正式版已更新，3d画风，非常烧的飞飞公主，权重推荐1，记得打关键词。请勿商用。

The official version has been updated, 3D painting style, very sexy Princess Feifei, weight recommendation 1, remember to type keywords. Do not use for commercial use.

Only initial vertical drawing is used as the training set, and breakthrough vertical drawing is not used.So there aren't many additional effects.

trained on danbooru drawings of beautiful women by tomato rice

• v2.0 = 70% Anything v5, 15% CamelliaMix, 15% NAI + Basic Filewords

• DBv = 40% CamelliaMix, 30% NabiMix, 30% Anything v5 + Deepbooru

v2.0 examples using trigger only and no negative on the same seeds

DBv examples using training prompts and EasyNegative on the same seed

https://huggingface.co/Powidl43

Chita-san [MeinaMixv9]

I trained this LoRa on MeinaMixv9 and I generated the sample images in the same model.

• Recommend weight for this LoRA is 0.6 to 0.7

• Recommended prompts if you want to more close to original character: 1girl, (((solo))), chita_(yancha gal), small breast, orange jacket, black skirt, school shoes, orange hair, long legs, (((large legs))), adult character

Now you can make all the images of Chita-san you want <3!!!

Senator Steven Armstrong from videogame Metal Gear Rising: Revengeance

"Don't fuck with this senator"
Really cool character, I'm glad with the results considering the lack of fanart and base images. I fucked up with the training and it's recommended to always pair the trigger word with "muscular male" to make him wider. I also tried to make his black spots with tag "nanomachines" but it's not possible, the AI doesn't understand it atm.

Trigger word: senatorarmstrong

Tips:

• For his topless outfit: topless male, veins

• For his shirt + necktie: white shirt, necktie, black pants, black belt

• For his suit: black suit, white shirt, necktie, black pants, black belt

• Extra: glasses, cigarette

整个烂活，哔哩哔哩董事长陈睿。

建议搭配丁真，孙笑川食用

Hiii ໒꒰ྀི´ ˘ ` ꒱ྀིづ !! Another cosplay costume !! my fav girl raven this time ! hope you enjoy ! Please let me know any issues and I will try and improve in the future ^^

My intended purpose of this Lora is to let any character or person you want cosplay as Raven from teen titans !! <3 Sorry if this one doesn't fully meet this expectation !

I used ChillOutMix to train and produce some of the images !! ^^

I recommend keeping the weight between (0.6-1) and using the negative prompts as needed !! Higher weights tend to heavily sway the images output to the original dataset so I recommend lower weights if your okay with losing the original style a bit !!

Important Trigger word Info (pls read)

Raven_Cosplay is essential to getting the exact style of the images above !!!

1. Raven_Hood (puts a ravens hood on head)

2. Raven_No_Hood (Takes hood off)

use trigger words Black leotard, purple hair, short hair, white skin

I also highly recommend using the inpaint face trick to fix the faces !!

Some known issues are Certain loras don't work well with this one, hard to get the hood sometimes depending on strength, has a face preference sort of. My apologies for all of these issues ^^

So that's all the colors! I mainly put this here so there's no confusion (ෆ˙ᵕ˙ෆ)♡

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

Please use negative prompt Embedding (Easy Negative) it works a lot better with it then without it !! I've noticed many issues with odd hands as well so you can also use Bad-Hands-5 !!

Please be kind with this Lora ♡

Oikawa Amane 及川あまね | ふぁみこん

https://www.pixiv.net/users/86166786

地点：(Night:1.7), Cyberpunk_CityView, Before Window,
动作：sitting at attention,sitting on couch,
服饰：collarbone,Black_shirt,jeans, (Full_body:1.2), cleavage,
角色：<lora:aiueOkaStyle_v10:0.1>, <lora:saiminSeishidou:0.1>,<lora:OikawaAmane-V2:0.7>,
头发：green_eyes,brown_long_hair,single_braid,bangs,
角色鎖定：
1 girl,Solo, 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: 2339949661, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original

The "main" LoRA for muv-luv, probably significantly less compatible than the individual character ones, but maintains the original style way better. Should cover all girls from the main series, Takeru should be promptable aswell. For prompts on individual characters check the individual loras that I uploaded. Nothing is baked in so stuff like "ahoge", breast sizes etc should be specified for best results.
Not sure which of the two versions is better, I think I like the final5e5 look more, but it's probably worse at following prompts. Let me know if you want any of the other versions.

Might be possible to use it at low strength with the individual character ones to maintain flexibility and add back the style but I haven't tested it much.

怪物王女の姫（リリアーヌ）を描いてくれるLoRAです。

トリガーは以下のようになります。

脱がせたりコスプレさせたい時は hime

原作のドレスを着せたい時は更に hime dores を入れてください。

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

I do not recommend using my LoRA with model RealisticVision

personally, I most often perform generation with parameters: DPM ++ 2M Karras, Sampling Steps 35, Denoising strength 0.35, CFG Scale 3.5-4.5

evaluation of the model is welcome

Style LORA based on the works of ER404. I was asked to give this one a shot and the requested liked it so here it is. The eyes look off at times, might need better/more data.

Game character LoRa model

Weight: 0.6 - 0.9

Tested dimensions:

• W600 * H600

• W600 * H900

• W800 * H800

Note: If you end up generating white bars on faces, add ((censored:1.2)), ((white box:1.2)), ((mask)), ((white bar)), dialog_bubble, ((white rectangles)), ((grey_scale)), dialog_box, to the negative prompt to mitigate it.

Legal Disclaimer:

This LoRA model is an artificial intelligence-based software that generates images. The user of this software is solely responsible for ensuring that the generated images comply with all applicable laws and regulations.

The user of this LoRA model agrees not to use the software for any illegal purposes, including but not limited to the creation or distribution of child pornography or any other illegal content.

The owner and developer of this LoRA model disclaim any liability arising from the use of the software for illegal purposes. The owner and developer of this LoRA model do not condone or support the creation or distribution of any illegal content, and the use of the software for such purposes is strictly prohibited.

By using this LoRA model, the user acknowledges and agrees to be bound by this disclaimer. If the user does not agree to the terms of this disclaimer, the user must immediately cease using this LoRA model.

你的林皇，无限猖狂

someone looks like

似てるの人

training model: chilloutmix_NiPrunedFp32Fix

weight: 0.8

Model: YesMix

Weights: 0.5-0.8

Keywords: ball gag, gagged

If you like my works, you can buy me a coffee to support
Buy me a coffee

Kofi

Base model:Anything4.5

Weight:0.8~1

Trigger word is not needed but you can add "chibi" when used with character lora

Trainned by AOM3 model and it works quite well under weight 0.9~1.0.

Suggested VAE as orangemix

Trigger Words:

xiaoroubao

If you would like to kindly support me, you can visit https://www.patreon.com/user?u=29398121

*Disclaimer: Please DO NOT use for any commercial usage.

Maylene from Pokemon Diamond/Pearl. Trained on NAI.

Tested at 0.7-0.9 strength.

Trigger word is "maylene".

• Add "blue tank top, white pants, fingerless gloves, barefoot" for her typical look.

• "Band-aid on nose" adds her band-aid, but is slightly inconsistent.

other characters from anime:

Astarotte Ygvar

Asuha Touhara

Naoya Touhara

Ingrid Sorveig Sorgrims

Judith Snorrevik

Griselda Reginhard

Elfreda Mirjasdottir

Cuthfleda

Olaf Frismar

Elika Drakul Draupnils

Lucca Austri

Unnbjorg Signar

Mistrune Asgrim

Mercelida Ygvar

Ursula Sumarlidi

Isold

Sigurd Sveinnsson Svarthaed

Isadora Finnsdottir

Porhelga Svarthaed

Backgrounds

Dragon

Fii

Enja Níuhali

Teacher

Atli Reginhard

Miina-sensei

Naoya Touhara_trap

Lama

Works fine with all sd 1.5 models. Trained on this

strength 0.6-0.8, hiresfix or img2img for best results(it's works for every LoRa).

Train models isn't easy, crypto donations on some coffee are welcome:

Donate

Boosty, subscribtion with additional content:

Boosty main

Boosty donate link

Five-seveN from Girls' Frontline

Features the primary character LoRA and 4 separate costume LoRAs
Switch between versions at the top of the page to download costume LoRAs

The primary is only trained on the character and should be used for undefined outfits and can be combined with the costume LoRAs to get their canonical appearances

Primary LoRA Keywords:
FiveSeven
FiveSevenLoli

Costume LoRA's Keywords:
FiveSevenNormal
FiveSevenCute
FiveSevenSwimsuit
FiveSevenWedding

Belgians for GF2 when
Recommend 0.8-1 Strength

If you've enjoyed my models, consider leaving a tip.
https://ko-fi.com/leontalon

This is a LORA of the art from Shin Megami Tensei 4 and 4A. It is entirely trained on art from Mayasuki Doi, and includes some training on demon art as well from the series.

To prompt this artsyle, enter "<lora:SMTDoiStylev1:strength>SMTDoiStyle,"

I like this LORA the best looking at 0.6 ~ 0.8 strength in my opinion. This model was trained on AnythingElsev4.5.

这是一个男生真人人脸特征训练的性转lora，由于训练集问题无特别指定负面提示词(greyscale:1.5)的话图片会偏灰，虽然我个人对这种画质挺喜欢的，肤质感觉会好看，但是有朋友说不是很好，可以附加其他lora或者负面提示词优化，如果后期反馈不错会尝试融合其他模型改善

该lora仅眉毛鼻子和嘴巴特征与真人相像，并且是美化的特征

majicmixRealistic_v2底模训练

V6版较v5低对比度有改善，融合比见版本说明

v9版权重1似乎脸部会有点油的感觉，但是测试下来似乎不用添加什么额外的提示词，预设的正常tag就可以

比如：

正面提示词：best quality, ultra high res, (photorealistic:1.4), 1girl,solo,looking at viewer

负面提示词：paintings, sketches, (worst quality:1.5), (low quality:1.5), (normal quality:1.5), lowres,

克苏鲁神话是以美国作家霍华德·菲利普·洛夫克拉夫特的小说世界观为基础，由奥古斯特·威廉·德雷斯整理完善，并由诸多作者所共同创作的一系列文学作品。

The Cthulhu Mythos is based on the novel world view of the American writer Howard Philip Lovecraft, compiled and perfected by August William Dress, and a series of literary works co-created by many authors.

古神 ancient gods：

1. 哈塔斯Hastur：黄衣之主 King in Yellow

眷族 monsters:

1. 深潜者DeepOnes: 大鱼人 Fish Monster

Natalia 異世界ハーレム物語 ナタリヤ V 2.0

https://www.pixiv.net/users/86166786

地点：(Night:1.7), IN THE FOREST, village, WOODEN HOUSE,
动作：sitting at attention,sitting on couch,
服饰：dress,green suit,DEEP_cleavage,
角色：<lora:Natalia-V2:0.7>,elf,pointy_ears,<lora:aiueOkaStyle_v10:0.1>, <lora:saiminSeishidou:0.1>,
头发：blonde_long_hair,bangs,green_eyes,
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: 178334103, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original, Denoising strength: 0.5, Hires upscale: 2, Hires steps: 18, Hires upscaler: R-ESRGAN 4x+ Anime6B

大家好，我是你们的斧头哥。

这次给大家带来的是我制作的第一款风格化Lora模型

我将以自己的id加上风格名作为风格化模型的命名方式，希望大家能够习惯

这是一款油画风格的lora，在表现景观上尤为出色，不过在制作人物上也是没问题的

推荐使用权重0.75

Hello everyone, I'm your axebro.

This time I bring you my first stylized Lora model.

I will use my ID and the style name as the naming convention for stylized models, I hope you can get used to it.

This is an oil painting style Lora, which is particularly good at depicting landscapes, but there is no problem with making characters either.

It is recommended to use weight 0.75.

Based on the style from the comic series I Hate Fairyland

This one's a bit rough, but it can output some interesting things sometimes.

(Not a generated image)

Discord >>> Inestable Difusion ver 1.0.5

• Strength: 1

• Checkpoint used: braBeautifulRealistic_v40

English:

If you want to use, read this:

-Set output scale to 512x512, and rescale the result yourself if you want.

-Remember to add the triggerword mc112block

中文:

若您想要使用，请注意：

-尺寸输出设为512x512，之后自行缩放图片。

-记得设置触发词“mc112block”，不然大概率输出的不是mc风格

I have finally made something useful! Less than perfect, but worth a try.

This LoRA gives 512x512 textures that good at creating full block textures.

TRIGGER WORD "mc112block" don't forget to add this, or you may get a mess

"Full Block"-That is, things like bush or flowers aren't supported. Trapdoors and doors are trained, but as sd is not good at handling transparency, so use it at your own risk. It won't hurt anyway. Items are not included and will be released in another model.

512x512 - Its theortically ready for a texture pack directly, but minecraft produces atlas in the size of 4096, huge textures easily crashes the attempt to mip map and may result in crashes. It is STRONGLY recommended that you rescale it into 16x16 textures before actually using.

1.12.2 - I am a 1.12.2-holic, so these aren't trained with the textures of the higher versions. This also helps in uniting the style, that's why I put mc112block in the trigger words, just in case you want to merge it with something else.

prompt S/R - please note that the trigger word has a "block" in it. If you use "block" as a placeholder, you will spoil the trigger word like this:

Lora for Helena from doa. This was trained with 饭特稀_v0.8 which you can find on civit. The lora should work ok on similar models but it looks best on 饭特稀_v0.8. Trigger word is "helena".

Also highly recommend EasyNegative which my sample images are using.

If you get bad looking faces at lower resolutions try inpainting only the face at higher resolution or using hires fix.

AL version is a little bit different from DOA version——DOA version has elbow gloves and thighhighs, and I trained them together.

Azur Lane version: put "elbow gloves" and "gloves" in Negativa Prompts.

DOA version: put "elbow gloves" , "gloves" and "blue thighhighs" in Prompts.

If you want to change her clothes more easily, just do not input "blue one-piece swimsuit".

舰B版本跟DOA版本有些许不同，DOA版本有长手套和过膝袜，我把它们练一块了。

舰B版本：把"elbow gloves" 和 "gloves" 放在反咒里。

DOA版本：把"elbow gloves" , "gloves" 和 "blue thighhighs" 放在正咒里。

如果你想换衣服就别输入"blue one-piece swimsuit"了。

Karina

trigger word : karina

3000 EVA theater version screenshots datasets,

which can better restore the EVA style.EVA characters such as Asuka, Ling Boli, and Dian Zhensi can quickly switch through the names of the characters. Later, they will further optimize the EVA style.

I will always like souryuu asuka langley！！！

lora recommend weight:0.5-0.7 CFG recommend:3.5-5

lora:koreanDollLikeness_v15

Hi I am Yamada San, I publish LoRA and models regularly.

More hot photo on discord and twitter! Please follow me if you like!

https://twitter.com/yamada_artist

https://discord.com/invite/NfXu9QDX96

This loRA is specific to generate cosplayers who wear act as a Red Girl Ninja Fighter.
They are just illustration and imagination.

This loRA is not depicting any specific persons or any organizations. It provides only the style and random outlook.

No commercial use for this LoRA.

Yuzuriha of Keishu - Hell's Paradise (Jigokuraku) - LoRA

Consider supporting me on Patreon for more models and content.

Anime character Yuzuriha from Hell's Paradise. You can produce the character from the anime with the right prompts.

Try tags that has to do with her appeareance.

(After full anime release I'll make a new version if necessary)

Example:

yuzuriha_jigokuraku, bangs, cleavage, bare_shoulders, red_eyes, hair_ribbon, ponytail, weapon, purple_hair, japanese_clothes, sleeveless, hip_vent, ninja

Required token: yuzuriha_jigokuraku

Trained using AnythingV5, but should be ok with similar models. Weight should be near 0.9-1.0 (used 1.0 most of the time)

Please credit me when using any of my models.
Specially when merging this model.
You cannot use my models in other generation websites.

Yamada Asaemon Sagiri - Hell's Paradise (Jigokuraku) - LoRA

Consider supporting me on Patreon for more models and content.

Anime character Yamada Asaemon Sagiri from Hell's Paradise. You can produce the character from the anime with the right prompts.

Try tags that has to do with her appeareance.

Example:

asaemon_sagiri, long_hair, black_hair, hair_between_eyes, brown_eyes, ponytail, weapon, japanese_clothes, sword, kimono, sash, obi, katana, high_ponytail, white_kimono

Required token: asaemon_sagiri

Trained using AnythingV5, but should be ok with similar models. Weight should be near 0.85-1.0.

Please credit me when using any of my models.
Specially when merging this model.
You cannot use my models in other generation websites.

原始数据集来自群友,我负责处理数据集和训练,触发词是:khorosho \(kancolle\)

The original dataset is from the my QQ group friend, I am responsible for processing the dataset and training, the trigger word is: khorosho \(kancolle\)

原始CKPT中也有响的tag，但是因为那个tag效果不太稳定所以炼了一个lora固定校服的形象，使用原始tag和触发词可以让角色形象更稳定,并且原始响tag会很可能出现响的舰炮

The original CKPT also has a ”hibiki \(kancolle\)“ tag, but because that tag effect is not very stable so me and my friend made a lora to fixed school uniform image, using the original ”hibiki \(kancolle\)“ tag with the trigger words can make the character image more stable, and the original ”hibiki \(kancolle\)“ tag will probably appear hibiki's ship gun

LoRA for Jinora from The Legend of Korra. Trained on AnythingV5 this time.

Recommended weight between 0.6-0.8; can go higher or lower depending on additional LoRAs, etc.

Outfits: wingsuit, airbender robes (easier to prompt on heigher weights)

Tattoo: blue-arrow forehead tattoo, blue-arrow body tattoos

Used Hi-Res Fix (Latent) 2x from 512x512 —> Extras (4x-UltraSharp) for final 2048x2048 resolution.

only test

它是在512x768分辨率进行训练的，我测试了一些图片，但是有点模糊，需要放大一点分辨率。

It was trained at 512x768 resolution, and I tested some pictures, but they were a bit blurry and needed to be enlarged a bit.

我测试了尽可能不影响衣服的情况使用其他的lora的脸，tifa的脸特征很明显，参考图3和图4。

I tested using other lora faces as far as possible without affecting the clothes, and the face features of tifa were obvious, refer to pic 3 and 4

我设置了一个触发词，它是fenglingCLo。但它没有效果，不引用也可以。如果你知道原因，请帮我解开疑惑

I set a trigger word, and it's fenglingCLo. But it doesn't work. It doesn't work. If you know why, please help me out

使用恋活人物卡训练,人物卡作者是haruki

I used the character card to train, the character card im using was created by haruki

因为是3D模型训练的,一般用0.7或更低比较好，lora的训练底模是原版大模型 model[e6e8e1fc]，触发词是: medeiros \(magiarecord\)

Because this lora was trained by 3d modle,use 0.7 or more less is batter.This lora was trained by original big model[e6e8e1fc]，Lora trgger words is ：medeiros \(magiarecord\)

This is a historical Japanese festival costume.

3Q Style

This style LoRA offers a wide-range of vibrant color combinations and good "unnatural" skin colors (ex. blue skin), good at generating female characters with stylized proportions.

Use danbooru tags in tandem with English caption for best result.

Recommend Strength: 0.7 - 1.0 CLIP Skip: 2

Base model: AnyLora fp16 Pruned (no VAE) VAE: OrangeMix VAE

Best used with: EasyNegative

LoRA Factory 2023

Please note that the pussy generated by this lora is not very tight and pretty.

It was trained using 71 realistic images and 65 anime images, with a training model of SD1.5 and a training clip of 1, so it should perform well in both the [realistic] and [anime] models.

It depicts [pussy] as [loose, gaping, rotten, disgusting].

I'm tempted to bury my head in it and taste the flavour, bro, would you want too?

Weight=1

Negative tips = white background, simple background

When you use a realistic style base model, enter realistic as the prompt

When you use the anime style base model, enter anime as the prompt

Of course, you can also use them together, the difference being that for training purposes, the anime style [big pussy] is marked by me as [anime] and the realistic style [rotten pussy] is marked as [realistic]

I will also be making more hardcore content related to feet, armpits and disease.

Note: Due to the specific nature of the training images used, you can try typing "dirty armpit" and it will also work well, it will produce a... er... rotten looking armpit, I like the taste!

Marley from Pokemon Diamond/Pearl. Trained on NAI.

Tested at 0.7-0.9 strength.

Trigger word is "marley".

Add "hair bow, gothic dress, detached sleeves, black leggings, black boots" for her typical look.

lora:koreanDollLikeness

Trained on NAI.
(Request)

Recommended weight: 0.8 - 1.

it can work in realistic models but results are not the greatest and need multiple tries.

Feel free to share your result on the comments and a review.

Also if you want you could you join my discord server, have talk, share result, feedback, tell me a request perhaps.
Discord

As with all my models, I tried to stay close to the original style.

"fake rabbit ears" for her Alternative ears. "fake animal ears" for her Altered Fable uniform "ears". Both in negatives should get rid of both, success with all of this can vary between models. If her alternative uniform is showing up too much just put "uniform" and "juliet sleeves" in negatives.

Added a v2, should be a major improvement.

UPDATED to 1.5.

Should work much better with other LoRAs now and give consistent results.

Tags: (>_<:1), closed_eyes

Recommended tags: (chibi:1) - greatly improves stability, I recommend using a lower weight such as 0.5 when making images of other styles. (>3<:1), (>o<:1), (>w<:1) also may give some results.

CFG should stay between 7-11 for best results, sometimes steps over 30 can make the eyes look more like they're normally closed.

lora:iu_V35

<lora:iu_V35:1>

Read the "about this version" for some useful information about the lora version

I'm going to rewrite this section to add more information and tips about lora, with poses, camera angles.

You can use, change and share any of the sample images.

If you have any suggestions I can try to implement.

Sharing my lora with you, it works with many of the models available, even photorealistic(horse_penis_V0), v1 wasn't tested on photo realistic models.

I was doing some tests with v12,v13,v13a with some 3d renders related to rear view, which works fine in v13a, the other versions work sometimes, I'm planning to add first person view.

Prompts without negative were made for testing purpose only.

Rear view should work fine with prompts like "futanari, 1girl, ass focus, ass, cum in ass, anus, testicles, horse penis " with "from behind, rear view" at the end.

Because it is trained using clip 2, on models trained with clip 1 it tends to produce strange results, however it is possible to make it work using other layers,

Negative embed are from the links below

https://huggingface.co/datasets/gsdf/EasyNegative

https://huggingface.co/NiXXerHATTER59/bad-artist

https://huggingface.co/yesyeahvh/bad-hands-5

Saya Takagi from Highschool of the Dead

• Trained with anime (full-final-pruned model)

• 3 versions; 7 epochs, 9 epochs, and 10 epochs

• Tip: Works well with ALL, MIDD, OUTD, and OUTALL LoRA weight blocks.

Lora for Tina from doa. This was trained with 饭特稀_v0.8 which you can find on civit. The lora should work ok on similar models but it looks best on 饭特稀_v0.8. Trigger word is "tina".

Also highly recommend EasyNegative which my sample images are using.

If you get bad looking faces at lower resolutions try inpainting only the face at higher resolution or using hires fix.

Hi everyone!
This LORA takes ultra realistic macro photos of human eyes!

You can use it directly in Prompt to take macro photos of eyes, or you can use it in inpaint.

Share your generations, I will try to improve this LORA as I'm very interested in it!

trained on danbooru drawings of beautiful women by piyodesu

• v3.1 = 60% CamelliaMix, 40% NAI + Basic Filewords

• v3.5 = 70% v3.1, 30% Anything v5 + Basic Filewords

• DBv = 40% CamelliaMix, 30% NabiMix, 30% Anything v5 + Deepbooru

v3.1-3.5 examples using trigger only and EasyNegative on the same seed

DBv examples using training prompts and EasyNegative on the same seed

https://huggingface.co/Powidl43

Role of Azur Lane

Weight 0.8~1.0 ,and you can ADJUST prompt to reductive character

TRIGGER:

1.BASE: dressLvzov;

2.DRESS: black dress,black thighhighs, hat

Original outfit updated

OG TRIGGER:

1.BASE: original outfit;

2.DRESS: thighhighs, hat, shirt,black footwear

I recently was trying to make some images of Rin Tohsaka and wanted to see if I could make a model so I could have more accurate clothing and faces.

If your model is guided towards NSFW art I suggest using nsfw, nude if you don't want don't want that. I specifically tagged the dataset to respond to those terms.

Do you like what I do? Feel free to support me on patreon

A Lora of Clair from pokemon.

Use Lora weight of 0.7

elle younger years

The Terraria Zoologist

Version 2 in progress

Works great with Corneo's 7th Heaven Mix & HeavenOrangeMix

Helpful Positives: 1girl, brown skin, fox girl, green shirt

A Lora of Karen from pokemon.

Use Lora weight of 0.7

Beatrix from Final Fantasy IX. Trained on NAI.

Tested at 0.7-1.0 strength.

Trigger word is "beatrix".

• Add "eyepatch, white sleeveless coat, cleavage, gauntlets, belt, fingerless gloves, brown pantyhose, boots" for her typical look.

• The eyepatch might be a little inconsistent, but it does work more often than not.

Lora quite simple that allows to simplify the creation of blowjobs with several added tools and with its own style. There are three variations of lora, but all three do the same thing, the only thing that was changed was the database to create it, so they give different results. The main focus is the blowjob with pov, but it can be used without pov if you are looking for different results. The style is required, as you created it to follow the drawing style seen in the examples and it is not neutral. However, if the intensity of the lora is lowered enough and a style lora is applied to it, I estimate that it would be overwritten (I did not try it). Play tag and have fun.

• Cuasinoto (103) penis (100) breasts (87) 1boy (84) dark skin (83) veins (73) veiny penis (72) dark-skinned male (69) blush (68) large penis (68) 1girl (65) solo focus (65) large breasts (64) erection (61) oral (59) fellatio (59) white background (57) testicles (51) nude (50) pubic hair (45) makeup (45) pov (44) black hair (43) simple background (43) long hair (42) cum (42) short hair (41) tongue (40) male pubic hair (40) interracial (40) tongue out (39) nipples (37) multiple girls (33) lipstick (33) blonde hair (31) lips (31) completely nude (31) blue eyes (30) group sex (30) looking at viewer (30) cum in mouth (29) ponytail (28) ass (28) brown hair (27) brown eyes (26) 2girls (23) huge penis (23) :>= (23) green hair (22) sweat (21) multiple boys (21) 2boys (20) cooperative fellatio (20) green eyes (20) censored (19) huge breasts (18) glasses (18) bangs (17) licking penis (17) licking (17) threesome (17) red hair (17) dark-skinned female (16) after fellatio (16) ffm threesome (16) mosaic censoring (15) rolling eyes (15) stray pubic hair (15) hair ornament (14) saliva (13) large areolae (13) penis on face (13) hair over one eye (13) half-closed eyes (13) black eyes (13) teeth (12) jewelry (12) open mouth (12) hairclip (12) collarbone (12) breasts out (11) tan (11) tanlines (11) earrings (11) cleavage (10) hand on another's head (10) cum in nose (10) mature female (10) 3girls (10) crown (10) star earrings (10) flower earrings (10) multicolored hair (10) penis grab (10) heart (9) penis awe (9) multiple penises (9) foreskin (9) yellow eyes (9) two-tone hair (9) pink lips (9) clothes lift (8) hair tubes (8) dark nipples (8) dark areolae (8) hair bun (8) male focus (8) tattoo (8) muscular (8) muscular male (8) yaoi (8) orange hair (8) looking at penis (8) siblings (7) 3boys (7) thighs (7) thick thighs (7) braid (7) cum on tongue (7) bandages (7) pectorals (7) facial (7) very short hair (7) buzz cut (7) skirt (7) nail polish (7) double bun (6) cone hair bun (6) green lips (6) eyelashes (6) flipped hair (6) bara (6) sideburns (6) purple hair (6) cape (6) dark penis (6) fingernails (6) trembling (5) lying (5) colored skin (5) handjob (5) grey skin (5) mature male (5) pov crotch (5) dress (5) purple eyes (5) swept bangs (5) crossed arms (5) blue-framed eyewear (5) red nails (5) colored sclera (5) hood (4) blurry (4) from side (4) hoodie (4) shirt lift (4) brother and sister (4) incest (4) mother and son (4) netorare (4) bowl (4) flaccid (4) underwear (4) on back (4) fishnets (4) red-framed eyewear (4) facial hair (4) clothed female nude male (4) thighhighs (4) belt (4) skirt lift (4) purple skirt (4) futanari (4) futa with female (4) forehead jewel (4) orange skin (4) precum (4) puckered lips (4) clenched teeth (4) facial mark (4) purple background (4) upper teeth only (4) bikini tan (4) lipstick mark (4) husband and wife (3) bed sheet (3) breast press (3) cum in container (3) all fours (3) headband (3) detached sleeves (3) cum on penis (3) female pubic hair (3) miniskirt (3) black dress (3) covered nipples (3) sisters (3) curly hair (3) teamwork (3) cleavage cutout (3) red lips (3) fishnet top (3) forehead mark (3) no pupils (3) bare shoulders (3) bottomless (2) blurry background (2) father and daughter (2) ahegao (2) sheet grab (2) after ejaculation (2) plump (2) kissing penis (2) mini crown (2) cum string (2) twintails (2) very long hair (2) necklace (2) twin braids (2) streaked hair (2) no panties (2) pendant (2) curvy (2) wide hips (2) green dress (2) big hair (2) multi-tied hair (2) pale skin (2) glaring (2) scowl (2) saliva trail (2) gloves (2) clothing cutout (2) red cape (2) smile (2) green sclera (2) fishnet bodysuit (2) black lips (2) large testicles (2) irrumatio (2) purple nails (2) butterfly hair ornament (2) very dark skin (2) school uniform (2) single hair bun (2) ejaculation (2) cum on hair (2) cum on clothes (2) overflow (2) mother and daughter (1) voyeurism (1) cum on body (1) cum on breasts (1) topknot (1) disembodied penis (1) gokkun (1) solo (1) x-ray (1) stealth sex (1) on stomach (1) black nails (1) 4boys (1) blue skin (1) gangbang (1) hair over eyes (1) television (1) double handjob (1) huge nipples (1) tears (1) goatee (1) long sideburns (1) spoken heart (1) no bra (1) looking up (1) ahoge (1) leotard (1) looking at another (1) fat (1) closed mouth (1) blue gloves (1) superhero (1) grin (1) light areolae (1) steam (1) steaming body (1) semi-rimless eyewear (1) ninja (1) emphasis lines (1) snot (1) purple cape (1) caressing testicles (1) forehead protector (1) gyaru (1) orange eyes (1) excessive cum (1) piercing (1) shirt (1) star (symbol) (1) mole (1) from above (1)

Mirrored Aviator Sunglasses

WHAT?!? The Instagram model porn star guy made a STYLE LoRA?

Indeed I did. I made this for a friend of mine who wanted realistic mirrored sunglasses for their artwork. I had no idea where to start so I went to the Discord and got some great advice, and voila! Mirrored shades!

This LoRA was trained to only do Aviator style sunglasses. I was going to include a bunch of different varieties (wraparound, circular, cat's eye, etc.) but this was way too much work. I now have a greater admiration for those brave souls who have given us LoRAs of buttholes pretty scenery.

DISCLAIMER: Sometimes you wont get any sunglasses, sometimes they wont be mirrored, sometimes they won't be aviators, sometimes they will be a mangled mess embedded in someone's skull. That's what INPAINTING is for. If you can't get a decent pair of shades with this LoRA, that's on you my friend.

Like my work? Buy me a beer!

and who knows, maybe I'll do something like this again someday.

Check out my Twitter where I post daily!
Commissions are open and my Patreon is up!

Harley Quinn Commission, based on Fortnite's in game model and other resources!

Some information about the images I shared:
On the last 2 images you can see they're the same. All of these images in the initial batch (nightclub images) were used in this way:

• txt2img: 512x768 -> upscale with ESRGAN 4x+ -> 640x960 (0.35 denoising)

• img2img: downscaled to 512x768 (0 denoising)

• img2img: upscaled to 640x960 without additional LoRAs (0.15-2 Denoising)

Its a bit complicated, but the last image uses one of the original prompts I use to generate the txt2img image. Hope this helps and if you have any questions, let me know!

Abella Anderson

She left us WAY DAMN too soon. Abella is alive and well, but she left the industry right at her peak, and that makes me sad.

However I DID IT! I've been working on this LoRA for months. Trying, failing, putting it away, coming back to it, trying again, failing again, repeat. However I am finally satisfied with her likeness. You see, kids? Never give up on your dreams!

I do not have an embedding for Abella, but I'll create one if the demand is there.

Like my work? Buy me a beer!

For Aponia Outfit use tags: aponia_outfit, 1girl, blue eyes, breasts, large breasts, looking at viewer, nun, veil, black dress, breast curtains, detached sleeves, long sleeves, frills, pelvic curtain, sideless outfit, thigh strap

If you want to add her wings to her outfit add: butterfly dress

For bikini, we use: aponia_bikini, aponia, 1girl, breasts, large breasts, white headwear, earrings, hair ornament, butterfly ornament, frilled swimsuit, frills, one-piece swimsuit, criss-cross halter
For casual: aponia, 1girl, blue eyes, breasts, large breasts

I also included a LoRA that's a bit overfitted on her original outfit, but it can make it accurate in like 95% cases even when using 4+ LoRAs, but it has really hard time making her wear anything else. Use it if you only care about her outfit or you can't get her outfit to look good with standard one.

For OG_outfit: aponia, aponiaoutfit, 1girl, breasts, large breasts, blue eyes, mole under eye, eyes visible through hair, mole, nun, blonde hair, long hair, bangs, hair between eyes, veil, black dress, breast curtains, detached sleeves, long sleeves, frills, pelvic curtain, sideless outfit, butterfly dress

Dies ist Christiane Friedrich, die deutsche Austauschschülerin aus der Visuellen Novelle und japanischen Animationsserie Majikoi (Maji de Watashi ni Koi Shinasai, Liebe mich wahrhaftig).

Use about 0.7-0.8 weight.

Trained on AnyLora model using mostly official art.

I gave all the preview pictures a 1:1 "upscale" using R-ESRGAN 4x+ Anime6B, for sharper lines.

Timido from Honkai Impact 3rd

Main trigger for the character is "timido cute, bangs, pink hair, grey eyes, short hair, bob cut, hair ornament, hairclip, large breasts".

Weight for all sample image is "0.7"

Lora quite simple that allows to simplify the creation of blowjobs with several added tools and with its own style. There are three variations of lora, but all three do the same thing, the only thing that was changed was the database to create it, so they give different results. The main focus is the blowjob with pov, but it can be used without pov if you are looking for different results. The style is required, as you created it to follow the drawing style seen in the examples and it is not neutral. However, if the intensity of the lora is lowered enough and a style lora is applied to it, I estimate that it would be overwritten (I did not try it). Play tag and have fun.

• Cuasinodo (100) penis (96) 1boy (90) breasts (89) veins (88) veiny penis (82) dark skin (78) 1girl (75) solo focus (75) large penis (69) makeup (67) nude (66) long hair (65) erection (64) blush (60) dark-skinned male (60) pov (58) lipstick (58) fellatio (53) oral (51) testicles (51) pubic hair (49) male pubic hair (49) looking at viewer (48) lips (48) large breasts (47) interracial (45) huge breasts (45) cum (36) blonde hair (32) ass (32) blue eyes (31) huge penis (31) :>= (30) collarbone (30) dark-skinned female (29) brown hair (26) black hair (25) white background (25) multiple girls (22) jewelry (22) earrings (22) gloves (22) stray pubic hair (21) green eyes (21) tan (21) cum in mouth (20) completely nude (20) group sex (19) tongue (18) pink lips (18) muscular (18) thighs (18) half-closed eyes (17) nipples (17) 3girls (16) hair over one eye (16) fingerless gloves (16) penis awe (16) mature female (16) cooperative fellatio (15) tongue out (15) precum (15) nail polish (15) colored skin (15) eyeshadow (15) simple background (15) looking at penis (15) foreskin (14) crown (14) star earrings (14) flower earrings (14) green hair (14) muscular female (14) purple gloves (14) green skin (14) open mouth (14) saliva (14) cleavage (14) rolling eyes (14) green lips (13) bangs (13) eyelashes (12) short hair (12) pink hair (12) after fellatio (10) teamwork (10) penis on face (10) puckered lips (10) dark nipples (10) mole (10) purple eyes (10) braid (10) spread legs (10) parted bangs (10) ponytail (9) flipped hair (9) tanlines (9) thick thighs (9) multiple boys (9) mole under mouth (9) on bed (9) bright pupils (9) french braid (9) white pupils (9) very dark skin (8) brown eyes (8) handjob (8) dress (8) bare shoulders (8) purple lips (8) clothed female nude male (8) yellow eyes (8) 2girls (7) sweat (7) kissing penis (7) orange hair (7) fishnets (7) fingernails (7) hair ornament (7) hairband (7) teeth (7) lying (7) on back (7) threesome (6) ffm threesome (6) star (symbol) (6) abs (6) penis grab (6) large areolae (6) disembodied penis (6) hair ribbon (6) bed sheet (6) cheating (relationship) (6) flower (6) dark penis (5) looking up (5) gloved handjob (5) red hair (5) from above (5) breast press (5) covered nipples (5) paizuri (5) underwear (5) cum in nose (5) holding (5) leash (5) red hairband (5) ribbon (5) ejaculation (5) censored (5) forehead protector (5) fishnet top (5) hair flower (5) puffy nipples (5) jacket (5) open clothes (5) sunglasses (5) breasts out (5) eyewear on head (5) sarong (5) hair slicked back (5) tank top (4) facial (4) lipstick mark (4) purple nails (4) food (4) steaming body (4) male focus (4) 2boys (4) pectorals (4) muscular male (4) yaoi (4) red lips (4) multiple penises (4) red dress (4) facial mark (4) forehead mark (4) holding leash (4) leash pull (4) panties (4) sitting (4) sleeveless (4) orange eyes (4) hair intakes (4) fishnet bodysuit (4) hair bun (4) double bun (4) nose blush (4) bikini tan (4) mosaic censoring (3) condom (3) used condom (3) elbow gloves (3) cape (3) clothing cutout (3) cleavage cutout (3) bara (3) red cape (3) 3boys (3) licking penis (3) curvy (3) collar (3) symbol-shaped pupils (3) headband (3) eyeliner (3) red eyes (3) curtains (3) huge nipples (3) cropped jacket (3) tinted eyewear (3) sagging breasts (3) anger vein (2) solo (2) strapless (2) steam (2) large pectorals (2) hat (2) top hat (2) nipple slip (2) licking (2) blue gloves (2) trembling (2) blue hair (2) aqua hair (2) purple hair (2) bracelet (2) all fours (2) thong (2) blue eyeshadow (2) k/da (league of legends) (2) overflow (2) smile (2) closed eyes (2) flaccid (2) gangbang (2) thighhighs (2) sidelocks (2) black gloves (2) black thighhighs (2) white panties (2) black dress (2) pantyshot (2) crossed legs (2) couch (2) two-sided fabric (2) gold hairband (2) two-sided dress (2) twintails (2) japanese clothes (2) tattoo (2) seiza (2) blue kimono (2) shoulder tattoo (2) glaring (2) hair between eyes (2) horns (2) indoors (2) demon horns (2) alternate breast size (2) tiles (2) red horns (2) bathtub (2) cross-shaped pupils (2) hibiscus (2) pink dress (2) visor cap (2) symmetrical docking (2) blue nails (2) spiked hair (2) arms behind head (2) furrowed brow (2) constricted pupils (1) no pupils (1) swimsuit (1) condom in mouth (1) upper body (1) breath (1) pantyhose (1) orgy (1) phallic symbol (1) looking down (1) one eye covered (1) belt (1) leotard (1) pussy juice (1) finger to mouth (1) after sex (1) white leotard (1) cum string (1) cum in clothes (1) cum on tongue (1) hairclip (1) bra (1) mole under eye (1) black bra (1) meme (1) huge ass (1) single sidelock (1) twitching (1)

What does it do you ask? Ok, I'm all aboard the Honkai: Star Rail train, despite not liking Genshin Impact at all. So. I need to make some Lora's of my main team of March 7th, Himeko, and Natasha I guess. Plus Kafka because I will 100% be pulling her. I know there are already some Lora's out there of March 7th, but I wanted one that I felt comfortably worked with my Locon's as well. So you choose whichever one works best for you!

How to use it:

(I use Sampling Method DPM++ SDE Karras and Upscaler R-ESRGAN 4x+ Anime6B).

0.6 to 1 depending on model/lora/etc combination.

Main Triggers

Mrch7 - Trigger word to get her base outfit. Remove it if you want to change her stuff.

multicolored eyes, skirt - Can also be used in some models to make sure these appear consistently.

Then add whatever else you want! Take it all away if you want to change her outfits

Extra, Extra!: Should be well trained and usable in most models. Images shown for example.

Picture Descriptions:

1-2: AnythingV3 with just this Lora and simple prompts.

3-4: AnythingV3 combo'd with my in progress Good Cop/Bad Cop Locons.

5: AnythingV3 combo'd with my in progress Racer Queen Locon

6: AnythingV3 combo'd with my in progress Succubus Locon.

7: AnythingV3 combo'd with my in progress Bound Yuri Locon.

8: AnythingV3 combo'd with my in progress Broken and Defeated Locon.

9: AbyssOrange using just this Lora.

10: RevAnimated using just this Lora.

11: MeinaMix using just this Lora.

How was it trained?

It was trained on animefull_final with 23 images with the usual self-pruning and tagging and a wide range of different images. I like keeping my prompts simple for the first few pictures so that you can see it without a whole bunch of mumbo jumbo. Make your pictures better than mine though!

What's next for this?

I think it's fine? LMK.

What's next that isn't this?

More character Loras. (Rupee and Volume from Nikke, Spectre Tenebria from E7, Zephia from Fire Emblem Engage, Ivy from Soul Calibur 5, Heine from Combatants Will Be Dispatched, Conqueror Lilias from E7, Evelynn from LoL (both base and All Out version), Tharja from Fire Emblem, Laharl-Chan from Disgaea 5, Miku from Idaten Deities, Kuribayashi Shino from GATE, Rin Sohma from Fruits Basket, Kasumi from P5, Ann from P5, Ikumi Mito from Food Wars, Alice Nakiri from Food Wars, Lulu from FF10, Lucoa from Dragon Maid, Laharl-chan from Disgaea, Dominique de Sade from Vanitas, and Dos from One Punch Man.) And maybe Wilhelm van Astrea, my favorite Re:Zero husbando or a Prinny from Disgaea.

V2's of some of the previous Loras. Specifically make them more diversely acceptable with other models and LORA's.

I also have some more quite NSFW Locon's. I decided to list them as well, so if people have preferences on which to fix up first let me know. Specifically, Bound Yuri (two girls tied up, breast to breast and likely kissing), Sultry Succubus Transformation, Tied up with vibrators (basically the waifu tied up with vibrators and being overwhelmed with pleasure), Beta Censorship (Automatically censors any images you create to make them BetaSafe), being stepped on, Robotification, Defeated (decently NSFW defeat scenario for the waifu getting abused), Petrified (I'm unsure if I should bother with this one anymore), Infection (this one is VERY nsfw, as the waifu is completely infected), Dominatrix (waifu becoming a dominatrix), Tentacles, Pillory, Masked Evil (A mask that possess' the waifu), Latex Corruption, Combat Drone (hard to describe but basically making them a mindless sci-fi combat warrior), JAV Supervillain, Blacked, Hypnotized Harem, Hourglass Figure (just to make your waifu curvy), and Holding a Leash.

And, uh, I want to make a few SFW Locon's so I have a Delinquent (The delinquent girls in japanese schools specifically), GoodCop/BadCop (Locon with outfits for being a good and for being a bad cop), Maid (both normal and naughty variants), Biker Chick (ride a motorcycle and look like you should), Race Queen (Chiling sexily on a car), forest fairy, Anime Officer, barbarian.

Or just give me some random suggestion or priority opinions. That works too.

I've also added a couple of my extreme NSFW preview images to my twitter (https://twitter.com/TwoMoreTimes89). Let me know if you think I should work on them more for a submission here.

LoRA is generated entirely from anime images.
Should work well with any model.
NSFW works nicely too.

Features 10 outfits from anime.
Check samples for prompts :)

Leave a like, comment, review but first of all, ENJOY ;)

¿Looking for Jeongyeon nsfw version?

Important read

This model is a early version, Gradually I will add more pictures and updating the model until make it better and more complete

Hello here some settings that can help to get some good results

Recommended resolution: 512x768

• lora weight: 1.0 - 1.1

• Simpling method: DPM++ SDE Karras or Euler a

• Steps: 30

• Model: "Chilloutmix_NiPrunedFp32Fix" or "Chilloutmix"

• CFG: 7

You can use some trigger words like: twjeong

also you can use: "blonde hair", "brown hair" or "blue hair" to get some variations

Trained on 806 images of Rika Jougasaki from iDOLM@STER

A little hard on position, use weight of 1.1 to 1.3 if not corresponding

No official outfit for the moment, may come in future version

Main prompt:

RikaJou, long hair, blonde hair, green eyes, small breasts, fang,

Mdf_an - Artist Style 🎨

This Lora is trained on the calm artworks of Mdf_an, you can checkout his works on Twitter .

Usage Tips :

• regular use is at the weight of 1 to 0.7 ,

• adding mdf_an to the prompt will give the LoRA a tiny bit more influence.

• try experimenting with removing quality tags like (masterpiece ) to more of that sketching effect

Samples Settings :

• Checkpoints : AOM3 A3, anyhentai 19

• Steps : 24 to 26

• Sampler: DPM++ 2M Karras

• CFG : 7

Hinata.
trained and generated on https://civitai.com/models/7240?modelVersionId=20322 with vae.

TWICE Momo V2 is here

1. You can download my LoRAs that haven't been released here a few days earlier. There is also other my premium LoRAs and unreleased LoRA / LE SSERAFIM Chaewon / VIVIZ Eunha / Loona Chuu / Taeyeon , etc

2. The commission is open! You can request LoRA here

3. Membership open! If you want to buy my LoRAs at a 30%~50% discounted price and request LoRA making to me once a month, check here

092.568.6789

Monet from One Piece LoRA

Weight 0.5 - 0.9 for best generation.

For best harpy generation use prompt: "harpy wings", "harpy legs"

Also use trigger words for add glasses and human form

The Juice is in the description below

Tips to use:

• Use LoRA weight between 0.6-0.8

• To summon her use the trigger Tags plus following tags: solo, twintails, multicolored hair, blonde hair, blue eyes, makeup, breasts, looking at viewer, blue hair, smile, jewelry, lipstick, red hair

We used one of Meina's model to generate the images.

If you have any questions feel free to ask!

Hope you guys like it! ❤️

You can ask for character LoRAs on our patreon page. Consider supporting us. 😘

This is Mayuzumi Yukie, the easily flustered swordmaster from the visual novel series and anime Majikoi (Maji de Watashi ni Koi Shinasai).

Use about 0.7 weight.

Trained on AnyLora model using mostly official art.

I gave all the preview pictures a 1:1 "upscale" using R-ESRGAN 4x+ Anime6B, for sharper lines.

Note that the feet generated by this lora are not very clean and fragrant.

It uses a large number of [soles] images as a training set (1272), so it is mainly used [to generate poses showing soles] as well as [to optimise soles].

It depicts a [soles] that is [dirty, dark or red, wrinkled, oily].

Weight=0.8-1

Negative prompt=white background, simple background

I will also produce more hardcore content related to feet, armpits and diseases.

Ronald is a well-known character from MC Donald Resutaurant. The Lora works with a force of (0.6 - 0.5) use those numbers to get the best results. The Lora adapts perfectly to various styles and is compatible with various models, due to the small amount of Dataset that I could find the Lora works better at low weights. It can be used individually to create the female version of Ronald or it can be used to dress your favorite characters in restaurant clothes. To make it work try with (0.5). To use the Lora use the following Tags (yellow gloves, striped thighhighs, striped sleeves), (makeup). This Lora gonna get a update soon!!

Restaurant Series: Wendys

Don’t forget to check out my other Disney Series Loras on my profile. YeiYeiArt.

Or

THE NEW !! Villain Collection: Maleficent, Ursula.

ADVICE!! “If you want a commission of something in particular, whether it’s a style or a character on request, contact me via email: jeanc.romero.ib@gmail.com.”

Lora quite simple that allows to simplify the creation of blowjobs with several added tools and with its own style. There are three variations of lora, but all three do the same thing, the only thing that was changed was the database to create it, so they give different results. The main focus is the blowjob with pov, but it can be used without pov if you are looking for different results. The style is required, as you created it to follow the drawing style seen in the examples and it is not neutral. However, if the intensity of the lora is lowered enough and a style lora is applied to it, I estimate that it would be overwritten (I did not try it). Play tag and have fun.

• Cuasimodo (105) penis (103) breasts (96) 1boy (87) veins (87) veiny penis (87) 1girl (80) solo focus (79) large penis (78) pov (76) pubic hair (71) male pubic hair (71) erection (68) dark skin (67) oral (67) fellatio (64) cum (57) blush (56) cleavage (50) makeup (47) lipstick (47) huge breasts (46) long hair (45) ass (45) large breasts (44) dark-skinned male (43) testicles (42) group sex (42) cum in mouth (41) nude (39) tongue (39) brown hair (38) stray pubic hair (38) interracial (36) blue eyes (35) tongue out (35) threesome (32) black hair (31) looking at viewer (30) huge penis (26) multiple girls (25) 2girls (25) ffm threesome (25) mature female (25) :>= (25) ponytail (24) red hair (23) teamwork (23) nipples (23) dark-skinned female (23) clothed female nude male (22) paizuri (22) areola slip (22) cooperative paizuri (20) lips (20) gloves (20) open mouth (20) breast press (19) cooperative fellatio (19) short hair (19) green hair (19) saliva (18) foreskin (18) blonde hair (18) disembodied penis (18) after fellatio (18) multiple boys (17) multiple penises (17) hair bun (17) lipstick mark (17) underwear (17) jewelry (16) single hair bun (16) penis on face (16) dark nipples (16) bare shoulders (15) panties (15) uniform (15) colored skin (15) green skin (15) green lips (15) licking penis (14) dress (14) hat (14) vest (14) baseball cap (14) black vest (14) covered nipples (14) thong (14) gangbang (13) thighhighs (13) multicolored hair (13) two-tone hair (13) precum (13) eyeshadow (13) brown eyes (13) completely nude (12) shirt (12) open clothes (12) purple panties (12) hair between eyes (11) cum in nose (11) trembling (11) facial (11) licking (10) white hair (10) white gloves (10) collarbone (10) wristband (9) torn clothes (9) purple lips (9) scar (8) earrings (8) red dress (8) hair ornament (8) cum on tongue (8) elbow gloves (8) gradient hair (8) twintails (8) handjob (8) looking at penis (8) pink dress (7) visor cap (7) symmetrical docking (7) thick eyebrows (7) 6+boys (7) on bed (7) sweat (7) simple background (7) white background (7) jacket (7) necklace (7) mole (7) orange eyes (7) half-closed eyes (7) sleeveless (7) dark penis (7) penis awe (7) mmf threesome (7) suit (7) 4boys (6) red eyes (6) animal ears (6) swimsuit (6) bikini (6) rabbit ears (6) animal print (6) leopard print (6) necktie (6) streaked hair (6) ejaculation (6) leotard (6) 3boys (6) military uniform (6) teeth (5) sitting (5) purple eyes (5) tears (5) mole under mouth (5) bodysuit (5) school uniform (5) freckles (5) wings (5) pointy ears (5) fairy wings (5) fairy (5) long tongue (5) green nails (5) dark areolae (5) cum string (4) rabbit girl (4) long eyelashes (4) nail polish (4) short twintails (4) very dark skin (4) red necktie (4) muscular (4) japanese clothes (4) orange hair (4) tattoo (4) seiza (4) blue kimono (4) shoulder tattoo (4) rolling eyes (4) long sleeves (4) hairband (4) puffy sleeves (4) clothing cutout (4) black hairband (4) all fours (4) white leotard (4) blindfold (4) black blindfold (4) thong leotard (4) double handjob (4) scar on face (4) heart (4) censored (4) 2boys (4) strapless (4) nipple slip (4) closed eyes (3) kneeling (3) bed sheet (3) red nails (3) smile (3) black thighhighs (3) kimono (3) indoors (3) from above (3) juliet sleeves (3) covered eyes (3) feather-trimmed sleeves (3) o-ring (3) poke ball print (3) green eyes (3) short sleeves (3) collared shirt (3) closed mouth (3) fingernails (3) hand on another's head (3) irrumatio (3) head grab (3) hood (3) piercing (3) deepthroat (3) spider web print (3) hooded bodysuit (3) 5boys (3) bangs (3) gloved handjob (3) thighs (3) tan (2) tanlines (2) bow (2) bowtie (2) glasses (2) abs (2) sleeveless kimono (2) bed (2) boots (2) thigh boots (2) cleavage cutout (2) star (symbol) (2) colored sclera (2) superhero (2) purple leotard (2) impossible leotard (2) spoken heart (2) imminent fellatio (2) pink lips (2) hood down (2) overflow (2) eyebrow piercing (2) torn bodysuit (2) minigirl (2) thick thighs (2) tongue piercing (2) speed lines (2) military (2) sidelocks (1) shorts (1) spread legs (1) bulge (1) erection under clothes (1) red bow (1) black nails (1) solo (1) covered navel (1) semi-rimless eyewear (1) hand on own face (1) futanari (1) arm under breasts (1) phimosis (1) condom (1) lace-trimmed legwear (1) cheek bulge (1) licking lips (1) headband (1) motion lines (1) motion blur (1) green sclera (1) nose bubble (1) ! (1) spoken exclamation mark (1) eyeliner (1) purple eyeshadow (1) pink nails (1) sex (1) vaginal (1) hoodie (1) cum in pussy (1) sex from behind (1) clothed sex (1) rape (1) doggystyle (1) emphasis lines (1) double penetration (1) x-ray (1) torso grab (1)

Trained on 55 images of Shinomiya Runa

Recommended settings : lora weight of 0.6 to 0.9

Main prompt :

ShinoRu, 1girl, solo, short hair, purple hair, yellow eyes, small breasts

Young Rena Tanaka

Only work best with henmixrealv10 , set 0.7-0.9

https://civitai.com/models/28673/henmixrealv10

Not really works well for chilloutmix or others~ you may try it yourself.
* But use controlnet if needed.

1st epoch works very well but too boring~

Sunny Leone

Sunny is a Canadian-Indian-American former pornographic actress now working as an actress and model in the Indian and American film industries. So I guess it would be correct to say that she left us too soon ;)

This is a LoRA, if you want a textual inversion of Sunny look HERE

Like my work? Buy me a beer!

Lora based on Kamen Rider Blade "Undead Green Blood" artbook by Yasushi Nirasawa. Made with Kohya Trainer.

• Lora Weight 0.6 to 0.8, CFG 5 seems best, up to 11.

• A Prompt exemple:

  Try to start and end with : (((concept art, drawing, sketch, Undd, solo))) (put here type, color), (((spikes))), belt, (((leather))), (((ring chains))) (((asymmetrical))), <lora:Undead Nirasawa II:0.*>

• You can experiment with many words like : animal print, antennae, armor, axe, fangs, claws, knife, sword, cyberpunk, dreadlocks, helmet, horns, spikes, leather, ring chain, chains, skull, tentacles, wings, feathers.

  Tiger, snake, bat, spider, peacock, lizard, dragon, bug, alien, animal, demon, robot, creature.

• Trained on Clip 2.

• Undead II is finetunned. I used automatic captions.

• Undead III was trained one more time on their actual specific names, instead of just automatic. It's kinda capricious. Results are not guaranteed.

  Names you can try to use : bat - beetle - wildboar - buffalo - capricorn - centipede - dragonfly - eagle - elephant - jaguar - jelly - chalice - lion - lizard - locust - mold - mole - moth - orchid - paladoxa - peacock - pecker - ivy - scarab - serpent - turban - spider - squid - tarantula - tiger - tortoise - trilobite - wolf

It also goes quite well with my other Blame! Lora !

Enjoy and share your results !!

Murasaki Shion (Hololive) LoRA

Thank you for deleting my previous LoRA without any warning (I discovered it delection 1 month later) because "Hololive" and "gothic_l*lita" contain a banned word and excluding the domain from wayback archive means that I can't copy paste my old LoRA description (this is annoying for recommended prompts)

01-05-2023/V2 Update:

• Reduce the model file size to 9MB.

• Use a way lower network dim (8 instead of 128) and Alpha (1 instead of 128) to prevent overfitting

• More flexible

• Rebalanced dataset (I uploaded it)

• More accurate official swimsuit

• Add a new token for each canon outfit (using booru token was a bad idea):
  ShionDefault

  ShionSwimsuit

  ShionGothic

V1 is a reupload of my prevous LoRA

This is a LoRA, this mean that you need to update Voldy/AUTOMATIC1111's webui to use it (use git pull) .

Weight:

You won't need to use a lower weight unless you want to use multiple LoRAs

Recommended prompt:

That LoRA can reproduice some of her models

For V2:

Default outfit:

black gloves,single hair bun, witch hat, vertical-striped_shirt,pinstripe_shirt,midriff, crop top,miniskirt, ShionDefault

Gothic L*lita :

twintails, animal_ear_fluff,gothic_lolita,lolita_fashion,halter dress, cat ears,cat_girl,long dress, halterneck, layered dress,frilled dress, center_frills,detached sleeves, ShionGothic

Official Swimsuit :

bikini, bikini_skirt, black_bikini,criss-cross halter, swimsuit, ShionSwimsuit

For V1:

Same thing as V2 but without ShionDefault,ShionSwimsuit and ShionGothic

Model used for the example:

V1 examples : AOM 2 and Pastel Mix (model version) .

V2 examples : AnythingV4.5

trained on danbooru drawings of beautiful women by anabone

all examples using trigger only with and without EasyNegative on the same seeds

https://huggingface.co/Powidl43

Luna by Zummeng Lora. This is a request from my lora request form. I recommend using 0.8~1.0 strength.

Helpful Tags:
General: furry, blue skin, antlers,
Glowing: glowing skin, glowing hair, glowing clothes, glowing eyes, glowing

Trigger word: luna_(zummeng)

I created a Lora request form. Please do not submit anything that breaks the tos. Thx!

对于个人来说，发布这个LoRA是个很难的决定，因为涉及到的版权问题和潜在纠纷是在令人咂舌。

所以，我的这个作品只在C站供大家私人学习使用，任何违反法律的，商业的行为都是被禁止的。

C站独家的LoRA，如果出现任何版权纠纷，我都会立刻删除下架。

For myself, releasing this LoRA is a difficult decision, because the copyright issues and potential disputes involved are staggering.

Therefore, this work of mine is only available at Civitai for your Private Study and use. Any violation of the law or commercial behavior is prohibited.

The exclusive LoRA at here, if there is any copyright dispute, I will delete it immediately.

欢迎关注，发现更多好用的模组。

Welcome to Follow and discover more useful mods.

1. 炭治郎tanzhilang/Tanjiro Kamado/かまどたんじろう

2. 吾妻善逸wuqishanyi/Zenitsu Agatsuma/あがつま ぜんいつ

3. 炼狱杏寿郎/Rengoku Kyoujurou/煉獄 れんごく 杏寿郎

4. 嘴平伊之助/Inosuke Hashibira/はしびらいのすけ

5. 蝴蝶忍/Kochou Shinobu/こちょう しのぶ

Actress

the training was on a model - CyberRealistic 2.0.

I'm Render on - EndlessReality

I do not recommend using my LoRA with model RealisticVision

personally, I most often perform generation with parameters: DPM ++ 2M Karras, Sampling Steps 35, Denoising strength 0.35, CFG Scale 3.5-4.5

evaluation of the model is welcome

I made this model because I like the character Boa Dararat who is a celebrity, and because no one has made it yet I took the initiative to make the LORA. hope it helps and you like the LORA. and the word to activate LORA is (boa_dararat)

if you want to see the selebgram search on instagram "Boa_Dararat"

folow my instagram "Katar_Legacy"

This Lora is gravelyn from adventure quest world trained by the Meinahentai model

A LORA model trained on David Oberto.

To use place the words "davidoberto person" in the prompt.

Trained on 6240 steps.

Model: chilloutmix_NiPrunedFp32Fix

Res: 512 x 832

DO NOT check "Restore faces"

マクロス７よりミレーヌです。

影響度を変えても崩れる場合img2imgに移動させてもう一度上乗せするとキレイになります。

This is a "toothbrushing blowjob" combined with cheek poking.

You can combine with other loras too,

Model: chilloutmix_NiPrunedFp32Fix

Res: 512 x 832

DO NOT check "Restore faces"

Singer

the training was on a model - CyberRealistic 2.0.

I'm Render on - EndlessReality

I do not recommend using my LoRA with model RealisticVision

personally, I most often perform generation with parameters: DPM ++ 2M Karras, Sampling Steps 35, Denoising strength 0.35, CFG Scale 3.5-4.5

evaluation of the model is welcome

If you'd like to support me and do more:

https://ko-fi.com/edgonkofi

Paladin Fashion

We are warriors, born from the light
An army for freedom, defenders of life

At night, high up in the heavens we fight
Faster than lightning we strike
Like fires that rip through the night
Surrounded by light

Raging thunder in the skies
Time has come to sacrifice

Those of you that recognize the lyrics of this song and are emboldened by its lyrics, welcome.

Those of you that do not, fear not brothers you will be put up to speed quite easily.

It's time to make your inner light SHINE!

Download this LoRA, for the light!

Trigger words

Paladin, paladin_glam, paladin armor

Use them!

Cindry from One Piece LoRA

Weight 0.6 - 0.8 for best generation.

( 680 x 1024 )

This is the result of messing around with merging a couple loras, thought it was good enough to share. (one of the Loras is the John Kafka Hard lora https://civitai.com/models/7090?modelVersionId=8335 cannot find the source for the other Loras)

Follows prompt fairly well and can be used with any weight (recommended 0.6-1.0) and can be used with other Loras

v3 : Train with new method, react better whtn prompts.

v2 : Uses only SFW pic to train, use better captioner

SAGE (alt. adult ver.) from Sonic Frontiers

Uses sageai to summon her, use red eyes, short hair, hair over one eye, white hair, red hair, multicolored hair to set her head, uses black dress to give her signature dress.

Disclaimer: The characters were depicted as 18+ years old and the training data used to make LoRA is an adultified version of SAGE.

Get your balls out for Saejima.

Based on his appearance in the Dragon Engine games.

I recommend .6 strength, but you should change it depending on what you're going for.

I know there are other models, but I tried making my own version.

Please do not prompt ahegao. Use one or more trigger words: ahg, rolling eyes, tongue.

activation prompt: shimokura megu

recommended prompt: shimokura megu, long hair, blue eyes, pointy ears, horns, halo, goggles on head, red hair, breast

best perfomance at 0.6~0.8 weight

Trained on YesMix 1.6

自己随便跑的一个 细节不到位，希望可以抛砖引玉

Trigger word: honokabkn + bikini

If you want her to wear other clothes or…… be nsfw, just do not input bikini.

如果你想让她穿别的衣服或是……nsfw，就别输入bikini了。

"one side up" might be needed at low weight.

低权重时可能需要"one side up"。

Well, I've discovered that some models may add "open clothes" to bikini.

我发现一些模型可能会给比基尼加一件薄外套。

LoRA for Kuvira from The Legend of Korra. Trained on AnythingV5 with both sfw and nsfw images.

55i*15r*10e/4b = 2063s

Used Hi-Res Fix (Latent) 2x —> Extras (4x-UltraSharp)

Recommended weight between 0.5-0.7; can go higher or lower depending on additional LoRAs, etc.

• basic outfit trigger: green metal outfit

• (optional prompts): braided ponytail, unkempt

Gives dragon scales and spikes on things.

Created as described here: https://civitai.com/models/52697/tutorial-konyconi-style-lora without any changes.


Perpetual coffee request: I signed up for Buy-Me-a-Coffee and Ko-fi, mainly because I thought it would be hilarious if someone as artistically untalented as myself received a coin for something creative. It's already happened, and it made my day. AI made the impossible possible :D. If you'd like to add to the amusement, please consider buying me more coffee through the following links:

https://www.buymeacoffee.com/konyconiN

https://ko-fi.com/konyconi7992

Mythmaker Irelia Lora

Tip word: MythmakerIrelia

Weight:0.7-0.8

other words: multicolored hair, hair ornament, facial mark, white hair, green eyes, bangs, jewelry, gradient hair,

My Pivix: KitteyLL - pixiv

You can use this mod to do anything you like

This mod is very good for semi-realistic style

Enjoy it！

LoRA for pulling back bikini tops.

Trained for realistic models, but can work with anime models as well.

Recommended weights: 0.4 - 0.7

If you like this LoRA and wish to support me, consider buying me a beer at https://www.buymeacoffee.com/WaifuWizardWong 🍺 Your support is much appreciated!

Be sure to check out my other LoRAs!

Fiore Brunelli is character from game Star Ocean 5

Best weight is 0.8-1 for ingame costume

For other clothes best is 0.6

Sometimes you need to lower the weight, especially in the case of other clothes

argyle or checkered pattern will not be perfectly generated due to how Stabble Diffusion works

Works with other anime models

Created by GeomanticArts

He's here. And he's hottttt. ! ~

Why did they decide to make him sexy? Who knows. I know. You know. Make some Ganondorf.

Trained mostly on his sexy new look, but has many examples of older versions of him as well in the training data. Tends towards the new look unless you prompt the older features. His little itty bitty pupils don't come out very often.

Example images all made with AOM2-7thV3_C_40 at weight 1

'Ganondorf, male focus, muscular, sarong' - Example prompt

LoRA metadata has all the tags used.

My (nsfw) Twitter

Trained on NAI.

Model: chilloutmix_NiPrunedFp32Fix

Res: 512 x 832

DO NOT check "Restore faces"

秦岚, one of my favourite actress. She is so graceful and elegant.

Trained with majicmix as base model, this Lora is able to generate beautiful portrait of 秦岚.

How to use LoRA:

• Copy the file to stable-diffusion-webui\extensions\sd-webui-additional-networks\models\lora

• Enable Additional Networks and select lan.safetensors as your lora

• No need to change the weight (use default value 1)

If you like my work and want to buy me a coffe :')

(updated link to ko-fi)

This is a LoRa with the style of DustyRayPaints.

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.

prone bone concepts

trained on:

• penis on ass

• vaginal

• anal

LoRa created with 100 images of full body (with tw vtuber-fullbody) and upper body vtuber models (with tw vtuber-halfbody).

Add full body or upper body at the end of your promt if you have problem with the result.

White background or simple background can sometime work but it tend to not change anything. If you want to have the character without the background, I suggest you to use an AI crop tool like Segment Anything from Meta AI.

To have better quality images, use hires fix. and upscalers (especially R-ESRGAN 4x+ Anime6B)

If you manage to have some great results or if you want to suggest some improvement for the model, you can send me feedback in the comment.

This is a LoRA for generating images of Macross Delta character Mikumo Guynemer.

Making models takes a lot of time and hard work. If you like this LoRA, please click the like button!

Recommended weights:

• Recommended weights: 0.5~0.7

  (Note: If the weight is below 0.4, the LoRA may not work effectively.)

這是一個用於生成超時空要塞Δ角色美雲吉內梅
的圖像的LoRA

製作模型需要花費很多的時間和精力。如果您喜歡這個LoRA，請幫忙按一個“讚”！

建議權重：

• 建議權重：0.5〜0.7

  （注意：如果權重低於0.4，則LoRA可能無法生效。）

I don't get any from my training but If you want to support my work. check patreon.com/goofyai

also my new merge model available https://civitai.com/models/53761/goofball-mix

all my Loras are available for free use at https://pixai.art/@goofy-ai

Lora weight of 0.7-1 works really good.

check the preview images for prompt ideas for both outfit

highres fix or mg2img at higher resolution is highly recommended.

drop tour review. it is much appreciated.

我发簪呢（恼）？
因为我是懒狗，所以训练集没优化好😆，等我什么时候想优化了再说罢！
推荐配合分层使用，
我使用的预设是：1,1,1,1,1,0,0,0,1,1,1,1,1,1,1,0,0

触发词：lingguang,china dress,(hair stick:1.2),
建议权重：0.7~0.9

Where did my hair stick go?

I was too lazy, so I didn't optimize the training set, and because the training set was relatively small, the training effect was not very good.😊

It is recommended to use with LBW,

I use the default is: 1,1,1,1,1,0,0,0,1,1,1,1,1,1,1,0,0

trigger:lingguang,china dress,(hair stick:1.2),

Weight 0.7~0.9

Princess Elise III of Soleanna from Sonic The Hedgehog (2006)

use princess_elise to summon her, use short hair, parted hair, red hair, blue eyes, hair ornament to form her head, use necklace, white dress, long dress to summon her dress.

Still trying to find the way to capture more detail, as far as I generate, it can't get her X shape pattern on her dress.

从现在起，列家人的lora齐了！（震声）是时候去炼其他人物去了yeah

权重大概在0.6到1，0.8也行，我一直用1权重感觉也没啥问题desu

也是在NAI上训练的，不过这次图片量比较少大概39张，

所以随便玩玩就行，可能也不是很还原？我不知道，我现在一心想要去学latent插件玩多人图和controlnet1.1，以及模型融合什么的....溜了溜了

trigger：constellation

Recommended weight is around 0.8, beware that his hair goes a little crazy at times

It's an original character i made with SD and now I am training the model of her

Note: Do not use a lot of Negative Textual Embeddings as It will effect the outcome.

I'm making a Discord Server. Join if you want - https://discord.gg/3mGXrS87cE
Also, I need helpers to make Free Requested Loras / My Idea's Loras. Join Discord if you wanna help. this will be not paid as I don't earn anything from it or from anywhere.

Please Review and Like this Model if you appreciate it. The least you can do to keep me motivated.

Also I'll appreciate your support on my Patreon and kofi.

Here is a Form you can request me Lora there (for Free too)

https://forms.gle/h9iuLCCnaEs6W9ZB8

A simple LoRA that increases breast size, strength should be between 0.3-0.7 (depends on model, start with 0.5 and experiment)

Can also enhance normal "big/large/huge breasts" inside the prompt in addition to increased word weighting.

Tested it with some 2d, 2.5d and 3d/realistic models and seems to work well with everything I throw at it. See examples. (but it's not perfect, especially with higher strength and realistic models it tends to over tune)

All the content I make is and will be completely free on CivitAI.
But if you want to support me with a coin, then you can do it here.

Trained on AbyssOrangeMix2 with Clip Skip 2.

Sample images was generated with Luma and AOM2

Recommended embeddings:
1. easynegative
2. bad-hands-5
3. badhandv4

Strength 0.6-0.8 is optimal.

PLEASE CHECK YOUR LORA NAME AFTER DOWNLOAD.

This lora works well without any trigger word, but you may use slavic style to increase strenght of it.

Also you may use tags sepia, border or monochrome, border to make it look like book illustration.

Clothes tags: kokoshnik headwear, white dress with ornate red frills, traditional clothes, circlet and etc. You may use any of these in any combination.

LORA by hana makihatayama in magical doremi, learning through official and fan art.

Trigger: 'makihatayama_hana'.

It is also possible to change clothes by using the 'hanacos01' trigger for everyday clothes and the 'hanawitch' trigger for witch clothes.

Casual Trigger: 'hanacos01'.

Witch Trigger: 'hanawitch'.

Weight: Between 0.7 / 1.

Example input:(casual)
(little girl), (masterpiece), highest quality, ((high detail:1.2)), (HDR), cel anime, (cinematic light:1.1), highres, (makihatayama_hana), 1girl, (full body), blonde hair, long hair, twintails, hair intakes, brown eyes, dot nose, (hanacos01), (blue shorts), (white shirt), (orange shoes), blue sky, park, <lora:hana_makihatayama_v2_5:0.8>

Example input:(witch)
1girl, solo, full body, cel anime, <lora:hana_makihatayama_v2_5:0.8>, (makihatayama_hana), blonde hair, long hair, hair intakes, brown eyes, dot nose, (hanawitch:2), white gloves, compact ,white dress, (yellow frills), puffy short sleeves ,white skirt, white boots, magical girl, hat, smile,

Recommended Strength 0.6~1

Success Rate isn't so good but you can get great images using t2t->i2i->inpaint features.

Serika Ito Lora from Uchuu Kyoudai/Space Brothers

I don't know if it is well made as it is my first LORA, but it seems to be capable with textual inversions. Rev Animated is used to plot examples.

⬇️ click show more for plots and examples

Activation

<lora:ItoSerika:0.9>

Depending distance, works well around 0.8~1

Suggested prompt :

1girl, brown short hair

Clothes

Dataset contained a few clothes description :

• casual clothes

• white suit (Jaxa outfit)

• Blue suit (Nasa outfit)

• Military suit

lowres plots :

Works well with astronaut suit on https://civitai.com/models/7371/rev-animated (for example)

Framing

Trained with some framing description with the following angles (warning : there is more close-up to medium-shot images than full body):

• close-up

• portrait

• medium shot

• cowboy shot (works poorly but I decided to let it go in reference to Mutta Namba "Houston" style 🤠 Try "cowboy hat" in negative prompt)

• full body

lowres plot :

Other recurrent stuff appearing in training set labels :

• light smile

• uneasy

• serious face

.. here's some expressions :

Stormi Maya

Stormi is a bombshell American model. She is, just, wow. Yeah.

(If you'd rather have a Textual Inversion, I got you covered.)

Do you like my work? Buy me a beer!

LoRA based on 30 high-def images, 1500 total steps, SDv1.5

Model: chilloutmix_NiPrunedFp32Fix

Res: 512 x 832

DO NOT check "Restore faces"

Click "show more" to read the English Introduction.

这个lora可以生成很漂亮的欧美小姐姐，建议底模使用chilloutmix

CFG Scale: 4~6.5

lora权重: 0.55~0.8 (权重越高越像欧美人，但我觉得0.55或0.6是最好看的)

其他参数在例图里都有

建议用高分重绘生成图像，参考如下

This lora can generate very beautiful white girls pics with chilloutmix

CFG Scale: 4~6.5

lora's strenth: 0.55~0.8 (The higher the strenth you set, the more she looks like a white girl, but I think 0.55 or 0.6 is the best)

All parameters you need are in the examples, including the generation data

Recommend using Hires.fix to generate images, the following pic is for reference

This is a New Style Lora I have created, it is suitable for cartoon styles.

I will suggest using 0.6 to 0.8 weight for this Lora. Using weight above 0.8 can look more like the original art however this can cause the hands to go wonky and messy.

In the example images I used the same prompt and experiment with different Models.

This Lora is more stylized therefore the hand part may need additional fixes or extra care.

Model: Most known models

Prompts: (hand well), (finger well), (long braided hair), (bangs), smile, with in frame, ultra-detailed, 8k, masterpiece, beautiful detailed face, smile, no names, (volumetric lighting), beautiful detailed eyes, (ambient light), realistic shadows, (fullbody:1), (1girl), (cowboy shot), shirts, pants, looking at viewer, perfect body, body well:1, leg well:1, arm well:1 , (GoodhandsV3)

Neg: (easynegative), (hands poor), (fingers poor:1), (badquality), (bad anatomy), (nsfw), (legs poor), (inaccurate limb:1), (bad-image-v2-39000), (2girls), bad composition, inaccurate eyes, (extra hands:1.2), (inaccurate limb:1.2), (extra arms:1.2), (bad quality), bad face, (bad eyes), deformed, ((deformed fingers:1.2)), ((extra fingers, extra limbs)), bad hands, bad fingers

Steps: 31, Sampler: Euler a, CFG scale: 9, Size: 512x768, Denoising strength: 0.37, Clip skip: 2

Jenna Ortega

What!? No LoRA of Jenna Ortega? Don't worry, I'm here to help.

Jenna is a fabulous actress who absolutely KILLED on Hot Ones, go watch it after you download

I used 30 hi-res photos, 50 steps each, against v1.5

Flat illustration with Colorful style

It is based on flat illustration model, 彩线2 model, chiloutmix model.

In order to make the flat illustration (in the style of james gilleard, abstraction-création, playful streamlined forms, wealthy portraiture, andrew hem, matti suuronen, playful animation).

However, I find it cannot understanding the abnormal human body structure, except img2img channel. Next version will be trained.

Alicia Silverstone

Actress and mother Alicia Silverstone was in the move Clueless and some music video things back in the 90s I believe.

This is the LoRA counterpart to my embedding.

The results, as they say, speak for themselves

This works best on 0.6-0.8 strength and the consistent-Factor model.

Do you like my work? Buy me a beer!

(all images used to train this model are of Ms. Silverstone 18-years old and older)

我的製作只作個人用途，請勿未經我的許可轉載到其他討論區或作盈利用途!

If you like our model, please buy me a coffee and supporting for future project.
https://ko-fi.com/apsrcevo
Please give us a 5 star rating.
Commercial customization LoRA project are welcome.

After download this model, try to share with us about your finding.
Again, not for commercial use and she is not a existing person.

Please use ChilloutMix and based on SD1.5

• 768,768 image

• Use DPM++ 2M Karras or DPM++ SDE Karras

• Use 18 sampling steps

• Strength: 0.7-0.8

• Hires. fix: on

• Hires step: 18

• Sampling step: 18-20

• Denoising: 0.7

1、此Lora包含了魔兽世界大酋长萨尔的信息，特别是盔甲细节；

2、生成大小建议设置为512×512，过大的尺寸会导致多腿多手的情况；

3、为了提高出图精度可以打开高分辨率修复，但重绘幅度 (Denoising strength)需要降低，具体数值推荐在0.5左右，以测试为准。

4、配合aZovyaRPGArtistTools_v2模型，可以生成具有写实感的美式漫画风格。

5、欢迎与我分享制作成果。

This is a LoRA for generating images of Macross Delta character Makina Nakajima.

Making models takes a lot of time and hard work. If you like this LoRA, please click the like button!

Recommended weights:

• Recommended weights: 0.5~0.7

  (Note: If the weight is below 0.4, the LoRA may not work effectively.)

這是一個用於生成超時空要塞Δ角色瑪基娜中島
的圖像的LoRA

製作模型需要花費很多的時間和精力。如果您喜歡這個LoRA，請幫忙按一個“讚”！

建議權重：

• 建議權重：0.5〜0.7

  （注意：如果權重低於0.4，則LoRA可能無法生效。）

Rebecca from One Piece LoRA

Weight 0.5 - 0.9 for best generation.

This is a lora model of the Honkai Rail character——桑博/Sampo。

权重一般用0.6比较好，想比较自由的0.5也不错，请务必配合使用controlnet，解锁更多姿势。

！请不要使用该模型进行商业用途

！Please do not use this model for commercial purposes

Contacts：

QQ交流群

专注练男人！

LoRA is generated entirely from anime images.
Should work well with any model.
NSFW works nicely too.

Check samples for prompts :)

Leave a like, comment, review but first of all, ENJOY ;)

I hate Tamase. Getting the hair right depends on model and luck. I guess you can inpaint or something. Not really sure what would I have to do to get the AI to learn that cursed hairstyle and ribbon, maybe lower learning rate would help, but I really don't care about her enough to try. Perhaps someday. As for actually using it, maybe resizing the LoRA will help if you're running into issues? Well, up to those who care enough.

Hello again my friends!

Today we have Minako, warm milf from hentai "Ano Ko no Kawari ni Suki na Dake" who flirted with her daugther's husband.

This is NSFW model

LoRA was trained on Animefull-pruned.cpkt due to best compitability with all anime checkpoint models.

89 arts were used with 8 repeats and 6 epoches. 2403 steps with batch size 2). It is a 5th epoch but because I suppose 6 is little overfit.

0.75-0.85 strength of Lora is the best radius. I prefet 0.85 as you can see in preview.

Some necessary tokens for better experience:

Default body:

Minako, mature female, 1girl, milf, cleavage, short hair, black hair, purple hair, huge breasts, puffy breasts, lipstick, big ass, earrings.

For her casual mommy clothes:

(pink tank top:1.1), tank top, yoga pants, tight clothes, big ass, bare shoulders (optional).

For cheerleader outfit:

cheerleader, stars, cleavage, bare shoulders, clothing cutout, pleated skirt, zettai ryouiki, white thighhighs, shirt, lips, cleavage cutout, huge breasts, covered breasts

If you like milf's pubic hair use these tokens:

female pubic hair, pubic hair.

Sometimes it need little changes with weight (pink tank top for excluding purple, small anus, more weight about stars on tits).

Also LoRA was trained with some arts including nsfw poses. You can use it but in many times scenes need to be repaired in img2img and photoshop.

Advices for better experience:

·         use highres.fix (If you are sure that your prompt is easy for SD); I prefer x4_Fatal_anime_500000_G, lollypop and animesharp.

use extention "dddetailer" (very useful extension for faces. You forget about face inpainting with it).

·         use SD Upscale script in img2img (For me 1.25-1.5x with AnimeSharp, UltraSharp or fatal anime is good on 0.2-0.3 denoising strengh);

·         Use inpaint and sketch for other details;

·         Use Photoshop.

I will be glad to receive feedback from you and a demonstration of your art with my LoRA!

Other works.

Neon Genesis Evangelion 1990s Anime style LoRA

Making models can be expensive. Do you like what I do? Consider supporting me on Patreon 🅿️ or feel free to buy me a coffee ☕

One of the best anime/manga ever made. Use this with AnyLoRA and similar models at around weight 1. Increase the fidelity with the trigger evangelion anime style. Could also help adding anime screencap, 1990s_\(style\) to get more flat colors and 1990s anime noise.

I also suggest you select "None" as the upscaler for highres.fix.
I posted examples using both "None" and R-ERSGAN AnimeVideo, so you can see the difference. The latter is too much "clean" and loses the noise effect.

Keep in mind this is not a character LoRA, so while it can imitate characters from their descriptions (as shown) you might need additional character LoRA/embedding to increase the accuracy. The purpose of this LoRA is to make any character with this style, not only Persona characters.

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)

Fishnet skindentation 东坡肉（渔网勒肉）

V1.1Slightly enhance the effect and improve the adaptability to real people a little.

V1.1略微提升效果，提高一点对真人的适应性

////

Test the effect of regularization.

This lora is possible to ignore the character's original body shape and show the skindentation  of leg. The effect is even more obvious for slim characters.

Weight 0.3-0.7.

thighhighs can be replaced by boots, thigh straps, garter straps

Recommend using head out of frame in negative prompt

测试一下正则化的效果。

本lora可以无视角色原本的体型，展现出腿部勒肉的效果。对痩的角色效果更加明显。

权重0.3-0.7.

thighhighs可以换成boots,thigh strap,garter straps

推荐在negative prompt 使用head out of frame

Gordon Freeman in Half Life

When 3...?

Otose Kotama, our cute wiretapper from Blue Archive

Use otose_kotama to sumon her. Use halo, blonde hair, long hair, bang, hair ornament, side braid, purple eyes, glasses, choker to form her head, Use jacket, open clothes, white shirt, necktie, white skirt, short skirt, belt, boots for her costume, take away keyword you do not want to appear.

So I was not too impressed with the already exisisting Loras of her, so I decided to make my own. Gathering training images was a little pain since despite the large number of images with her, a lot of them are crayon, pencil, lowres etc. So I had to do a lot of img2img and retouching to get usable images for the training.

This is trained on the Novel AI model (animefull-final-pruned), so it should produce good results on models which are based on this model.

For her default attire/look also include the following trigger words in addition to the main trigger word: long black dress, fur-trimmed dress, cat tail, facial mark, whisker markings and feather earrings. Whenever you want to add the tail, you can also add miqo'te to make sure that you don't get 2 tails and stuff like this. Making her hold a staff it's a hit or miss, you need to experiment.

It can also do NSFW ofcourse, as you will see in the examples provided.

Don't forget to use Clip Skip 2 when using the Lora.

Checkpoints used in the examples:

Corneos7thHeavenMix_v2

CounterfeitV25_25

AnythingV3

Abyssorangemix2NSFW

EpicMixAnimeRealism

Anyhentai_19

Kotosmix

GrapefruitHentaiModel

Abyssorangemix2_Hard

This is a LoRA for generating images of Macross Delta character Kaname Buccaneer.

Making models takes a lot of time and hard work. If you like this LoRA, please click the like button!

Recommended weights:

• Recommended weights: 0.5~0.7

  (Note: If the weight is below 0.4, the LoRA may not work effectively.)

這是一個用於生成超時空要塞Δ角色要巴卡妮雅
的圖像的LoRA

製作模型需要花費很多的時間和精力。如果您喜歡這個LoRA，請幫忙按一個“讚”！

建議權重：

• 建議權重：0.5〜0.7

  （注意：如果權重低於0.4，則LoRA可能無法生效。）

Nana Shimura from My Hero Academia - Character LoRA

If you like what I create, feel free to support on Ko-Fi!🍵

Make sure to leave a review if you like it, would love to see all the pictures everyone creates!

Pro Hero from My Hero Academia, Nana Shimura! Trained on the NovelAI model, and tested with many different anime specific models such as AOM3, AOM2_Hard, NovelAI, Anything v4.5, mistoonAnime, Universal Solvent Mix, Based64Mixv-V3, and a few more. This LoRA definitely can produce images of Nana Shimura in her Hero outfit very well. However, you can also feel free to generate her in other outfits, such as a tank top and shorts in one of my example images. I found that this LoRA works well with a weight of 0.9 - 1. I have generally found a weight of 1 works perfectly fine in all circumstances. Have found a CFG between 7-9 works well, along with Step around 25-30, but feel free to experiment! Also Clip Skip of 2.

• Triggers are: NanaShimura and boku no hero academia

• Description Tags that may help: black hair, blue eyes, hair between eyes, sleeveless, red belt, white cape, yellow gloves, whole bodysuit, yellow footwear
  

• Some Negatives that may help: EasyNegative, badhandv4, extra digit, fewer digits, (extra arms:1.2), bad hands, ((badhandv4)), watermark, signature, text, patreon username, username

这是一个做中国古代国风女子的lora，

把权重设置在0.65-0.95之间，看着调吧；

不同大模型出来的效果不一样，可以多尝试；

低模用的是3Guofeng3_v33，尺寸是516X768；

描述文案不用太复杂，类似写一个女孩，穿什么颜色的裙子，在干什么，环境有什么；

遗憾的是这个lora会出现下半身容易错误，手指畸形，眼睛有些奇怪这些毛病；

The following is a machine translation:

This is a lora that does the ancient Chinese state style of women.

set the weights between 0.65 and 0.95, and watch the adjustments;

Different large models come out differently, try more;

The low model used is 3Guofeng3_v33, size 516x768;

the description text should not be too complicated, similar to writing about a girl, what colour dress she is wearing, what she is doing and what the environment has;

regrettably this lora can suffer from the faults of being prone to errors in the lower half of the body, deformed fingers and somewhat strange eyes;

I've started a community on Deviant Art to share images made with my models/LoRAs! Join here if you are interested!

IF YOU ARE NEW TO STABLE DIFFUSION OR YOU HAVE PROBLEMS WITH MY MODELS, GO HERE: Guide

IF YOUR PICTURES HAVE PROBLEMS, CHECK MY FAQ SECTION: FAQs

What's this LoRA

This is a meme LoRA. One day I was looking at my kitchen plates, so I took a few photos and turned them into a LoRA. For the people that are usually concerned that my LoRAs do not look like the source material, I can guarantee that these plates are really similar to the ones I have in the kitchen 🤣.

The images in the examples are created using only this LoRA and different models (mainly my custom):

• https://civitai.com/models/24149

• https://civitai.com/models/26332

• https://civitai.com/models/28322

• https://civitai.com/models/32022

• https://civitai.com/models/33707

You don't need to specify a word to trigger this LoRA.

Guides

Learn how to create pictures like mine with my step-by-step tutorials:

• Beginner's Guide

• Prompting

• My Workflow

• Inpainting

If you want to create images with a high LoRA weight, I highly suggest you to check the tutorial.

Support

Do you want a custom LoRA or picture? Check out my commission page:

LoRA and Picture Commissions

Do you want to get access to my LoRAs before everyone, and contribute to my next models? Check out my supporter tiers on Deviant Art:

If you want to support my work for free, you can also check out my music/art here:

• Bandcamp

• Instagram

• DeviantArt

The skirt and similar clothing can interfere with the generation of handles. May make faces worse. Use Hires fix in that case.

Handles without furry and without inflatables (mostly)

Inspired by LoRA but didnt used its images: https://civitai.com/models/36929/

特殊浴場の風景をイメージしたLoRA

ソープマットとスケベイスも学習させました。

soapmat, sukebeisu, tiles, indoors, tile wall, faucet, tile floor, bathtub, scenery, bath, water, mirror, bath stool, shower stool with a u-shaped hole in the middle

洗面器＆ローションなんかも学習させてあります。

soapland, bath bowl

ブルマ卿のブルマLoRAの完成度にはかないませんが、青い体操服を学習させてみました。個人的な趣味でサイドにストライプが2本入ったタイプ多めです。

体操服、襟と袖に紺の縁取り付

taisoufukuN, gym uniform,

JP530タイプ、紺、サイドに2本ストライプ入り

BK2S, JP530S2N,

マツウラ601、紺、サイドに2本ストライプ入り

BK2S, M601N,

コマツ714、紺、サイドに2本ストライプ入り

BK2S, K714N,

キャッツアイ4301、紺、サイドに2本ストライプ入り

BK2S, 4301N,

モデルを分けて学習させたものの…期待の物が描かれる事は少ないです。もうどうやったら完成度上がるのか分からん…

プロンプト例：

masterpiece, best quality, ultra-detailed, illustration,

BK2S, taisoufukuN, gym uniform,

1girl, glasses, solo, long hair, black hair, smile

<lora:TaisofukuN_V2:1>

Wings

Suggested Prompts: yellow eyes, fingerless gloves

You know, AI or many models now is fond of "Arknights' open jacket". If you really want to take off her jacket, use "bare arms" (high strength?), and put "jacket" "open jacket" in Negative Prompts.

你懂的，AI或者说许多模型很喜欢“粥式外套”。如果你真的想脱了外套，使用"bare arms"(高权重), 并把"jacket" "open jacket" 放在反咒里。

This is a LoRA for generating images of Macross Delta character Freyja Wion.

Making models takes a lot of time and hard work. If you like this LoRA, please click the like button!

Recommended weights:

• Recommended weights: 0.5~0.7

  (Note: If the weight is below 0.4, the LoRA may not work effectively.)

這是一個用於生成超時空要塞Δ角色芙蕾雅薇恩
的圖像的LoRA

製作模型需要花費很多的時間和精力。如果您喜歡這個LoRA，請幫忙按一個“讚”！

建議權重：

• 建議權重：0.5〜0.7

  （注意：如果權重低於0.4，則LoRA可能無法生效。）

Balinese Kebaya LoRA

Trained using Chilloutmix Checkpoint

v1.0

30 Dataset Images,10 Repeats,10 Epoch and (i forgot the steps)

v2

73 Dataset Images,18 Repeats,10 Epoch and 2190 Steps

Use weight 0.6-0.8 for best result

available at huggingface.co

https://huggingface.co/meepmeepow/Lora

"Silly spider, cats always land on their feet."

Been a while since my last LoRA and needed a Black Cat to play around with. It seemed to favor a weight of 8.0 for my taste but feel free to lower for more variation.

凤眼也可以很美的!Fox eyes can also be beautiful.フォックスアイも美しいです.

鹤田一郎,日本画家,风格非常明显的东方美人画家.第一次尝试,优点是文件很小.

Tsuruta Ichiro , a Japanese painter, is a distinctly styled Eastern beauty painter. First try.One of the advantages is that the file size is very small.

鹤田一郎は、日本の画家であり、独特のスタイルを持つ東洋美人画家です。初めての試みです. の一つは、ファイルサイズが非常に小さいことです。

POV Blowjob - LoRA

All Preview Images are generated with very Simple Prompts! Try your Prompts its very flexible!

I seperate Train White and Black Penises check promts

Version 0.1 i focus on clear Faces.

In the next Version:

• Reduce Filesize (sry about that)

• Better Body (on 0.1 i focus on face)

• Fix the "Long Penis" Bug

• Clothes / Backgrounds more Realistic

• and more

All Preview Images are 0.3 - 0.4 Weight!

Generate Big LoRA Pack cost many Time and Training ... so pleace write a 5* review when you support me for free for other Support: BuyMeACoffee

Trained on: W:480 H:720

Best result at 512x512 because "to long Penises" will fixed at 0.2

Triggerwords:

povblow

bbc_povblow

See Preview Images for my Prompts.

Preview Images generate:

LoRA Load over: additional-Networks (work without it)

Model: Uber Realistic Porn Merge (URPM) 1.2

VAE: vae-ft-mse-840000-ema-pruned

Txt2Img and Upscale settings:

• EULER A

• Steps: 20

• 512x512

• CFG: 7

• LoRA Weight: 0.3 - 0.4

Role of Girls' Frontline

Weight 0.8~1.0 ,and you can ADJUST prompt to reductive character

TRIGGER:

1.BASE: original_outfit;

2.OG Costume: black headwear, cross, fur trim, bodysuit,

ᶘ ᵒᴥᵒᶅ
ᶘ ᵒᴥᵒᶅ
ᶘ ᵒᴥᵒᶅ

The A-545, also known by its GRAU designation of '6P67', is a Russian-designed select-fire assault rifle chambered for the 5.45x39mm intermediate cartridge. The A-545 is a successor variant of the AEK-971 line of rifles, and features numerous internal and external improvements over earlier AEK-971 models.

A loRA of Maya from Azur Lane
Use my loRA at 0.6,this seems to be the best value

《我的世界》物品图标

A LoRA trained with all 1.12.2 minecraft item textures scaled to 64x64.

That is, all the pics in the item folder, no blocks, no block items.

Note that it only works under 64x64. Smaller or bigger simply won't fit.

More like a warning and experiment than actural texture maker. I will put it here so people can satisfy their curiousity without fumbling this themselves again. There are so many pictures, it takes quite long to train it.

You will get background even if you tell it to go "white background", no use, I tried but have no avail in removing that black BG.

First try. Please refer.

When you're a cis or bi guy, they say you're either a 'boob man', or a 'butt man', or a 'leg man'. Me ? I'm definitely the latter !!! I don't mind a nice rack nor a shapely derriere, and as a sometime-glamor-photographer I've shot plenty of all three.

But show me a set of long, toned legs in a pair of sexy (overpriced) heels, and my heart rate starts going up, my mouth starts going dry, and I find it very hard to focus on anything else.

SD is pretty good at giving me my fix of good gams 'out of the box' ... but if you're anything like me, you're probably thinking "yeah, but it could still be better".

So ... here's my first attempt to address the 'issue'. It's my first working Lora, trained on a set of 244 images of sexy legs, shapely feet, seductive heels, and some appropriate lingerie thrown in. I recommend using a strength of between 0.3-06, and I've found this will generally determine if your shot keeps the rest of your subjects body, or just focuses in on the legs.

"Legshow" is the magic woid, "heels" or "high heels" also encourage good results.

It's definitely not perfect (mixed results with toes, typically require multiple generations, I'm sure inpainting would fix 'em too but hey this is only Week 3 or maybe 4 for me using SD, so proper use of inpainting is waaaay above my skill-level RN haha). I'm sure I'll tweak it some more and update, but pretty happy with it for now ... so ... go nuts :P

Tested in a buncha models as with all my releases, and works. Hell ... even Pixar got the legshow treatment and it worked. Yeesh !

Very popular singer of the 80s.

The model was trained on photographs of that time.

Works best in the 0.7-0.8 range.

📌Training model：AnythingV5V3_v5PrtRE.safetensors

📌LoRA set：0.6~0.8

📌Prompt：Arima Kana, hat, blue headwear, beret, blunt bangs, shirt, long_sleeves, school_uniform, bowtie, pink bow, gray skirt, thighhighs, white legwear, loafers, brown footwear

———🔰🔰🔰🔰🔰———

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

———🔰🔰🔰🔰🔰———

Recommend tags: blonde hair, blue eyes, bangs cover one eye

Can't really generate the scope on deidara's eye or the mouths on his hands

Here is the Lora for straddling paizuri!

The trigger words are penis, straddling paizuri

The suggested weight is 0.7 - 0.8

Hope you enjoy my work!

You are welcome to browser my other works if you are interested!

#

New Announcement (19th April)

Hello my followers. I am sorry that I bring a bad news to you today. The update frequency of Lora may be dropped dramatically or even be stopped for some reason. You can still leave your request for free (discord: qqqw123#4777) but I am not 100% sure I will make it. Thanks for your support and appreciation!

#

You are also welcome to leave your request for free and I will try to make it into a Lora if it is possible.

If you appreciate my works, you can support me via donation on ko-fi.com/dpp12

Here is the Lora for POV paizuri !

The trigger words are penis, paizuri, pov

The suggested weight is 0.7 - 0.9

Hope you enjoy my work!

You are welcome to browser my other works if you are interested!

#

New Announcement (19th April)

Hello my followers. I am sorry that I bring a bad news to you today. The update frequency of Lora may be dropped dramatically or even be stopped for some reason. You can still leave your request for free (discord: qqqw123#4777) but I am not 100% sure I will make it. Thanks for your support and appreciation!

#

You are also welcome to leave your request for free and I will try to make it into a Lora if it is possible.

If you appreciate my works, you can support me via donation on ko-fi.com/dpp12

The Joker LoRa is meant to closely resemble the appearance of The Joker in the 2019 Joker movie.
Works best between a weighting of: 0.5 - 0.8

Trained on NAI with 46 pictures.
Best result with weight between : 0.8-1.
Prompt activator : quolkeh
Other prompts : purple hair, purple eyes, ponytail, armor, belt, gloves, midriff, navel, shorts, thighhighs

Commissioned by lasserine on fiverr

Reviews are really appreciated, i love to see the community use my work, that's why I share it.
If you like my work, you can tip me here.

Got a specific request ? I'm open for commission on my fiverr gig *! If you provide enough data, OCs are accepted

*Any paid request will not be presented on this page, unless you give me permission to.

Asoul成员珈乐模型

v1版本适合画头像图或者大头图，全身半身图也能画，有时候可能会出现崩坏

v2版本可以画半身图，但是也有可能出现画面崩坏

v3.2.5版本最稳定，头图，半身图，全身图还有各个角度的图都可以画，对不同模型适配性还可以，但是建议用原模型融合其他模型后再使用，有时候会出现齐刘海

v3.2.6针对上一个版本进行了优化，分为有触发词版和无触发词版，有触发词版头部和身体有不同触发词。

若在使用此版本的过程中出现齐刘海问题，可以添加forehead这个tag解决。

v3.2.8是最完善的版本，建议用这个，全身图出现齐刘海时加上forehead就行。

v3.5添加了水手服，可以用c-serafuku,short sleeves,white cropped jacket,触发，要帽子的话加上hat就行

IF YOU JUST DOWNLOADED (8:20 EST MAY 1ST) PLEASE REDOWNLOAD!!! I UPLOADED THE WRONG FILE BY MISTAKE SORRY

InflateBot is a joke that went way too far; originally an attempt to fine-tune GPT2 on inflation erotica, its latest shape takes the form of a LoRA based on the excellent Fluffusion model.

The goal of InflateBot is to provide multi-target inflation and full-body (spherical) inflation that isn't just a featureless blob. Its training data is multi-concept, with folders for the following tags:

• "air inflation"

• "blueberry inflation"

• "floating" (where it coincides with an inflation tag)

• "hyper inflation"

• "inflation fetish"

• "spherical inflation"

The issues targeted by InflateBot are:

• Over-focusing on belly inflation (ignoring body/spherical/hyper tags)

• Creating blank spheres with no character at all

• Creating a sphere that is detached from the character, who may be relegated elsewhere in the image

• Not generating arms or legs.

In the past, putting "balloon fetish" in the negative prompt helped alleviate some of the issues, but the goal is for that to no longer be necessary. The first demo image below uses this technique, the second does not (Private Version 2 "Oxide" used for these demos)

While InflateBot isn't done yet, I feel like it's far enough along that I'm willing to share some snapshots and see what you guys can do with it! (Note that InflateBot is a LyCORIS, specifically LoHa, and works best with KohakuBlueleaf's a1111-sd-webui-lycoris extension.)

有人喜欢XXMix_9realistic_v20的这张脸，所以我单独把脸的LORA提取出来供大家Mix,希望大家喜欢。

Maria Robotnik from Sonic Adventure 2, Shadow The Hedgehog, Sonic X.

Use maria_robotnik to summon her, use long hair, blonde hair, parted bangs, blue hairband, blue eyes to form her head, use shirt, dress to summon her default costume.

这是我在civitai上上传的第100个lora，

但是说实话效果有些抽象捏

prompt中描述的主体可以是1girl或者1boy或者monster

然而我建议你在Negative prompt里面写上butterfly（问就是打标的时候脑子抽了一下）

玩的开心！

This is the 100th lora I've uploaded on civitai.

But to be honest the effect is a bit abstract

The subject of the prompt can be described as 1girl or 1boy or monster

However, I would suggest you put butterfly in the Negative prompt (I'm just asking because I had a brain fart when I typed it)

Have fun!

Puparia style LoRA

Making models can be expensive. Do you like what I do? Consider supporting me on Patreon 🅿️, feel free to buy me a coffee ☕ or just leave a ❤️

You might remember this amazing video by Shingo Tamagawa called "Puparia". I tried to reproduce that art style with this LoRA.

Works perfectly on AnyLoRA at weight 1. Increase the accuracy with the trigger puparia style.
I also suggest you select "None" as the upscaler for highres.fix.

I hope you have as much fun using this as I did. Enjoy!

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 according to the instructions (by default it's :1)

Kaworu Nagisa model

Trained on animefull

works well with anythingV4.5 Enjoy !

weight 0.7~0.8 is best.

I created this LoRA in response to a request.

Be sure to read the description.

mer1, tiara, sailor senshi uniform, blue sailor collar, bow, knee boots, choker, white gloves, blue choker, elbow gloves, jewelry, earrings, blue skirt

mer1, tiara, sailor senshi uniform, blue sailor collar, bow, knee boots, choker, white gloves, blue choker, elbow gloves, jewelry, earrings, blue skirt, 1990s (style)

1990s (style) prompts can be used to create an animated effect.

LoRA is generated entirely from anime images.
Should work well with any model.
NSFW works nicely too.

Check samples for prompts :)

Leave a like, comment, review but first of all, ENJOY ;)

A low-effort experiment, this LoRA is trained on pictures of realistic dildo. It is unstable, low-quality and prone to generate cursed images. Use at your discretion.

Human interaction with dildo was not trained. There are better LoRAs for that.

Trained and tested on SD 1.5.

Olivia from Fire Emblem: Awakening, trained on animefull.

Around 0.6-0.8 weight works well.

After messing with various LoRA, I was impressed by some from @LordOtako (very good stuff) and so I decided to make my own. While I think mine turned out pretty okay, it's very stubborn about using different clothes on higher weights (try adding 'white dress' to the negative) and on lower weights it can really mess up the hair. Also, the dataset used didn't include anything that wasn't her dancer outfit from Awakening, because otherwise it was consistently generating amalgamations of her normal outfit and her FEH alt outfit.

Also, if anyone has any tips for getting well drawn hands more consistently, I'd be glad to hear them. On my setup it takes quite some time per generation, and sometimes it feels like I have to go so many generations just to get something decent and then I gotta fine tune it and it just takes forever lol. Also just any tips in general would be cool, I've only been doing this for a few days and could use the help haha.

Finally, I'll probably make more of these! Super fun to do, and I've already got a couple characters in mind.

训练集是动画里面截图的

不想再炼莎拉萨了，这个背包，项链，发卡，太求恶心了，这次增加了图片数量，稍微好了这么一点，看来还需要改进训练和打标等等的方法，希望下一版能更好

图跑出来崩了，记得高清修复和降低权重

触发词是小写，直接点它后面的复制标志

Trained with majicmix as base model, this Lora is able to generate beautiful portrait of Mire, my favourite actress, especially closeup and upper_body.

How to use LoRA:

• Copy the file to stable-diffusion-webui\extensions\sd-webui-additional-networks\models\lora

• Enable Additional Networks and select mire.safetensors as your lora

• Make sure to change the weight (around 0.4-0.8), choose as your own wish

If you like my work and want to buy me a coffe :')

(updated link to ko-fi)

Goal is staying close to official artstyle.

事实证明，防毒面具确实可以让角色变的神秘（或者涩气？）

玩的开心！

As it turns out, the gas mask does make the character mysterious (or shibboleth?)

Have fun!

Привет люди, я создал новую LoRA с персонажем. Автор дал полное согласие на использование его артов, также поддержал меня в этом. Для img2img нужно использовал значение 0.5-0.6

Hi people, I created a new LoRA with a character. The author gave full consent to the use of his art, also supported me in this. For img2img, you need to use the value 0.5-0.6

I am currently working with @Philip on a LoRA collection.

Zinogre - Zinogre | Stable Diffusion LoRA | Civitai

Delga - You are here.

Description:

40 images trained on NAI.

Tested on my model:

Indigo male Doragoon mix - v16_hybrid | Stable Diffusion Checkpoint | Civitai

doing best on this model. (And only on this model)

Nice nsfw image generator.

Poor little delga :)

Issues:

• Can't do proper tattoo.

• Compatibility is relatively poor, doing worse on other models. (anything, AMO3, even on NAI)

Add a ❤️ and follow to support my future works.

model little overtrained, and even after it it's don't be sure about ears color/form->

"black animal ears" prompt for better stability of results.

Trigger word is "mei-mei"

Boosty Donation link

This LoRA allows you to generate girls who pee in their panties.

You can see several weights in the last image.

Trigger words: peeing self, have to pee, peeings

I think v2 is quite an improvement. Some noise_offset to improve color, higher network_dim to improve likeness, and more high quality images for robustness.

I always try to put my settings in the "About this version" section, so check that out.

Not an easy LoRa to work with, but it can work.

v2 is a slight improvement, but I still think I can do better. I think the training data needs higher quality.

Would you like to support, it will help me to continue making this type of content ❤️

This model is a early version, Gradually I will add more pictures and updating the model until make it better and more complete

Hello here some settings that can help to get some good results

Recommended resolution: 512x768

lora weight: 1.0 - 1.1

Simpling method: DPM++ SDE Karras

Steps: 30

Model: "Chilloutmix_NiPrunedFp32Fix", "Chilloutmix" or "NeverEnding Dream (NED)"

CFG: 7

You can use some trigger words like: bunnyunnie, twnayeon

also you can use: "blonde hair", "brown hair" to get some variations

High res is not very recommended but you can use it

I recommend use remacri to upscale

> <

I made this LoRA because not even anime checkpoints seem to be able to generate this expression on their own.

Usage (important!)

Compatibility

Should work with most anime-like checkpoints. Tested with AOM3 (works best), and MeinaHentai. more testing will be done later.

Weight

I recommend 0.75~1.0; however, should you use other LoRAs with this, try a higher value such as 0.85~1.1

When in doubt, increase weight~! This LoRA impacts artsyle minimally.

Trigger Words

> < is the trigger word. Recommended to put emphasis on this, like (> <:1.2)

Also try one eye closed in your prompt with the above! (may be hard to generate from.)

Limitations

On lower weights and with other LoRAs, this may fail!

Avoid using prompts about the eyes, unless you're using one eye closed with this.

Bias to open mouth. Try (open mouth:1.2) in negative prompt if you need to get around this.

炼机甲第二档

做的吉姆模型，用了大概20来张图片吧，抠成白底炼的

权重大概在0.5-0.7的时候自由度会比较高且部分还原，例如变猫变机娘啥的，往上就是背景可换但是机甲比较定性

Refining Mech Armor Second Gear

The Jim model I made used about 20 pictures, cut out into a white background

When the weight is about 0.5-0.7, the degree of freedom will be relatively high and partly restored, such as changing from a cat to a mecha, etc., the background can be changed at the top, but the mecha is more qualitative

Take a small website and like to like it: https://www.bilibili.com/video/BV1Sa4y137TY/?vd_source=fd14ef3c7aa9e8cb3e3f08b04caed768

Self Wedgie (Pulled by Self) Concept for one-piece swimsuit & leotard. Please tell me if work with other uniform.

Use badhandv4 for better hands.

LoRa was trained using Hollowstrawberry's Colab.

Prompt masterpiece, best quality,1girl, <lora:pulledwedgie-06:0.8>,pulledwedgie,pulled by self,wedgie,one-piece swimsuit

Negative prompt: badhandv4,(worst quality, low quality:1.4), (greyscale, monochrome:1.0), text, title, logo, signature

📝Biography📝

(generation advice below) 👇

🌟Violet Monroe 🌟

Megan Violet Carlson, who would later be known as Violet Monroe, was born on May 8, 1988, in Minneapolis, Minnesota. She attended a school for music and theater and started her career working at a grocery store at the age of 17. After doing some mainstream modeling, she was contacted by someone in the adult entertainment industry and began performing in adult movies in 2009. Violet has worked for various notable companies in the industry. She has also worked for various renown websites. In her spare time, Violet enjoys singing and studying astronomy.

Violet Monroe, also known as Violet Raye, is a busty and shapely 5'7'' redhead of Dutch and Native American descent, with black and red hair and brown eyes. Her real and natural 34C breasts are complemented by a toned 23-inch waist and shapely 35-inch hips. She has a heart tattoo on her right upper arm and a star tattoo (happy to say, so does my LORA 😏). Violet is based in Los Angeles but has worked all over. With her porcelain skin and sultry presence, Violet Monroe has made a name for herself in the adult entertainment industry.

💡Generation Advice💡

🔧Prompt and Sampling🔧

• Most images have been generated with LORA weighted @ .75.

• Most samples done with CyberRealistic v2 and jimEIDOMODE with sampling steps best above 20.

🚀Upscaling🚀

• Using CyberRealistic, this LORA seems friendly to Hi-Res fix upscaling and the Ulti SD upscale script.

• Often, 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.

🗒️Side Notes🗒️

• 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.

As with the rest, goal is reproducing the OG muv luv style.
I usually post the 8th epoch but it seemed to have some issues, uploaded it as a separate version with a grid.
You can PM me on pixiv or here (if this place allows it) for the other versions, too much of a hassle to upload them here.

Harukaze Doremi from Ojamajo Doremi, the first magical girl franchise in my childhood.

Right now we only have LoRA with her casual outfit, her witch outfit will train and be uploaded later.

use hrkzdrm_cs to summon her, uses red hair, double bun, musical note hair ornament to form her head, use hood, pink vest, pink shorts, belt to summon her casual outfit

You need closed mouth, smile if you want her little smile, since she loves to :D.

LoRA is generated entirely from anime images.
Should work well with any model.
NSFW works nicely too.

Check samples for prompts :)

Leave a like, comment, review but first of all, ENJOY ;)

Rain and Umbrella

Legal & Risk

In principle, I do not support the use of this model to generate any adult content. When using this model, users must abide by laws and regulations, and do not infringe other people's reputation, privacy or portrait rights.

Changelogs

• v1.0: Experimental, do not expect too much

Features

• good-shaped umbrella

• multicolored umbrella with adding color prompt

• compatible with 2D and photorealistic models

Usage

• compatible with all sampling methods, but i’m using my models: https://civitai.com/models/44418/eclipse and https://civitai.com/models/41141/halcyon , just because i’m trained this LoRA with it.

• Use 15-30 sampling steps

• CFG: 4.5-7

• Less complicated, short negative prompts are better

If using Hires:

• Hires step: <20

• R-ESRGAN 4x+

• Denoising: 0.3-0.7

After download this LoRA, try to share with us about your findings.

Please give star reviews with your generated picture if you like it.

p.s. It's not necessary but If you like my creations please consider to support me to create more projects by donate in here:

https://ko-fi.com/gorillamonsooniii

I'm also accept commisioning for LoRA/Others

doraemon_\(character\)

哆啦A梦，机器猫，小叮当，蓝胖子

doraemon, ドラえもん

A trail of my laughing big doraemon plushie. Note that do not use normal negative prompts for human onto it. Doraemon has different structure than human.

• sakuragi_sd_v1 Lora:权重:0.6-0.85，触发词 sakuragi sd 生成相应的湘北队服，触发词 sakuragi head 生成头像照，V1是测试版还需要进一步改进，希望大家多提意见。

• sakuragi_sd_v1 Lora: Weight: 0.6-0.85, trigger word sakuragi sd generates the Shohoku uniform, trigger word sakuragi head generates a profile photo, V1 is a test version that needs further improvement, hope everyone will provide more suggestions

Lora I made of Hwasa from MAMAMOO

大肥蛇的模型

触发词:Holhaya

lora权重:0.8-1.0

lora model of arknights-ho'olheyak

Necessary trigger words:Holhaya

Lora Weight: 0.8-1.0

It's perfect for the "beautiful realistic asian" checkpoint,

Weights 1 or slightly below (0.85 - 0.95), 25-30 sampling steps

Uncheck restore faces for better results.

Resolution 512x512 | 512x768 | 768x768

Support me at Ko-fi

https://ko-fi.com/aibyia

Followe me on instagram @aibyia.p

I don't get any from my training but If you want to support my work. check patreon.com/goofyai

also my new merge model available https://civitai.com/models/53761/goofball-mix

all my Loras are available for free use at https://pixai.art/@goofy-ai

Lora weight of 0.7-1

highres fix or mg2img at higher resolution is highly recommended.

drop tour review. it is much appreciated.

The default trigger for the character is "inui sajuna, bangs, pink hair, long hair, hair between eyes, pink eyes, crossed bangs, small breasts"

Weight for all sample image is "0.7"

If you don't know what it is, just read wiki.

The base model is Suzumehachi.

Singer

the training was on a model - CyberRealistic 2.0.

I'm Render on - EndlessReality

I do not recommend using my LoRA with model RealisticVision

personally, I most often perform generation with parameters: DPM ++ 2M Karras, Sampling Steps 35, Denoising strength 0.35, CFG Scale 3.5-4.5

evaluation of the model is welcome

DEWEY FINN - The Chaotic Anarchist.

No, this is NOT Alex Brightman OR Jack Black.

This is a fusion of both of the main actors who were in the KOREAN tour, and predominately has data mostly of the more popular actor who's also toured in a secondary version of Beetlejuice.

In a sense this is a CELEB and CHARACTER lora, but also it was an accommodation lora for us who has a Dewey Finn fictive who needed a better face representation.

WORKS ON ANY MODEL - was trained on ANY LORA.

If you got requests, or concerns, We're still looking for beta testers: JOIN THE DISCORD AND DEMAND THINGS OF US: https://discord.gg/Da7s8d3KJ7

Listen to the music that we've made that goes with our art:
https://open.spotify.com/playlist/00R8x00YktB4u541imdSSf?si=b60d209385a74b38

We stream a lot of our testing on twitch: https://www.twitch.tv/duskfallcrew

any chance you can spare a coffee or three? https://ko-fi.com/DUSKFALLcrew

Experimental!

It's perfect for the "beautiful realistic asian" checkpoint,

Weights 1 or slightly below (0.8), 25-40 sampling steps

Uncheck restore faces for better results.

Resolution 512x512 | 512x768 | 768x768

Support me at Ko-fi

https://ko-fi.com/aibyia

Followe me on instagram @aibyia.p

HEIL IRONBLOOD!

Guten tag shikikan! Today is the wonderful day for all shikikan out there! I am releasing the KMS Brünhilde with her costumes and rigging that you can use for your stable diffusion! Now you can make your own fanart of this beautiful shipfu. o(*^＠^*)o

So this Lora have 2 versions, one exclusive for my Pixiv Fanbox supporters and one is free for the public use. The differences are the training methods that are used in the premium and the free version. The premium version also unlocks the special chinese new year outfit and special Iron Blood Lingerie

ハイル・アイアンブラッド！

グーテンタッグ式鑑！今日は四季館にとって素晴らしい日です！ KMS ブリュンヒルデの衣装とリギングを安定拡散に使ってリリースします！これで、この美しい船風のファンアートを作成できます。 o(*^＠^*)o


したがって、この Lora には 2 つのバージョンがあり、1 つは私の Pixiv Fanbox のサポーター専用で、もう 1 つは一般向けに無料で使用できます。違いは、プレミアム版と無料版で使用されるトレーニング方法です。プレミアムバージョンでは、特別な旧正月の衣装と特別な鉄血のランジェリーもアンロックされます

If you like what I do, consider supporting me on my fanbox

https://mariegold.fanbox.cc/ or buy me a coffee

SUGGESTED LORA STRENGTH

20 epochs >>> 0.3-0.8

推奨されるローラの強さ

20 エポック >>> 0.3-0.8

Here are the trigger words of each costumes:

各衣装のトリガーワードは以下の通りです。

1) Default: Bruenhilde, 1girl, black coat, BRN_costume, bodysuit, metal gauntlets, black boots, red eyes

2) Rigging >>> Bruenhilde, 1girl, black coat, BRN_costume, bodysuit, metal gauntlets, sitting on a robot, black boots, red eyes, rigging_background

3) Brünhilde: Grain-Seizing Grani (Donut) >>> Bruenhilde, 1girl, SPD_outfit, white shirt, red hoodie, red shorts, wet, sweaty, red eyes, red headband, holding a donut with her mouth, hands behind body, black sneakers, white background, full body shot,

4) Brünhilde: Grain-Seizing Grani (Bagel) >>> Bruenhilde, 1girl, SPD_outfit, white shirt, red hoodie, red shorts, wet, sweaty, red eyes, red headband, holding a bagel with her mouth, hands behind body, black sneakers, white background, full body shot,

基于无限版沙也加进行训练

STD 版本只用了剧场版截图进行训练

DOUJIN版本追加了部分同人图进行训练

I can call this model one of my prime works, suitable for both generating images of Emma Watson, and Hermione Granger. Based on an extended catalogue.

If you want to use Emma just write: realistic

If you want to use Hermione write: ermione

I realise on request: cassulof@gmail.com

Star Lord - Lord of Stars? The one that literally cannot stop being a dork?

Yea, the one that was born of the comics, starred by Chris Pratt- and is literally in every Marvel game since 2017.

This does a mixture of Old-ish Quill, GOTG Video game Quill, Comics and Chris Pratt.

If you got requests, or concerns, We're still looking for beta testers: JOIN THE DISCORD AND DEMAND THINGS OF US: https://discord.gg/Da7s8d3KJ7

Listen to the music that we've made that goes with our art:
https://open.spotify.com/playlist/00R8x00YktB4u541imdSSf?si=b60d209385a74b38

We stream a lot of our testing on twitch: https://www.twitch.tv/duskfallcrew

any chance you can spare a coffee or three? https://ko-fi.com/DUSKFALLcrew

Based on the portrait style of the video game Stellaris. There are two different versions since the training results varied quite widely. The versions posted are the top performing epoch from the best two training runs.

Sample images were generated using the simplest possible prompts just the overall style trigger word and a species trigger.

This was my first attempt to put multiple training concepts into a single LORA so some of them don't seem to have taken quite as well as others. I ended up needing to change the trigger word spellings in some cases as using standard spelling was causing issues, some words just couldn't seem to break out during training from whatever the existing meaning for that word in the base model.

A bit finicky with getting the sword and shield right. Sometimes Hair color can go wrong, but specifying Pink hair helps.

All tags are in the metadata.

LoRA is generated entirely from anime images.
Should work well with any model.
NSFW works nicely too.

Check samples for prompts :)

Leave a like, comment, review but first of all, ENJOY ;)

基于NAI底模训练的符玄角色lora

使用方法：

推荐权重：0.7~0.9，建议使用高清修复

核心触发词：1girl, official, purple and yellow eyes

推荐触发词：1girl, official, purple and yellow eyes, purple hair, bare shoulders, detached sleeves, hair ornament

translate by GPT：
Based on NAI training, the model of the Honkai Star Rail character, Fuxuan.

Usage:

Recommended weight is 0.7-0.9, and it is recommended to use high-definition restoration.

The core trigger words are: 1girl, official, purple and yellow eyes

It is recommended to use trigger words: 1girl, official, purple and yellow eyes, purple hair, bare shoulders, detached sleeves, hair ornament

-praise tag shuffling-