original character: seele(honkai star), 1girl, (((solo))), purple eyes, purple hair, hair between eyes, long hair, breasts, cleavage, medium breasts, scarf, black shorts, sidelocks, curvy body, white leotard, boots, ((open clothes)), long legs
This LoRa will be update in the future when I can look for a better image dataset :D
seele(honkai star)
It makes objects made of ivory and gold.
Created as described in the tutorial: https://civitai.com/models/52697/tutorial-konyconi-style-lora
Perpetual coffee request: Keep me motivated and caffeinated:
IvoryGoldAI
breastsheart hands
This is a cup of damn fine coffee.
DaleCooper
quality may not good well due to insufficient number of datasets.
activation prompt: overdose
recommended prompt: overdose, sitting, chair, instrument, microphone, cable, computer, monitor, holding instrument, electric guitar, poster (object), keyboard (computer), mouse (computer), amplifier
best quality at 0.8~1.0 weight
overdose
权重0.7-0.9出图
负面词用了EasyNegative
请勿用于任何商业用途和其他非法用途，严格遵守各国法律。
下载即表示您同意上述协议。
最后补一次，出意外不再补了，权重0.7-0.9出图
负面词用了EasyNegative
请勿用于任何商业用途和其他非法用途，严格遵守各国法律。
下载即表示您同意上述协议。
" If I decide to be an idiot, then I'll be an idiot on my own accord ! ” - J. S.Bach
..............................................................................
JOHANN SEBASTIAN BACH = Iconic Musicians Series
.
Johann Sebastian Bach (born March-1685 – died 28-07-1750) was a German
composer, organist, harpsichordist, and violinist of the Baroque Era.
Although he was admired by his contemporaries primarily as an outstanding
harpsichordist, organist, and expert on organ building, Bach is now generally
regarded as one of the greatest composers of all time.
.
Bach produced countless masterpieces, including solo keyboard works, cantatas,
concertos, oratorios, and solo and orchestral suites. Also renowned for his
ingenious technical ability in contrapuntal composition.
In his lifetime (of 65 years), Bach composed an incredible 1,128 pieces of music !
There are a further 23 works which were lost or unfinished.
He is celebrated as the creator of the Brandenburg Concertos, The Well-Tempered
Clavier, the Mass in B-Minor, and numerous other masterpieces of classical music.
.
Appearing at a propitious moment in the history of music, Bach was able to
survey and bring together the principal styles, forms, and national traditions
that had developed during preceding generations and, by virtue of his synthesis,
enrich them all.
Bach ranked among the best musicians in Germany in his time, as he was
active as an organist, teacher, director, instrument technician, and composer.
.
Other interesting facts about Bach...
Bach was the father of 20 children. He was married twice.
...
Possibly one of the most famous pieces of classical music, Bach's "Toccata and
Fugue in D minor" has made an appearance in some interesting formats,
including car alarms and Walt Disney's 1940 film "Fantasia".
...
In 1977 NASA sent into space two Deep-Space-Probes. Voyager 1 and 2 each
carried on board a golden record - an interstellar collection of human sounds and
images meant to represent life on Earth for any alien civilization that might come
across it.
Like a message in a bottle, hoping to reach out to other intelligent life.
The music included on the disc had to bear the burden of the entire history of
humanity’s music, because these Deep-Space-Probes might last longer than our
civilization here on Earth !
Considered by many to be the crowning composer of all time, the music of Bach
demonstrates humanity’s appreciation of balance and symmetry.
Perhaps unsurprisingly, two pieces of J. S. Bach music are on the gold-Disc.
.
..............................................................................
..............................................................................
.
I used 21 well prepared high-resolution images (some fixed in Photoshop), from
life-size wax-figures of J.S.Bach wearing his Baroque era 1750 hair wig
and Baroque era clothing.
.
Images were Trained on the Stable-Diffusion-v1-5-pruned.ckpt,
to 1050 steps, with a batch of 2.
.
For Best results use LoRA weight parameters at = 0.7, 0.8 , 0.9
You may test other settings while using diferent Checkpoints for a better
A.i. image render, and for Ai-realistic-Photos or Ai-Artworks.
As usual, the less weight number , the less it will look like him.
.
Made using Kohya-ss. Place it in
"stable-diffusion-webui\extensions\sd-webui-additional-networks\models\lora" folder
in Automatic1111 WebUI root folder.
..............................................................................
The original LORA name is
j07ansb1c4_v1.0
- you may change it to another name if you wish.
-
Other useful words you can use in your Prompts ...
a man with a white 1750 hair wig,
a man with a hair wig,
a baroque era coat,
a baroque era jacket,
a baroque era suit,
..............................................................................
..............................................................................
This is a S.F.W. (Safe for Work) Lora ,
but can also create N.S.F.W. images (Not Safe For Work).
.
.
To MAKE SURE the image renders in a S.F.W. manner, include in your "Negative Prompt"
the following words....
((NSFW)), ((Nude)), ((Naked)), ((Vagina)), ((Pussy)), ((Nipples)), ((Ass)) , ((Penis)),
.
.
To MAKE SURE the image renders in a N.S.F.W. manner, include in your "Negative Prompt"
the following words....
((Clothes)), (Clothing), (pants), (bikini), ((bra)), (underwear), (dress), (shirt),
..............................................................................
..............................................................................
a man with a 1750 hair wig
Alicia Melchiott from Valkyria Chronicles. Trained on NAI.
Tested at 0.7-1.0 strength.
Trigger word is "alicia melchiott".
Add "headdress, red necktie, military uniform, blue thighhighs, brown gloves, boots" for her typical look.
alicia melchiottheaddress, red necktie, military uniform, blue thighhighs, brown gloves
She is an influencer in Tiktok and Instagram:
二次元角色海报立绘，推荐权重为0.8-0.4，多多尝试多多五星好评，欢迎分享相关作品或提问。
Masterpiece, poster design
Chisa Kotegawa from then Grand Blue anime and manga.
5 versions; 6, 7, 8, 9, and 10 epochs, each with 18.1 MB.
Works best with ALL, MIDD, OUTD, OUTALL LoRA block weights, and with 0.5+ weights.
kotegawa chisabrown hairahoge
haruna,four horns,brown_hair,twintails,long_hair,red eyes,hair_ornament,zettai_ryouiki,thigh_boots,short_dress,wide_sleeves,black pants, 
kongou,blonde hair,long hair,blue eyes,white_gloves,garter_straps,zettai_ryouiki,pleated_skirt,white_skirt,stoking,high_heels,white shrite,long_sleeves,black_cape,short_dress,
Train on anything v 4.5, recommend using it to gen image, feel free to play around,
This is an art style I like on Pinterest , also CFG should around 5 to have more dynamic pose
Euler a, Euler, 2M Karras, SDE Karras are recommend
in the style of (lora)
helena,blue hair,long hair,puple eyesblue dress,thigh_boots,elbow_gloves,short dress, 
Hey, everyone. This is Sex Eyes and this is my Queen Draco model. This model has limited dataset so it's not very flexible if you ask me. Follow and join my Patreon to get an updated version in the future.
This model is trained with Anything v4.5
To get an accurate look add this to your prompt
hair over one eye, cape, fur trim, horns
If you like my model, why not join my patreon and support me? You'll get early access to my models.
https://www.patreon.com/SexEyes
I am also on Pixiv
queendracoqueendracohead
Recommended weight 0.75
1girlhair bunschool uniform
Model based on the character Dovakini-chan by Artist: Nisetanaka
Prompt: dovakini-chan, 1girl, red hair, boots, horns,necklace, armor,pubic hair peek, horned helmet, bikini, red hair,white bikini, female pubic hair, abs, helmet, micro bikini, armlet, brown gloves,
dovakini-chan
Well... This Lora draws Panin
Tag: pnnn
Weight: 0.8-1
Checked on Deliberate 2.0
Euler a, 32 steps, 6-9 cfg
⚠️ Be sure to try my other Loras here: https://civitai.com/user/milkomt/models
😈 by milkomt
pnnn
This is a LoRA version of this LyCORIS - https://civitai.com/models/22696/s12-v42-update-or-fine-tune-lora-or-extended-lora
Please, if you want to read description - check it out. (And you better use it as well)
This is version for those who are not using LyCORIS, or just doesn't know how to use it.
Also, if you can use LyCORIS, but prefer to use LoRA version instead - you're shooting yourself in the foot. Extended LoRA version captures composition better, has much more small details, more stable and higher quality.
皇帝的利刃lora（这种眼睛比较多的角色，确实容易出问题，但影响不大）
触发词：empire blade
lora weight：0.8-1.0
lora of empire blade
Necessary trigger words：empire blade
lora weight：0.8-1.0
empire blade
Baby, it turns everything into voodoo dolls :D
Its well cooked but can be a little fiddly to prompt. Use the main tag and take a look in the metadata with additional network to see what else it will recognise.
I'll probably do a v2 next week to add some sticking pins in and some more variety prompts
Based of course on the konyconi style, check out their stuff.
Obligatory kofi link ... buy me a machine oil today!
https://ko-fi.com/theartificialanalyst
and if you read down to here, keep an eye out for my rediculously cool cthulhu lora... any day now i swear....
v00d00t3ch
Worker Drones and Disassembly Drones from Liam Vickers and Glitch Production's show Murder drones, all credit to them. Unfortunately, the drones are extremely dissimilar from most model's perception of a person so you need to crank the lora up to 1.4-1.6 to get any decent results. Use "workerdrone" to get worker drones and "muderdrone" to get disassembly drones. I highly suggest adding "android, robot, robot joints, glowing eyes" to the prompt as well for best results.
I've tried to include as many of the characters as I can in the training data. so I've included their triggers as well. I also tried to get some like stuff the XD facial expression of the disassembly drones, the Absolute Solver Eye, and Error symbols, and drooling oil from the mouth. However, I haven't had any luck with that so give it your best shot if you want.
glitchproductionsworkerdronemurderdronemechanical tailhollow eyesuzidronendronevdronethaddronelizzydronekhandrone
LoRa trained on 30 images focused on Hungarian model Barbara Palvin's face. Recommend using with weights less than 1.5 and with semi-realistic model (dreamshaper etc.)
vampire
dog daya model fine tuned to the character eclair martinozzi
eclair_martinozzi
参考了水墨画《水浒传》人物谱，可以生成自己喜欢的角色。欢迎来小红书（我的账号：Sholyang_）和我交流呀！这个模型我会继续完善，直到能够更完美的呈现好的角色。
推荐参数：
Model：abyssorangemix3AOM3
VAE：kl-f8-anime2
Steps：55
Sampler：DPM++2M Karras
Size：768*1024
CFG：7
尝试用下面这些关键词：
通用描述：ink， white background， simple background， full body
男性外貌：1boy，male focus，beard，facial hair， barefoot， closed eyes
男性着装：armor， helmet， sheathed，boots，belt， sandals，hat， wide sleeves， robe，hair ornament，cape
男性姿态：standing，jumping，holding weapon，fighting stance， sleeves past fingers
武器种类：polearm，sword，mace，spear，scabbard
女性外貌：1girl，black hair，twintails，smile， blush，hair ribbon，long hair， bangs， hair bun， makeup， collarbone
女性姿态：outstretched arms，dancing，standing
女性着装：hanfu，long sleeves，chinese clothes，ribbon，jewelry，hair flower，floral print，bracelet，hair stic，earrings，updo， red lips
ink
Recommend weight: 0.5-0.6
Gardenia's LoRA for character from Wandering Witch: The Journey of Elaina LN Volume 5. The training data was based on the original illustrations of the novel and xLapin's column 'Journey of Illustration Coloring' on Bilibili, with some modifications made to the hair color. "
The link to 'Journey of Illustration Coloring (Simplified Chinese)' is: www.bilibili.com/read/cv11533791/
Note: This LoRA model is not permitted to Commercial Use unless you get permission from the original author of this character, Jougi Shiraishi and Azure. I am creating this model for fan-art only, and not crediting the LoRA to myself.
推薦權重：0.5-0.6
這次我試著訓練了《魔女之旅》第五卷登場的梔子小姐。訓練資料參考了小説原插圖和Bilibili上的xLapin的《插畫上色之旅》這個專欄，并在基礎上修改了部分髮色。
《插畫上色之旅》鏈接（簡體中文）：www.bilibili.com/read/cv11533791/
注：在獲得白石定規先生和Azure先生的原創角色授權之前，此LoRA模型不允許商業使用。我只是為了創作粉絲藝術而創建此模型，並不將LoRA歸功於自己。
推奨ウェイト：0.5-0.6
今回、私は『魔女の旅』第五巻に登場するクチナシさんをトレーニングしました。トレーニングデータは、小説の挿絵とBilibiliのxLapinさんの『插画上色之旅』という専門記事を参考にして、一部髪色を変更して修正しました。
『插画上色之旅』（簡体字中国語）のリンクはこちら：www.bilibili.com/read/cv11533791/
注：このLoRAモデルは、白石定規さんとあずーるさんのオリジナルキャラクターの許可を得るまで、商用利用は許可されていません。私はファンアート用にこのモデルを作成しており、LoRAのクレジットは自分に帰属しないことをご了承ください。
gardeniamajonotabitabi, 1girl, light green short hair, golden eyes
lora of wisp from warframe
wispframehoodno eyesfaceless
Lora trained on Khun Aguero Agnis from Tower of God
small dataset (29 images mostly close-up or upper body images) it might be tricky with full body shots
works great with upper body and potraits though
tags: khun aguero agnis, 1boy
optional: androgynous, light blue hair, blue eyes, shirt, necktie
use streaked hair for the regular hair or short ponytail
models used:
2,5D: MeinaUnreal - V3, majicMIX
realistic: fantasticmix_real
negative embeddings: embedding:bad-hands-5, embedding:ng_deepnegative_v1_75t, embedding:bad-artist, embedding:EasyNegative
Recommended settings:
Make sure you are aware on the usage instructions of LORA
VAE: vae-ft-mse-840000-ema-pruned, kl-f8-anime2, blessed2.vae (for anime models)
Sampler: DPM++ SDE Karras or DPM++ 2M Karras
Steps: 20
Clipskip: 2
Upscaler : img2img in comfyui using 4x-UltraMix_Smooth
CFG Scale : 5 to 9 (mostly 7 and 8)
LORA weight for txt2img: 0.6-1
Denoise strength for img2img: 0.4 to 0.7
If you want the exact same settings, just drag and drop the picture into ComfyUI
for the pictures i used comfyui
khun aguero agnis
by Alphonse Mucha
If you having any problems while using it, check settings the sample images first.
v2 319 images used! My current record. Often more images make models worse, but I found enough high quality, and used a bunch of new settings (Check version info for settings). I think it is an improvement on v1.
Want a LoRa? Hit me up on ko-fi. Link in bio.
Lora has been trained on a few concepts.
Triggers: pawg, gympawg, bodybuildling, tomboy.
This is my checkpoint model. FitCorder. Check out my other Lora's.
Body Type: Double Thick Thighs, Perky Breasts, Round Breasts, Max Cleavage, Amazon Goddess,
Fashion: Crop Tops, Choker Minidress, Scrunchbutt leggings, Cybergoth, Nerdy
Environment: Undersea Depths
pawgbodybuildinggympawgtomboy
Lora trained on Zafina from Tekken
tags: zafina, 1girl
optional: bodysuit, skirt, long hair, jewelry, magic, aura, big hands, blue bodysuit
models used:
2,5D: MeinaUnreal - V3, HELLmix, PsyGon Animerge - v1.1, CamelliaMIx_2.5D_V2
realistic: fantasticmix_real
negative embeddings: embedding:bad-hands-5, embedding:ng_deepnegative_v1_75t, embedding:bad-artist, embedding:EasyNegative
Recommended settings:
Make sure you are aware on the usage instructions of LORA
VAE: vae-ft-mse-840000-ema-pruned, kl-f8-anime2, blessed2.vae (for anime models)
Sampler: DPM++ SDE Karras or DPM++ 2M Karras
Steps: 20
Clipskip: 2
Upscaler : img2img in comfyui using 4x-UltraMix_Smooth
CFG Scale : 5 to 9 (mostly 7 and 8)
LORA weight for txt2img: 0.6-1
Denoise strength for img2img: 0.4 to 0.7
If you want the exact same settings, just drag and drop the picture into ComfyUI
for the pictures i used comfyui
zafina
Please consider to support me by buying me some coffee so I can make a new LoRA faster. I do open commissions there too if you want to make your own custom LoRA: My BMAC page
Trigger word: staycsumin
Best checkpoints/models to use:
Realistic: VisionGen, ChillOutMix, ArtEros, Snapdd00, Henrealmix
Fanart/painting style: Dreamshaper, LucidMix, SunshineMix, Majicmix
Join KPOP Diffusion Discord: https://discord.gg/NrBDMX79Da
staycsumin
一条碎花小裙子
希望您能尝试并分享生成的图片和反馈
A flower dress
Hope you can try it and share the image you generated. Eager for your feedback~ Thanks~
Recommend config:
LORA weight for txt2img: reccomended 0.8
Keys: flower_dress
flower_dress