Train based on AnyLora model.Training materials are basically official arts and the leader live2D.

Strongly recommend weights from 0.6 to 0.8,and enable hires.fix.

鬼爷求求你摇了我吧😭😭😭我v你50吃疯狂星期四

[4 of the sample images is being moderated :S]

A compilation of most beautiful cosplayer, now in Ai.

Training cosplayer isn't the easiest thing to do, since their data sets most of time are in cosplay, it becomes unique and inconsistent at the same time, so most of the time, in my case this messes up the training. This thread is dedicated to improve the outcome of the cosplayers lycoris, in the hope to have be able to generate an accurate representation of the cosplayer with many cosplays they did in real life.

IMPORTANT NOTE: Please look at the ABOUT THIS VERSION to see the Prompting Guide

ver 1.0 - this is actually a failed version, however, the results are still quite decent therefore i decided to share and hopefully with some discussion and inputs it can finally be perfected in further iterations. as usual, i will post update in this thread for future updates or addition.

===

If you love my work, and wanto to treat me a coffee : ko-fi.com/tauron you dont have to, but if you want, it helps me stay awake until midnight (which is the only time im able to work on these stuff lol) and cope with life..

My model is out, check out: Tauron Hybrid Mix - TauHybrid_Anime_Final | Stable Diffusion Checkpoint | Civitai

--

DISCLAIMER:
All of my models are to be used with 21+ subjects only, I'm doing this for fun only, DO NOT use for commercial purpose. Nudity generated with this model DOES NOT DEPICTS REAL PERSON.

===

NOTE : I will keep updating more actresses in versions update

Open to suggestion or request

Request Rules:

1. No requesting minors or anything that has underage as their datasets! please don't make civitai or ai in general into trouble.

2. If requesting a person, that person should have a public images (ex. no private instagrams), they themselves has to consented sharing their images online for public.

3. Be a decent human being, if you requested, please at least give a review, since I'm doing this for free. and dont be a dh and review bomb if you don't get what you want, not all lyco is perfect.

4. Do not start discussion that is pointing towards anything illegal.

===

Usage Guide

--

A. Lora Weight

• 0.1 ~ 0.4 -- is to combine multiple faces or features, or if you experience overbaking

• 0.5 ~ 0.7 -- is a balanced weight to get still accurate feature but with more flexiblity

• 0.8 ~ 1.0 -- is to create accurate representation of the features.

B. CFG Scale

• Lower CFG Scale means AI gets more creative and less overbaking problems.

• Higher CFG Scale means results get very accurate with training sets and more prone to overbaking problems.

• Dynamic CFG Threshold, using this method allows you to bump up quality and achieve photorealistic easily, but with lycoris i found out recently that this method actually lower success rate, so only use when you are composing with a Lora / model that works fine with this method.

C. Sampling Steps, High Res Fix & Upscaling

• I like to use sampling steps at 50, with high res step at 20, or 1/2 of sampling steps.

• I find that 2M Karras works better and faster than SDE Karras. with ERSGAN 4x+ method + ERSGAN 4x for Upscaling. This method actually gives quite good result without being too sharp or too plasticky

• Upscaling - GPFGAN @0.25, Codeformer @0.3 & Codeformer Weight @0.15 seems to yield superior results to me.

--

Combining with Attribute LoRa / LyCoris

You may need to do your own testing, but using character weight 0.2 below attribute lora weight yields superior results, if for some reason one features persists not showing or overbake, you may want to lower the weight of the overbaked feature Lora, and then use emphasis () to boost the specific attribute you want.

DDetailer Script

If subject face is small / far away use: DDetailer Script with bbox anime face with default settings except denoiser at 0.2-0.3. You can combine this LoRa with a lot of other Style / Attire / Pose / Acts LoRa, Also works well with character LyCoris.

How to install this script? I kinda forget how because its a bit complicated beacuse you need to install stuff from venv. but a quick google search will land you somewhere in tutorial page.

Please leave a review if you have any critiques or if you find this LoRa good.

your need Additional Networks to use it.

Can't draw kos-mos like choco,Can't draw fairy knight lancelot \(fate\) like choco.

but it still like choco. it's just a style.

"wtf"

Thanks LAX help me.

weight 0.5~0.6 is best.

english↓

This is lora by inui toko, a member of the virtual youtuber group nijisanji.

The strength of lora is recommended to be around 0.8~1.

Trigger words:1 girl,initk

If you want her to wear a kimono→ "japanese kimono" and "wa maid"

if you want to make it easier to give her dog ears,→ "dog ears"

if you want to make it easier to give her odd eyes,→ "heterochromia", etc. in the prompts.

It also works well if you put the words that occur frequently by putting lora itself through analysis.

This may not be a lora with satisfactory accuracy since it was made as a trial.

Please enjoy it if you like.

https://www.anycolor.co.jp/guidelines/

japanese↓

これは、バーチャルyoutuberグループのにじさんじに所属している戌亥とこのloraです。

loraの強度は0.8~1くらいがおすすめです。

トリガーワード:1 girl,initk

着物を出したい場合は「japanese kimono」と「wa maid」

犬の耳を出しやすくしたいなら「dog ears」

オッドアイを出しやすくしたい場合は「heterochromia」等をプロンプトに入れると効果があります。

また、lora自体を解析にかけて頻出するワードを入れてみると効果があります。

お試しで作ったので、満足のいく精度のloraではないかもしれません。

よければお楽しみください。

... and the god of Ai said: Let there be waifus! and there were Ai waifus.

There are anime loras out there what is the difference?

My version focused heavily on cosplay -- this means that it has a bulk of cosplayers data sets in instead of arts / illustrations - the ratio in general are 60-40. Secondly, with Lycoris we only need 11MB, more space for more waifus! Third, there are NSFW datasets mixed in, so have fun and enjoy!

===

If you love my work, and wanto to treat me a coffee : ko-fi.com/tauron you dont have to, but if you want, it helps me stay awake until midnight (which is the only time im able to work on these stuff lol) and cope with life..

My model is out, check out: Tauron Hybrid Mix - TauHybrid_Anime_Final | Stable Diffusion Checkpoint | Civitai

--

DISCLAIMER:
All of my models are to be used with 21+ subjects only, I'm doing this for fun only, DO NOT use for commercial purpose. Nudity generated with this model DOES NOT DEPICTS REAL PERSON.

===

NOTE : I will keep updating more actresses in versions update

Open to suggestion or request

Request Rules:

1. No requesting minors or anything that has underage as their datasets! please don't make civitai or ai in general into trouble.

2. If requesting a person, that person should have a public images (ex. no private instagrams), they themselves has to consented sharing their images online for public.

3. Be a decent human being, if you requested, please at least give a review, since I'm doing this for free. and dont be a dh and review bomb if you don't get what you want, not all lyco is perfect.

4. Do not start discussion that is pointing towards anything illegal.

===

Usage Guide

--

A. Lora Weight

• 0.1 ~ 0.4 -- is to combine multiple faces or features, or if you experience overbaking

• 0.5 ~ 0.7 -- is a balanced weight to get still accurate feature but with more flexiblity

• 0.8 ~ 1.0 -- is to create accurate representation of the features.

B. CFG Scale

• Lower CFG Scale means AI gets more creative and less overbaking problems.

• Higher CFG Scale means results get very accurate with training sets and more prone to overbaking problems.

• Dynamic CFG Threshold, using this method allows you to bump up quality and achieve photorealistic easily, but with lycoris i found out recently that this method actually lower success rate, so only use when you are composing with a Lora / model that works fine with this method.

C. Sampling Steps, High Res Fix & Upscaling

• I like to use sampling steps at 50, with high res step at 20, or 1/2 of sampling steps.

• I find that 2M Karras works better and faster than SDE Karras. with ERSGAN 4x+ method + ERSGAN 4x for Upscaling. This method actually gives quite good result without being too sharp or too plasticky

• Upscaling - GPFGAN @0.25, Codeformer @0.3 & Codeformer Weight @0.15 seems to yield superior results to me.

--

Combining with Attribute LoRa / LyCoris

You may need to do your own testing, but using character weight 0.2 below attribute lora weight yields superior results, if for some reason one features persists not showing or overbake, you may want to lower the weight of the overbaked feature Lora, and then use emphasis () to boost the specific attribute you want.

DDetailer Script

If subject face is small / far away use: DDetailer Script with bbox anime face with default settings except denoiser at 0.2-0.3. You can combine this LoRa with a lot of other Style / Attire / Pose / Acts LoRa, Also works well with character LyCoris.

How to install this script? I kinda forget how because its a bit complicated beacuse you need to install stuff from venv. but a quick google search will land you somewhere in tutorial page.

Please leave a review if you have any critiques or if you find this LoRa good.

This is a multi-training program that combines the characters of Ensemble Stars! into one.

I received the source from "tamaokaima" and then created it upon his request.

Note that I did not know who to make, so I made everything one.

This work has brought me from ignorance to slight acquaintance with this work (laugh)

*****Caution!*****

ex) 1boy, solo, <lora:ensemble_stars:0.6>, ensemble stars, [character name], some tags

Due to the large number of characters, contains learned tags in zip.

Please use it as a reference for generatation.

Recommend weights = 0.8~1

Recommand Negative prompts = EasyNegative, (worst quality:1.4), (low quality:1.4), extra fingers, fewer fingers, (speech bubble:1.2, bar censor:1.3)

v1.0 Train Settings

• Type: LyCORIS/LoCon

• Dim: 128

• Learning rate: 2.5e-4

• Text encoder learning rate: 1.25e-4

• Unet learning rate: 2.5e-4

• Repeat: 60 (iom 1girl)

培训讯息

第一版本v10训练步数1800步(第二版本V20训练3600步)，基于sd1.5模型训练，请根据使用的模型调整权重。 Clip跳过层1。

训练使用86张高清无码无水印去除背景的NSFW照片,全部皆为裸体照片，手动标记提示词,300张正则化照片。

擅长分开阴部、分开肛门、分开肛门。

培训照片的解析度512x768,768x512.(脸和全身和阴部特写)

请勿再评论发布裸体照片。

这是一个 LyCORIS (LoCon/LoHA) 模型，需要在 Automatic 1111 中额外扩展才能工作。

范例测试使用的模型:

braBeautifulRealistic_brav5

chilloutmix_Ni

realisticVisionV20_20

触发词

触发词:A woman named Anya01,Anya01

额外可用提示词: asian,Detailed skin texture, detailed skin pores, high skin detail,mole,mole_on_breast, nude,completely naked, a woman with a hairy pussy, on all fours, legs apart, simple background, White background, LABIA MINORA,nympha,LABIA MAJORA,everted labia,clitoral hood,urethra,Goosebumps,urethral opening,vaginal opening,

sitting, lying, Moist pussy,small breasts,rough areola,long hair, tattoo

spread pussy,gaping, spread anus,spread ass, puckered anus,indoor, upper body,medium close shot,pussy close-up,face close-up,photo_\(medium\),

photorealistic,Realistic,anal hair, pussy juice,pussy_juice, fingernails,fingers,hands, hairy pussy,excessive pubic hair,mismatched pubic hair, from behind,ass focus,butt lift,m leg,ass in the air, laughing, legs up,from side,

如果你喜欢我的工作并想支持我：https://ko-fi.com/O5O2K9TV1

免责声明：

我不支持公开散布此模型中所有人物生成的任何不雅图像，我不支持出售此模型中人物生成的图像。请不要使用他们在网上发布的不雅内容。

TRAINING INFO

The first version v10 has 1800 training steps (the second version V20 has 3600 training steps), based on the sd1.5 model training, please adjust the weight according to the model used. Use Clip skip1.

Training using 86 high-definition uncensored NSFW photos without background removal. All are nude photos. Label the prompt words manually. 300 regularized photos.

Good at spread pussy,spread anus,spread ass.

The resolution of the training photos is 512x768, 768x512. (face & full body&pussy close-up)

Don't comment on posting nude photos anymore.

This is a LyCORIS (LoCon/LoHA) model, and requires an additional extension in Automatic 1111 to work.

Models used for example tests:

braBeautifulRealistic_brav5

chilloutmix_Ni

realisticVisionV20_20

TRIGGER WORD

Trigger Words: A woman named Anya01,Anya01

Additional prompt words available: asian,Detailed skin texture, detailed skin pores, high skin detail,mole,mole_on_breast, nude,completely naked, a woman with a hairy pussy, on all fours, legs apart, simple background, White background, LABIA MINORA,nympha,LABIA MAJORA,everted labia,clitoral hood,urethra,Goosebumps,urethral opening,vaginal opening,

sitting, lying, Moist pussy,small breasts,rough areola,long hair, tattoo

spread pussy,gaping, spread anus,spread ass, puckered anus,indoor, upper body,medium close shot,pussy close-up,face close-up,photo_\(medium\),

photorealistic,Realistic,anal hair, pussy juice,pussy_juice, fingernails,fingers,hands, hairy pussy,excessive pubic hair,mismatched pubic hair, from behind,ass focus,butt lift,m leg,ass in the air, laughing, legs up, from side,

If you love my work and want to support me: https://ko-fi.com/O5O2K9TV1

Disclaimer:

I do not support the public distribution of any indecent images generated by all the characters in this model, and I do not support the sale of images generated by the characters in this model. Please do not use the indecent content they post online.

anime model，weight around 0.7

S12 V4.2 Technidrawing | Side-training

All images generated use this model without any additional embeddings or loras(Only G Mix + technidrawing) - https://civitai.com/models/67130/acmar-g-mix-or-amateur-deep-merge-model

Training based on processed S12 dataset of V4.2 itteration. Inherits most benefits of S12, while concentrating on specific style, technical drawing(i think) in this case. If anyone knows better - please tell.

Extremely editable and generalized. Capable of combining seemingly unrelated tags in lots of cases that it was not trained on.

Fully manually tagged. 100% SFW.
Original model without technidrawing stylization: https://civitai.com/models/22696/s12-v42-update-or-fine-tune-lora-or-extended-lora

Hey, if you like what i'm doing, consider giving me a dollar or two - https://www.patreon.com/anzhc :D
This will help me produce more and better.

Elna Stongs (Ore dake Haireru Kakushi Dungeon)

All uploaded images are generated by YesMix

options

LoRA weight 0.6~0.7

Sampling steps 30 or Higher

Recommend Hires

Trigger words

• Default

  belt,elna stongs, elna, ore dake haireru, belt_buckle, black_hair, black_pants, long_sleeves,pants,white_shirt,

This incorporates

• FairyTaleAI

• TikiAI

• VikingPunkAI

• ZuluPunkAI

• OldEgyptAI

• PiratePunkAI

In future versions, I'll add TotemPunkAI and BronzePunkAI. Must redo them first.

The preview pics were generated with prompt:

<lyco:CulturePunkBundle-000007:.8> {FairyTaleAI | TikiAI | VikingpunkAI | ZuluPunkAI | OldEgyptAI | PiratePunkAI } [concept]

Perpetual coffee request: Keep me motivated and caffeinated:

https://www.buymeacoffee.com/konyconiN

https://ko-fi.com/konyconi7992

Well, for this one it would be needed this: https://civitai.com/models/47085/envybetterhands-locon

For the VAE I'm using on the images its: vae-ft-mse-840000-ema-pruned

Check the metadata of the images for the examples of some poses.

If you'd like to support me and do more:
https://ko-fi.com/edgonkofi - commissions open! I can make a model for you!

Royal Gowns

We do like to dress to impress, don't we? What's more, we are lacking in beautifully detailed clothing.

Fear not friends. I've got your back.

More beautiful clothing is incoming, beautiful royal gowns to please your eyes and dress your women with style!

Special shout out to @Zovya, I am very impressed with the results of the clothing on her model as I am showcasing it here.

I highly recommend you download either her model A-Zovya RPG Artist Tools - V2

or my Amixx - v1.0-pruned-fp16

when prompting with these models.

Have fun!

Download this LoCon and get some gowns!

LoHA trained on Kayn from League of Legends

models

anime: MeinaAlter

2,5D: MeinaUnreal, HELLmix, PsyGon Animerge - v1.1, ReV Animated

Kayn: kayn, 1boy, male focus, split theme, braided ponytail, topless male, armor, pauldrons

Shadow Assassin Kayn: shadow assassin kayn, 1boy, male focus, blue skin, forehead mark, multicolored hair, very long hair

Rhaast: rhaast, monster, horns, red skin, glowing eyes, red eyes, armor,

negative embeddings: embedding:bad-hands-5, embedding:ng_deepnegative_v1_75t, embedding:bad-artist, embedding:EasyNegative

Recommended settings:

• Make sure you are aware on the usage instructions of LORA

• VAE: orangemix.vae (for 2,5D pictures), vae-ft-mse-840000-ema-pruned (for realistic models), kl-f8-anime2, blessed2.vae (for anime models)

• Sampler: DPM++ SDE Karras or DPM++ 2M Karras

• Steps: 20 to 35

• Clipskip: 1 or 2

• Upscaler : img2img in comfyui using 4x-UltraMix_Smooth

• CFG Scale : 5 to 9 (mostly 7 and 8)

• LORA weight for txt2img: 0.5-1 (for the examples i mostly used 0.7)

• Denoise strength for img2img: 0.4 to 0.7

If you want the exact same settings, just drag and drop the picture into ComfyUI

for the pictures i used comfyui

Let us embark on a journey to embrace some thot. Sometimes they are simply breath-takingly beautiful.

If you love my work and wish to buy me a coffe: ko-fi.com/tauron it helps me stays awake and get more content up

My model is out: Tauron Hybrid Mix - TauHybrid_Anime_Final | Stable Diffusion Checkpoint | Civitai

DISCLAIMER:
I'm doing this for fun and not for commercial purpose, I do not support selling imageries generated with all subjects inside this model. I do not support indecent generation images of all subjects used in this model. DO NOT post online explicit nudity, sexual or suggestive images using this model. All subjects are 21+.

===

Open to suggestion or request

NOTE : I will keep updating more actresses in versions update

===

Usage Guide

--

A. Lora Weight

• 0.1 ~ 0.4 -- is to combine multiple faces or features, or if you experience overbaking

• 0.5 ~ 0.7 -- is a balanced weight to get still accurate feature but with more flexiblity

• 0.8 ~ 1.0 -- is to create accurate representation of the features.

B. CFG Scale

• Lower CFG Scale means AI gets more creative and less overbaking problems.

• Higher CFG Scale means results get very accurate with training sets and more prone to overbaking problems.

• Dynamic CFG Threshold, using this method allows you to bump up quality and achieve photorealistic easily, but with lycoris i found out recently that this method actually lower success rate, so only use when you are composing with a Lora / model that works fine with this method.

C. Sampling Steps, High Res Fix & Upscaling

• I like to use sampling steps at 50, with high res step at 20, or 1/2 of sampling steps.

• I find that 2M Karras works better and faster than SDE Karras. with ERSGAN 4x+ method + ERSGAN 4x for Upscaling. This method actually gives quite good result without being too sharp or too plasticky

• Upscaling - GPFGAN @0.25, Codeformer @0.3 & Codeformer Weight @0.15 seems to yield superior results to me.

--

Combining with Attribute LoRa / LyCoris

You may need to do your own testing, but using character weight 0.2 below attribute lora weight yields superior results, if for some reason one features persists not showing or overbake, you may want to lower the weight of the overbaked feature Lora, and then use emphasis () to boost the specific attribute you want.

DDetailer Script

If subject face is small / far away use: DDetailer Script with bbox anime face with default settings except denoiser at 0.2-0.3. You can combine this LoRa with a lot of other Style / Attire / Pose / Acts LoRa, Also works well with character LyCoris.

How to install this script? I kinda forget how because its a bit complicated beacuse you need to install stuff from venv. but a quick google search will land you somewhere in tutorial page.

Please leave a review if you have any critiques or if you find this LoRa good.