Fairy among flowers in the moonlight

Note #1: Wanted to try to mix it with my in development, quite NSFW, Broken and Defeated Locon but TOS wouldn't allow it to be uploaded. Check it out on my twitter. If the Locon won't be able to be uploaded here once it's ready, I'll have to find somewhere else to put it.

What does it do you ask? Who could this be? I'm not sure but it's definitely not Zelda dressed as Sheik. Totally not. And totally not spoilers. So put this unknown figure into your Stable diffusion so you can make some dope images with them. To change outfits, don't use the trigger word SheikZ and lower weight to 0.5 to 0.7, as I had to train it hard and removing those two things give flexibility..

How to use it:

(I use Sampling Method DPM++ SDE Karras and Upscaler R-ESRGAN 4x+ Anime6B). Will need ~0.5 to 0.7 weight to get good clothing changes (as well as probably removing the trigger word of SheikZ), as I had to train it to get all the details right on the outfit she usually wears.

Main Triggers

SheikZ - The main trigger word and then

WMASK or NMASK - Depending if you want to emphasize her mask appearing or not.

Highly Recommended Tags

bodysuit, braid - Bodysuit helps make sure her bodysuit shows up well. Don't use it when changing clothes. Braid is if you want to keep her hair between clothing changes.

And then whatever else you want!

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

Picture Descriptions:

1: AbyssOrange with just this Lora.

2-3: AnythingV3 with just this Lora.

4: RevAnimated using just this Lora.

5: MeinaPastel using just this Lora.

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

7: AnythingV3 combo'd with my in progress Masked Evil Locon (Basically get an evil masked place on you and become evil).

8: AnythingV3 combo'd with my in progress Anime Officer.

9: AnythingV3 combo'd with my in progress Biker Chick Locon.

10: AnythingV3 combo'd with my in progress Delinquent Locon.

How was it trained?

It was trained on animefull_final with 26 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?

Should be good to go with this one but let me know of errors.

What's next that isn't this?

More character Loras. (Viper and Rupee 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, Chika Fujiwara from Kaguya-sama, Lulu from FF10, Momo from My Hero, 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), Demonic Transformation, 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), Brainwashing Machine, 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, Mind Flayer Defeat, 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 Goth Chick, 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.

ilulu anime kobayashi work with 0.6 weight enjoy it

follow me on Instagram @creativehotia

推荐模型：Anything3.0

推荐权重：

单独加载时：0.65-0.75

配合其他lora加载时：0.7-0.8

LoRA I made of my wife, made with Koyha with out about 25 images. 15 of her face and 10 of her body. Full body shots were taken nude to get the most accurate model. It's about 90% correct at this point. Can be used for both sfw and nsfw

(I don't play Elden Ring...not sure who are those characters...)

This Lora is for human characters only.

A lovely French singer.

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.

Sample images was generated with 🌕Luma

Recommended embeddings:
1. easynegative
2. bad-hands-5
3. badhandv4
Strength 0.6-0.8 is optimal.

PLEASE MAKE SURE THAT YOUR LORA FILENAME MATCHES LORA NAME IN PROMT.

Classic outfit: aqua hair, glowing eyes, white dress
Alternative outfit: aqua hair, twin braids, ponytail, glowing eyes, white armor, pauldrons
Optional: jewelry, earrings, tiara
Hair color: blue, green, purple, aqua
Eye color: blue, black, purple, yellow
Negative prompt: loli, kid, child, old

Hi ! Today i give you Sejuani !

If you like what i do, you can buy me a coffee to support me.

If you want a specific character that doesn't exist yet,
I can create your own character, tell me which character would you like me to create, you can read all the informations in this Cerate your Lora

I used Kohya to train this LORA

Thanks

Just for fun

Lora of a Big Daddy character from "Bioshock"

If you like my work you can buy me a Ko-fi, I would be grateful 🥰

https://www.buymeacoffee.com/Crabeon

https://ko-fi.com/crabeon

最好带上barefoot词条，不然容易出现不应该出现的鞋子

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.8 - 1.0

Trigger + Usual Appearance: vegeta, black spiked hair, black eyes

Default Outfit: armor, white gloves

Super Saiyan: super saiyan, blonde spiked hair, blue eyes

This next section concerns my thoughts and addresses one of the feedback I had received from my form (and it'll also explain why the file size is larger than usual, but it's loooong read)

I first want to start off by saying thank you to everyone that sent requests, left positive reviews, and comments. I have released so many character LoRAs (with a few styles sprinkled in there) in a span of about a week. This is all because I have Hollowstrawberry's Colab to thank for, which is always linked at the top of each model description. I am also able to get out these models so quickly because I have chosen to pay out of my own pocket to get Colab Plus. This is indicative of an addiction as you can tell lol. You should've seen how I scour and work with images (and god forbid, webp files) not typically found on booru sites.

Now moving on to the one feedback that I received overnight. I received some resources concerning the training parameters of a LoRA model. I of course always use the base NAI model, it's the default setting in Hollowstrawberry's Colab and it's just a click or two away from seeing it. It makes the most sense because then it translates well to other mixed models (that or I am just speaking out of my ass, I just like to generate pictures). Here's the parameters that I used for the Vegeta model:

• I had 133 images of Vegeta to work with, all of them taken from Gelbooru using Hollowstrawberry's dataset maker.

• Resolution: 768 (been that way for 95% of all my models)

• Number of Repeats: 14 (maybe could've done 20)

• Epochs: 3

• unet_lr: 1.5e4

• text_encoder_lr: 1.5e5

• lr_scheduler: cosine_with_restarts

• lr_schedular_number: 3

• lr_warmup_ratio: 0.05

• min_snr_gamma: checked

• network dim/network alpha: 128/64 (went off the guide a bit since the Colab has recommended that alpha is half the dim)

What I had done in my previous models was pretty much left everything as it was in the original Colab. Those gave me good results in CitrineDreamMix. Honestly, if I had a faster computer and more time, I would definitely include more sample images with MeinaMix, AniReality-Mix, and AmbientMix. I also went back to my Clementine model and checked the reviews. Some might've not gotten the armor completely right (I think it was the ranking tags that make up the armor? Been a while since I watched Overlord), and some might've lost her features like hair color and eye color (though, I should start saying add this or that if necessary from now on). One of the reviews also affected a part of her armor by removing the breastplate to see her breasts. I try my best for flexibility and customizability for my character LoRAs but I know it doesn't always meet this case whenever I try modern streetwear on a fantasy character.

Now I pretty much have zero clue on what an overbaked LoRA looks like. I downloaded more than a hundred models since I created this account 3 months ago, and most of them work pretty fine and then I just kinda forget about it afterward. The only hint I got from Googling what an overcooked/overbaked LoRA looks like is the immense sharpening it has on the image, which is the result of overtraining it or something? I should've paid more attention in my machine learning course.

The resources I got from the feedback, I'm gonna keep it. I was actually going to see if I can bake a LoRA on my laptop (the VRAM's good enough, but the time it'll take to bake isn't amazing as you can imagine). I used those parameters for here and it works fine as usual, so why not on my laptop as well right?

If you're a reader that made it all the way to the end (or you skipped it, which is fine anyway, you're not missing a whole lot if you don't know anything about training a LoRA, because I don't either!), then leave me your thoughts below! Are the models with the default parameters I worked with fine or do you think the parameters I worked with here are far superior? This shouldn't be of concern to me because I have a terabyte of free space on my hard drive (and 3 more on my other hard drive, but that's for something else entirely), but I'll probably turn back down the network dim and alpha. Helps to save some storage space, but if you want me to leave it alone, just let me know.

P.S. Since I'm writing a bunch of stuff down here, I'm just gonna announce that I'll be taking a break. Sorta. Kinda. You won't be seeing any releases from me for a while, but know that I am still gathering datasets from your requests (it's usually quick anyway, depending on how much art has been published on Gelbooru). The form is still undergoing revisions. Again, thanks for all your support!

P.P.S. Someone requested Aomine Daiki and I was planning to bake a character that shares the same voice actor as him, so... 👀

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 1.0

Trigger: cinder fall

Appearance #1: long hair

Appearance #2: short hair, hair over one eye

Sample Outfit: red dress, elbow gloves

You may add black hair and yellow eyes if necessary.

Sorry if I didn't directly contact you for help on training this character (I like to keep myself on the down low), but I can leave you with my parameters and my training set.

• This model is trained on 77 images found on Gelbooru

• Resolution: 768

• shuffle tags: checked

• activation tags: 1

• num_repeats: 4

• Epochs: 10

• train_batch_size: 2

• unet_lr: 5e-4

• text_encoder_lr: 1e-4

• lr_scheduler: cosine_with_restarts

• lr_scheduler_number: 3

• lr_warmup_ratio: 0.05

• min_snr_gamma: checked

• network dim/alpha: 32/16

Trained on 12 photos

1596 steps

1 epoch

Trigger word is: mdchnmck

Fire Emblem Engage Veyle

Anythingv3 or AbyssOrangeMix2_hard

Our favorite captain from the Nadesico. She is the last of my planned Nadesico characters as Inez Fressange and Erina Kinjo Won are already complete. Interestingly Yurika is like a yandere but with zero violent tendencies. Instead of killing your harem she would help you recruit and manage it :P The lora is trained mostly on Watanabe Yoshimasa's and Hell And Heaven's work. So support those artists if you can. Specially Watanabe Yoshimasa who has made dozens of great Nadesico doujins.

The Lora is rather forgiving and works fine with: <lora:YurikaMisumaru:0.75>, YurikaMisumaru, CaptainUniform,

It seems stable from .3 to 1.2, at 1.3 it is very interesting and somewhat artistic(not good for general use but nice in it's own way).

As for her outfits, images of her out of uniform are scarce so i didn't bother to add triggers for them only 2 are properly added:

• CaptainUniform

• CaptainHat

You may want to set her hair color and eye color as it varies and a good part of the dataset is in monochrome. Honestly I am confused if her eyes are green or blue and her hair changes from blue to purple whenever, so YMMV just pick what you like.

You might be able to possible make the lora spit her idol costume, wedding_dress and naked_apron with some coaxing as they are inside the dataset.

All in all a very nice and mild tempered LORA, I seem to have a harder time when using screencaps of 90s anime than with modern stuff or artwork. Might be all the noise and graininess they have even after filtering them. The ones I have made mostly from screencaps have been a bit of a nightmare(Erina Kinjo Won and Ifurita... I also blame their lipstick).

I'll add the tag summary with the dataset so you can experiment.

Some interesting tags from the dataset:

• white_gloves

• military_uniform

• thighhighs

• pantyhose

• garter_belt

• garter_straps

• capelet

• stockings

• pencil_skirt

• school_uniform

• wedding_dress

Trained with 50 images of chibi characters, probably not too much for a good model.

The results depends a lot on the base checkpoint that you use.

face

.

Contains two swimsuits for Theresa Apocalypse from Honkai Impact 3rd

包含崩壞三德麗莎的兩套泳裝

.

Morning Sunshine（初升晨日）:

(masterpiece:1.2), (best quality:1.3), morning_sunshine, 1girl, blue_eyes, white_hair, swimsuit, pink frilled_bikini, bikini_skirt

.

Shallow Sunset（落日淺岸）:

(masterpiece:1.2), (best quality:1.3), shallow_sunset, 1girl, blue_eyes, white_hair, straw_hat, swimsuit, dress, shorts, flower, starfish

.

The second one works better than the first one

第二套生出來的圖比較好一點

.

Although only the second suit has a straw hat in the design, sometimes it appers too when generating the first one. I think it is fine, but you can add it to negative prompt if you don't want it.

雖然在原設計中只有第二套有戴草帽，但在生成第一套的圖片時，有時候也會出現。我是覺得ok，如果你不想要的話就寫在negative prompt裡面吧

.

德麗莎世界第一可愛！

兩套服裝我都沒有，而且不知道有沒有復刻可能（悲

下次應該會上傳觀星的

.

It's an early version, may generate wired hands/penis, and has low successful rate (~20%). But when it works, it works great, with very good results. So I think it still worth to share.

Always try multiple time to get best.

(多试几次，如果有效，图片会非常不错）

A model for "licking" only blowjob, with sexy tongue of woman.

===============================================================

Please rate it if you like it, this is great help.

If you like this LoRA consider buy-me-a-coffe, it will encourage and support me to create more models like this.

https://buymeacoffee.com/xyzdali

===============================================================

Usage 用法:

licking my dick, <hair style, eyes, any other styles, background you like>

Recommend to always add some description about the hair and face, otherwise, the model will likely to generate 2 penis~

Model 模型:

Trained and tested in model: https://civitai.com/models/6424/chilloutmix

Not sure if other model works.

Recommend Weights 权重:

0.7~1

Example 例子:

licking my dick, white woman, red head, clear eyes

Negative prompt: two penies, illustration, 3d, sepia, painting, cartoons, sketch, (worst quality:2), (low quality:2), (normal quality:2), lowres, bad anatomy, bad hands, normal quality, ((monochrome)), ((grayscale:1.2)), futanari, full-package_futanari, newhalf, collapsed eyeshadow, multiple eyebrows, vaginas in breasts,holes on breasts, fleckles, stretched nipples, nipples on buttocks, analog, analogphoto, signatre, logo,2 faces

Steps: 30, Sampler: Euler a, CFG scale: 7, Seed: 331484549, Size: 512x768, Model hash: fc2511737a, Model: chilloutmix_NiPrunedFp32Fix

工藤 静香 Shizuka Kudo

Those who are not familiar with this celebrity, should google search it. A very beautiful women when she is young. And don't disturb by the front result. Check the photoes when she is young.

2023,04,23 v3.1

提升了一点泛用性，其余的和下面一样。

Google Translate:

A little generality has been improved, and the rest is the same as below.

2023,04,20 v3.0

三个衣服能正常画出来了，和服最容易(素材给多了)，另外两个难

和服：Renge-ki, kimono

私服：Renge-pc, black dress

水手服：Renge-se, serafuku

这三个衣服nsfw有点费劲，低强度描述一下人物在画。

（模型练的很强，建议小于0.5）

Google Translate:

The three clothes can be drawn normally, The kimono is the easiest (more materials are given), the other two are difficult.

Kimono: Renge-ki, kimono

Private Clothes: Renge-pc, black dress

serafuku: Renge-se, serafuku

The nsfw of these three clothes is a bit difficult, the intensity is low, describe the characters again, and then go to draw.

(The model is very strong, it is recommended to be less than 0.5)

OLD

v1.0 更像 不好换衣服，v2.0不像 更好换衣服， 强度0.6-0.8，例图用的 Cetus-Mix V3

两件衣服：kimono, dress

Google Translate:

v1.0 is more similar, but it is not easy to change clothes, v2.0 is easy to change clothes, and the strength is controlled at 0.6-0.8, the example picture is used Cetus-Mix V3

two clothes: kimono, dress

提示词：Megumin,megumin's style

可以通过加cape和hat，来让她穿上斗篷和帽子

想要换衣服的话依然是只保留yunyun，然后删除其他的词

但是由于Lora训练方式的问题，很可能一些涉及到训练集里的tag的时候就会无法彻底替换，这是没法避免的，除非欠拟合，但是这样的话很多模型跑出来的效果基本上约等于没有

有一说一，惠惠真的好可爱

tigger words: yunyun, yunyun's Style

She can be made to wear a cape and hat by adding cape and hat

If you want to change clothes, you can still just keep yunyun and delete the rest of the words

However, due to the problem of Lora's training method, it is likely that some of the tags in the training set will not be replaced completely, which cannot be avoided unless they are underfitted, but then many of the models will run out with approximately the same effect as nothing

There is one thing to say, Megumin is really cute

Ahoy me crews!

Shiver me timbers! How are ye all doing? I hope that everyone is doing well and fine this week. After long days of testing and preparing the LORA. I am happy to announce that Trieste Lora can now be used by all of you fellow shikikans. \(￣︶￣*\))

Thank you so much for all of your love and support. I have made some beautiful art with this Lora and this Lora can both be used on realistic or anime models on stable diffusion.

This Lora is equipped with 3 official costumes and 1 fanart costume that you can use. I've made this Lora with 20 Epochs and 14 epochs. You can test out which ones that you like the most. I prefer to use the 20 epochs for more strict official outfits and the 14 epochs for scenes that do not use the official costumes. YOU CAN GET THE 14 EPOCHS ON MY PIXIV FOR FREE

乗組員の皆さん！

震えろ材木！皆さんお元気ですか？今週もみんな元気で元気に過ごせますように。 LORA のテストと準備に何日も費やした後。 Trieste Lora が式管の皆さんに使用できるようになりました。 \(￣︶￣*\))

皆様の愛とサポートに心より感謝申し上げます。私はこの Lora でいくつかの美しいアートを作成しました。この Lora は、安定した拡散の現実的なモデルまたはアニメ モデルの両方で使用できます。

このローラには、公式衣装3種とファンアート衣装1種が装備されています。この Lora を 20 エポックと 14 エポックで作成しました。どれが一番好きかをテストできます。より厳格な公式衣装には 20 エポックを使用し、公式衣装を使用しないシーンには 14 エポックを使用することを好みます。MY PIXIV で 14 のエポックを無料で入手できます

I'VE PUT THE TRIGGER WORDS FOR THE COSTUMES IN MY FANBOX

>>> CLICK HERE <<<

ファンボックスにコスチュームのトリガーワードを入れました >>> ここをクリック <<<

郑重声明：本模型禁止用于训练基于明星、公众人物肖像的风格模型训练。

Hanfu Song Style Lora model, feel the Song Dynasty style!

汉服宋风Lora模型，感受宋朝风采吧！

汉服总模型：https://civitai.com/models/15365/hanfu

汉服宋风文档: https://sleepy-oyster-204.notion.site/song-style-766a8b462f654d7597164351581a4f7b

基础模型BaseModel: chilloutmix、六只粽子 sixZongzi、Basil mix、cardosAnime、meinamix、RealisticVision

Hanfu ALL model: https://civitai.com/models/15365/hanfu

Hanfu Song Style Document: https://sleepy-oyster-204.notion.site/song-style-766a8b462f654d7597164351581a4f7b

QQ交流群 149774926

微信交流群添加小助手 hanfuaigc

宋Song Tags

☑️ 长衫 long shan☑️ 短衫 short shan

☑️ green/white/pink/red/blue/black/purple/yellow

☑️ 绿/白/粉/红/蓝/黑/紫/黄 + 长衫或短衫
☑️ 浅蓝 light blue ☑️ 浅绿 light green ☑️ 黄白 yellow with white
☑️ 浅黄和白绿相间的短衫 light yellow with green with white short shan

☑️ 百迭裙 pleated skirt

☑️ white/red/green/blue/pink + pleated skirt

☑️ 白/红/绿/蓝/粉 + pleated skirt

☑️ 白金 white golden ☑️ 浅蓝 light blue ☑️ 渐变绿 gradient green

☑️ 半透明白 translucent white

☑️ 红青渐变 red cyan gradient ☑️ 红绿渐变 red green gradient

☑️ 白绿相间的百迭裙 green with white

☑️ 浅黄和白绿相间的百迭裙 light yellow with green with white pleated skirt

☑️ 宋抹 songmo

☑️ red/green/blue/white/orange/pink/yellow/purple

☑️ 红/绿/蓝/白/橘/粉/黄/紫 + songmo

☑️ 绿色碎花宋抹 green floral songmo

☑️ 紫色碎花宋抹 purple floral songmo

☑️ 青色碎花宋抹 cyan floral songmo

Thank Support Contribution 感谢以下人员的支持与贡献

• @liaoliaojun @明月明月 @南陌 @现代汉服研习小组 @小千 @善果心花 @WEN

Activation word ”alexbotez”

trained with URPM v1.3

A model of the virtual character Lu Ming from Miha Game Company，Need to be paired with an orange model to achieve better results。

Tabatha Swatosh

Tabitha in an American internet person on the TikTok and Instagram.

This is a LoRA. If you'd rather have a textual inversion of Tabatha, please go HERE.

Like my work? Buy me a beer!

Erica model

This is my first LoRA so be aware it may not functioning very well.

Somehow it works better with lower CFG scale(around 5) and I have no idea why this happens. :(

recommended strength is 0.6-0.9

More details on my ko-fi blog post regarding Niiji and the next upcoming model. For model makers, I definitely recommend trying the method I spoke about in the post on utilizing Midjourney if you're not already.

Feel free to follow me on Twitter at Tacki_NO, and if you feel like saying thank you or you want to say hello, leave me a few hearts or dm on Discord at Tacki ●﹏●#2322

For those that would like to go the extra mile and support the creation of my models and their updates, please check out my ko-fi, where I'll be posting early access content, exclusive content surrounding models, things I'm working on, and more!

All models I make will ALWAYS BE FREE.

Thank you for your time and consideration! 😀

PDM Clara

Clara is an oddly-named German internet person on the TikTok and Instagram.

This is a LoRA . If you'd rather have a textual inversion of Clara, please go HERE.

Like my work? Buy me a beer!

Brooke (Mleth)

There is not any trigger words but it does really like to make portraits unless you specify another framing/perspective.

It also pretty lewd, if you don't specify that you want clothes you often get little to none.

Niji-Expressive V2 launch!!

灵动优雅，张力尽显

本次升级泛用性，能在更多模型和采样方式上适用

选取500+优质且风格统一的Nijijourney图片，进行更高精度训练

推荐关键词：letterboxed, illustration,

目前仍然存在手部和肢体的bug（甚至更严重）

十分推荐使用“bad promt，easynegative，badhand”等负面embedding

======================

Vivid and elegant, full of tension

This upgrade is generalized and can be applied to more models and sampling methods

500+ Nijijourney pictures with high quality and uniform style were selected for higher precision training

Recommended keywords: letterboxed, illustration,

- There are still hand and limb bugs (even worse)

It is highly recommended to employ "bad promt, easynegative, badhand" and other negative embedding

Lenka Dragounova

Lenka is a former Miss Czechia, ambassador, and current internet person.

This is a LoRA. If you'd rather have a of textual inversion of Lenka, please go HERE.

Like my work? Buy me a beer!

看看ai能生成啥样的漫画，喂ai漫画看看ai会生成啥样的漫画，果然现在还是没啥逻辑，喂的是一组漫画，生成不出来剧情

This is a simple Lora, just use it with weight around 1. I've tested, and works well most of the time. Enjoy!

This model is a LORA model made using real-life models and contains a large number of cosplay photos for training. As the photos are owned by myself, there is no other risk of copyright. Please do not use them for illegal purposes!

Best at 0.5-0.7!!!

For online test, you can use my Discord Bot which has Chinese support: https://discord.com/api/oauth2/authorize?client_id=1095809481411145749&permissions=277025507328&scope=bot

Katie Sigmond

Katie is an American social media person on the TikTok and Instagram.

This is a LoRA. If you'd rather have a of textual inversion of Katie, please go HERE.

Like my work? Buy me a beer!

Sayaka Kanzaki 神崎紗耶香 | Ryoujoku Famiresu Choukyou Menu

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：red skirt, white shirt,white_bow, white_ribbon,black choker, pantyhose, cleavage,
角色：<lora:SayakaKanzaki-V1:0.85>,
头发：pink hair,long_hair,bangs,Purple_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,
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: 518423302, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original, Denoising strength: 0.5, Hires upscale: 2, Hires steps: 18, Hires upscaler: Latent

TLDR (Because I don't care)

Use star butterfly <lora:starButterflyWith_v11:1> in positiv promt. (But checkout the usage section as this LoRa provides additional outfit options ;D)

If you like the LoRa, please feel free to share your generated images here and leave a review (or at least a rating) ;D

About

Well, this week I realized that there are more than 2 sessions of this series. (I watched it years ago, and it was nice back then ^^).

So I went back and watched the first two sessions, and occasionally made screen caps of the characters, because I was thinking: I want to make LoRa available to all the characters that I care about.

I start the first batch with Star Butterfly the magic crazy girl, in most of her outfits (she changes them quite often, which is nice, considering that this is a western animation series). And the existing LoRa didn't tag the outfits properly, which is the main reason why I created my own ^^.

Additionally there will be Marco's Mom and Jacky Lynn coming up this weekend, they are currently in training :D

Training

This was fun and due to the fact I wanted to train most of her outfits (there maybe some missing) I ended with 155 images. (v1 had just 128 but I actually forgot to put two more outfit to the training data. So I created v1.1)

Additionally instead of using repeats I moved to use epochs instead... this way I have way more control over the end result ^^ (But I think I nailed it with 32 epochs ;D)

This is the first time I've specifically tagged the model for specific outfits, it worked quite well in most cases (first I used deepbooru for tagging, but there was just too much unnecessary junk, so I started from scratch, and although the dataset tag editor helped a lot, it was still a pain).

Usage

There is no need to adjust the weight (Which is quite important for me):

star butterfly <lora:starButterflyWith_v11:1>

Additionally I tagged her outfits in training so there are following (trained) options.

The outfits that did go to well:

• For all outfits you can add or remove her horn hairband and her wand but it likes to multiply a lot xD

• green squid outfit Her iconic green outfit with the pink squid on it for completion add purple striped pantyhose

• green spider outfit Her green dress with spider necklace, for completion add purple striped pantyhose

• rainbow outfit The outfit with the rainbow top

• shroom outfitOne episode outfit... but I got quite a lot of data, it even comes with a shroom hairband

• nightgown Her nightgown, sadly the moon on the gown did not stick, so there will be random motives but you can add or remove her sleep mask

The outfits that didn't go to well:

• cloud outfit The blue outfit with the cloud pouch, didn't stick that good

• princess dress Was only trained with 6 images from the opening and first scene

• bloodmoon outfit The outfit she wore on the bloodmoon ball, didn't really stick :<

• pie outfit Didn't not stick so good (was only shown in one episode, with many obstructions :/)

Maybe I'll make an updated version to get a bit more training data for the outfits that didn't go so well. I only used the animated series, but maybe fanart / the comics could solve the problem of those missing outfits ^^

animelike_2D_v2 で学習

Trained on animelike_2D_v2
Good results with 0.8-1 weigths.

鸣杆资源，摆烂补档，抓紧时间。不要船波，不会再补了。

该Lora的目的是提振亚洲美学，鼓励亚洲的创作者和鉴赏家们回归正统审美，提升个人的审美志趣。

Lora包含以下要素：

·         重视少女感，缓和的五官和带点婴儿肥的圆脸，没有过于夸张的马脸和整容锥子脸。

·         符合亚洲人的匀称身形，脱离清一色大码、重口味的鸥镁审美。

·         细腻无瑕疵的皮肤，没有大片的雀斑和皱纹。

·         纯享的审美体验，不受任何政治正确成分干扰！

·         不需要大量提示词也能生成高质量成品，可作为亚洲美学风格搭配其他lora使用。

·         示例人物为虚拟成人。请大家自重并端正自己的审美态度，不要把伦理道德问题强加给艺术活动。

个人推荐使用三次元基础模型（是否能用于二次元还未测试，感兴趣的朋友可以反馈一下），lora权重建议在0.6左右，搭配高清修复和较低的降噪，能得到更佳的使用体验。

示例（sample prompt）：

Sony a6600 Mirrorless Camera,highres,masterpiece,(photorealistic:1.34),1girl, keai, <lora:LOL_style-000014:0.6>,wavy hair,

Negative prompt: bad-picture-chill-75v, negativeembed,nude,loli,

Steps: 30, Sampler: Euler a, CFG scale: 7, Seed: 4009842147, Size: 600x768, Model hash: fc2511737a, Model: 真人_chilloutmix_NiPrunedFp32Fix, ENSD: 31337

我的抱脸账号：https://huggingface.co/BigXiang

相关问题、或者渲染出来的优秀作品，都欢迎在评论区留言反馈。

隐藏功能：后期测试后确认，在三次元大模型的基础上使用该lora后，通过调整描述人物身形特征、服装的prompt，能实现：？？感←少女感→少妇感 的覆盖。对此功能感兴趣的请自行探索，本人不会再做详述。

注意！一旦下载这个lora，即证明你自愿遵守以下条款：

使用事项：

·         本Lora仅供个人学习交流，禁止任何形式的转载、传播；禁止用于任何商业用途。

·         Lora禁止用于从事非法活动，使用时请遵守所在地的法律法规。对于Lora使用者的非法行为，本人概不负责并坚决反对。

·         不鼓励将Lora用于生成NSFW内容。

EN

This lora emphasis on soft-core oriental aesthetics, for distinct from the aesthetic pursuits of western. The generated character has soft facial features and a medium body type which is in line with the traditional aesthetic concepts of East Asia. All example characters are virtual. So please respect yourself and your aesthetic attitude, and do not impose ethical and moral issues for artistic activities. And you are welcome to show your work in the comment section.Hope you enjoy.

Attention! By downloading this lora, you certify your voluntary compliance with the following terms:

Terms of Use:

·         My Lora is only for personal learning and communication.Any form of reproduction and dissemination is prohibited; any commercial use is prohibited.

·         Use for illegal activities is prohibited, and please observe the laws and regulations of your location when using Lora. I am not responsible for and strongly oppose any illegal actions by Lora users.

·         it's discouraged for generating NSFW content.

Handsome Male LORA making mature and handsome male character in 2.5D style.

I recommend you to use the weight of this LORA between 0.7 and 1.

***

☕You can buy me coffee

If you support me, I will make many AI male models and try to improve them more beautifully.

This model got the face of Zhouhuimin and a style of 90s I think. It has a downside that doesn't work well with clothes and envirement.

Weight around 0.7 -1 is good.

映画「マトリックス」に登場した道場、仮想トレーニングルームみたいな感じ

Like the dojo/training room in the movie "The Matrix".

DOJO, sliding doors, architecture, east asian architecture, indoors, tatami, shouji, scenery, 2boys, male focus, brown hair, shorts, sunlight, running, dougi,

Kimberly "Kim" Wexler or Mrs. Goodman (also known as Giselle St. Claire) — is a fictional character from the AMC drama series "Better Call Saul". She is played by Rhea Seehorn.

I got better results with endlessReality checkpoint.

Trigger word: Kim Wexler

LoRA weight: 0.8-1.2

CFG: 8-10

CLIP Skip: 1

请不要使用该模型用于商业用途。

Please do not use this model for commercial purposes.

请注意选择版本使用。

Please pick the appropriate version to use this LoRa.

角色原型来自游戏：Blue Archive

以下内容使用机器翻译 The following uses machine translation（To English)

*注意：我在本地进行生成时使用的文件名和上传的文件名不同，请不要复制范例图中的tag。

The file name I use when I build locally is different from the file name I uploaded, so please do not copy the tag in the example image.*

角色原型来自游戏：Blue Archive

The character prototype comes from the game: Blue Archive

用合适的tag去描述符合角色原本的年龄、性别特征可能获得更好的成果。

Using appropriate tags to describe the original age and gender characteristics of the character may obtain better results.

这种LoRa应该没有提示词，请直接添加<lora:HanaokaYuzuV009:1> or<lora:HanaokaYuzuV010:1>使用。

This LoRa should have no prompt, please add <lora:isamikaedeV010:1>to tags for use it.

推荐使用的Tag/Recommended tag：

版本 V009

• 这个版本比较泛用，但是没有添加合适的标注，如果认为V010版本饱和度或者曝光度过高且难以调整，请尝试使用该版本

  This version is more general, but does not add suitable annotations, if you think that the V010 version is too saturated or exposed and difficult to adjust, try this version

版本V010/V011

• 通用/General

  <lora:HanaokaYuzuV010:0.7>,red_hair,purple_eyes, small_breasts,long_hair

• 如果认为画面饱和度过高，请进一步降低权重使用

  If you think that the picture is too saturated, please further reduce the weight usage

• 尝试出现光环 Try to make a halo appear in the picture

  halo

• 常服 /clothes

  hood,jacket,white_jacket

• 如果连续出现同一姿势的情况 请尝试在负面提示词中使用以下tag/If the same pose occurs in succession, please try to add the following tag in the Negative prompt:

  tachi-e, tachi-e_stand

敬告：之前的障壕因溪芳的郑智正确而被广理反复疯浸，故而开新号重新上床。内容暂时没有改动，诸位无需重复下载。

介绍：

站内其实已经有几套类似的lora了。但要么炼得太老，要么炼出眼疾。跑出的图总感觉怪怪的不太合我胃口，反正对颜控来说很难受就是了。于是我自己昨晚熬夜炼了一套（深更半夜强迫症发作）。这次的Lora主题为日系美少女。是的，有些像她，又有些不像她。所以命名为飞鸟集。训练集内包含三种主要风格：和服、针织毛衣、双人。

由于C站更新后无法自动识别prompt，这次我投放了大量炼出来的原图。如果你对我上传的某张样图比较感兴趣，你可以另存为后用SD自带的“图片信息”功能识别，里面自带所有的prompt参数。可以按照既有参数自行调整。具体的效果请看样图。

相关问题、或者渲染出来的优秀作品，都欢迎在评论区留言反馈。

如果您喜欢并想更多地接触到我的作品，可以在我的主页找到我往期的lora。当然，您也可以在huggingface找到我

我的抱脸账号：https://huggingface.co/BigXiang

使用事项：

注意！一旦下载这个lora，即证明你自愿遵守以下条款：

• 本Lora仅供个人学习交流，禁止任何形式的转载、传播；禁止用于任何商业用途。

• Lora禁止用于从事非法活动，使用时请遵守所在地的法律法规。对于Lora使用者的非法行为，本人概不负责并坚决反对。

• 不鼓励将Lora用于生成NSFW内容。

prompts of sample:

<lora:FB_style-000008:0.75>,keai,keai_\(keai3\),Sony a6600 Mirrorless Camera,masterpiece,highres,photorealistic,photography,cowboy shot,1girl,solo,pale skin,round face,small lips,small breasts,

Negative prompt: bad-picture-chill-75v, negativeembed,watermark,bad eyes,huge eyes,bad anatomy,(bad arms),broken arm,

Steps: 27, Sampler: Euler a, CFG scale: 7.5, Seed: 3895585265, Size: 568x776, Model hash: fc2511737a, Model: 真人_chilloutmix_NiPrunedFp32Fix, Denoising strength: 0.34, Clip skip: 3, ENSD: 31337, Hires upscale: 1.5, Hires upscaler: Latent

EN

The theme of Lora this time is a Japanese Kawaii girl. Yes, once seen the sample, you'll felt a little familiar and a little unfamiliar. That's why the name is Flying bird(飞鸟). Three main styles are included in the training set: kimono, striped sweater, and double girls.

Since this website can't automatically identify the prompt, this time I put a lot of original pictures. If you are more interested in a certain sample picture I uploaded, you can save it as and then use the picture information function to identify it, which comes with all the parameters of the prompt. You can adjust it by yourself according to the existing parameters. Please see the sample image for the specific effect and parameters.

Related questions, or rendering out excellent work, are welcome to leave feedback in the comments section.

Terms of Use:

Attention! By downloading this lora, you certify your voluntary compliance with the following terms:

• My Lora is only for personal learning and communication.Any form of reproduction and dissemination is prohibited; any commercial use is prohibited.

• Use for illegal activities is prohibited, and please observe the laws and regulations of your location when using Lora. I am not responsible for and strongly oppose any illegal actions by Lora users.

• It's discouraged for generating NSFW content.

说明 | specification：

• 原模型 | Link to original model web page:

  yden - v3.0 | Stable Diffusion Checkpoint | Civitai

• 效果更好的ckp大模型版本 | A better ckp model version：

  yden V3 TrainFix - yden-T | Stable Diffusion Checkpoint | Civitai

• 我非常喜欢这个风格的模型，但是很多我想要的效果无法表达出来。我花了一点时间去处理了一下这个模型，使其效果得到了一定的提升。

  I like this style of model very much, but many of the effects I want can not be expressed. I spent a bit of time working with the model to improve it.

• 基于此，我制作了此LoRA

  Based on this, I made LoRA

模型使用Anything V2.1 N训练 | The model was trained with Anything V2.1 N

This is LoRA of the character Alyn shir

best weight to use is from 0.6-0.8 more if you try to add more the face quality will not match the body

If you are like the models and LoRAs I create consider supporting me on Patreon

Have fun with this LoRA 😉

This model depicts the character Astro Boy or Atom from the anime "Astro Boy".

(Disclaimer: The images showcased from the model were post processed so your results may be different.)

Decided to make my own ass on wall to learn more about loras. Works pretty good with both aom2 hard and kotosmix.

Trained on aom2 hard with 68 images and thumbnails made using kl-f8 vae with hi-res fix.

You can use "Feet" to get some feet.

*New prompt "hands" but it looks weird most of the time.

Hope you can make good use of it!

This is beta version. May continue or not.

Recommended value:

Lora Weight: 0.6-0.8

Cfg scale:7-8

Sampling steps:15-30

512*512 or 512*768

***For non-commercial use only***

敬告：之前的障壕因溪芳的郑智正确而被广理反复疯浸，故而开新号重新上床。内容暂时没有改动，诸位无需重复下载。

介绍：

这次的Lora主题为少女风的卧室摄影。所以起名为：闺房

首先我自我批判一下。这不是一件优秀的成品lora，不像我之前发的两个傻瓜式lora，这个lora需要使用者对prompt有一定的调试能力。真心不推荐新手使用！

由于是清明节前一天晚上腾出时间制作的，过于仓促，挑选的训练集内侧脸图片偏多，而正脸图片较少。导致生成的女孩容易闭眼或侧脸。本来不想放出这一版，但在大量测试后发现，只需对prompt进行一些简单处理，还是能得到不错的风格化图片。所以还是决定放出来供各位学习交流。

光说不练假把式，所以具体的效果和参数请看样图。

相关问题、或者渲染出来的优秀作品，都欢迎在评论区留言反馈。

您也可以在huggingface找到我

我的抱脸账号：https://huggingface.co/BigXiang

使用事项：

注意！一旦下载这个lora，即证明你自愿遵守以下条款：

• 本Lora仅供个人学习交流，禁止任何形式的转载、传播；禁止用于任何商业用途。

• Lora禁止用于从事非法活动，使用时请遵守所在地的法律法规。对于Lora使用者的非法行为，本人概不负责并坚决反对。

• 不鼓励将Lora用于生成NSFW内容。

样图参数（Sample image prompts）:

(realistic:1.1),masterprice,keai, 1girl, dress, small breast,<lora:LDC_style-000002:0.6>, on bed,looking at viewer, (detail eyes:1),

Negative prompt: bad-picture-chill-75v, negativeembed,bad anatomy,side face,bad eyes,

Steps: 30, Sampler: Euler a, CFG scale: 7, Seed: 2980421217, Size: 568x776, Model hash: fc2511737a, Model: 真人_chilloutmix_NiPrunedFp32Fix, Denoising strength: 0.35, Clip skip: 2, ENSD: 31337, Hires upscale: 1.5, Hires upscaler: Latent

ps：不是我不愿意分享样图参数，而是c站现在不会自动读取图片的prompt，只能自己手动填写，真的没这么多精力和时间。请各位理解。如果你对我上传的某张样图比较感兴趣，你可以另存为后用图片信息功能识别。这个lora上传的样图全部是原图，里面自带所有的prompt参数。

EN

The theme of this Lora is a lady's bedroom photography. So I named this lora as “lady's chamber”

First let me be self-critical. This is not an excellent finished lora, unlike the two foolproof lora I posted before, this lora requires the user to have some debugging ability for prompt. Really not recommended for newbies!

Since it was made in the night before the Qingming Festival to spare time, it was too hasty, and there were more side face pictures and fewer front face pictures in the selected training set. The result is that the generated girls tend to have closed eyes or side faces. Originally I did not want to release this version, but after a lot of testing, I found that I only need to do some simple processing of the prompt, but still can get a good stylized picture. So I still decided to put it out for guys to communication and learn.

Please see the sample picture for the specific effect and parameters

Related questions, or rendering out excellent work, are welcome to leave feedback in the comments section.

Terms of Use:

Attention! By downloading this lora, you certify your voluntary compliance with the following terms:

• My Lora is only for personal learning and communication.Any form of reproduction and dissemination is prohibited; any commercial use is prohibited.

• Use for illegal activities is prohibited, and please observe the laws and regulations of your location when using Lora. I am not responsible for and strongly oppose any illegal actions by Lora users.

• It's discouraged for generating NSFW content.

This is a lora based on my favourite artist donaught. He/she is well known for their very thick/curvy proportions and I wanted to try and replicate it.

• works best at 0.8-1.0 strength

• Some inpainting was done of the pictures with more than 1 subject.

This model works best with abyssorange2hard

https://huggingface.co/WarriorMama777/OrangeMixs/tree/main/Models/AbyssOrangeMix2

Commission is open for model training.

patreon.com/goofyai

support me there if you like my work it inspires me to continue

also available at https://pixai.art/@goofy-ai

use lora weight of 0.7-1

use the preview image for better prompt ideas

please share and drop a review

Variant of my other Lora, dedicated exclusively to this fetish

Warning: In this case the program generates well if you ask it well, otherwise it generates badly. And to top it off there is the incredible problem of the hands, long objects apparently generate several problems and it is difficult to train them if they must be held by the hands, in any case I leave the Lora in case anyone needs it or it works. Use it with patience

To access the full uncensored version of my Loras visit:
 +) https://huggingface.co/Pepomariuomddddd

敬告：之前的障壕因溪芳的郑智正确而被广理反复疯浸，故而开新号重新上床。内容暂时没有改动，以前下载过的诸位无需重复下载。

介绍：

这次带来的是海边少女（SSS_style_v10），使用了更大的素材图片，20个批次练了3小时，成品很不错。不多说，请直接看效果。考虑到上次少钕崇拜因为不同人群喜好而在评论区引起了争议。这次的炼丹覆盖到了不同的使用受众，做了“大小”两个版本的lora。

同样的，不需要过多prompt就能得到优秀成品，荤素兼备，大道至简。触发词也一如既往可爱（keai）。

• 喜欢的话请给个五星好评作为鼓励!

• 你们的积极反馈不仅能帮助我改进我的作品，同时也是对我创作的价值的认可。所以，有相关问题、或者渲染出来的优秀作品，欢迎在文字区或图片区回复。

BTW:由于我之前贴出历史资料来反驳寒郭人的纹画稻妾行为，并且鄙夷溪芳那套虚萎的郑智正确，我已遭到馆里园多次翩毯式的疯浸，我正在逐步将往期lora补档。当最糟糕的情况发生时，你也可以在huggingface找到我的作品。

我的抱脸账号：https://huggingface.co/BigXiang

看板娘样图示范（sample prompt）：

keai, <lora:SSS_style-000018:0.6>

Negative prompt: bad-picture-chill-75v, negativeembed

Steps: 30, Sampler: Euler a, CFG scale: 7, Seed: 4231914231, Size: 560x700,

使用注意：

一旦下载这个lora，即证明你自愿遵守以下使用事项：

• 本Lora仅供个人学习交流，禁止任何形式的转载、传播；禁止用于任何商业用途。

• Lora禁止用于从事非法活动，使用时请遵守所在地的法律法规。对于Lora使用者的非法行为，本人概不负责并坚决反对。

• 不鼓励将Lora用于生成NSFW内容。

EN

This Lora is used to generate shaonv (which means "young lady" in Chinese) at the beach . Simple and efficient. Good output with less prompts. And you are welcome to show your work in the comment section. Hope you enjoy.

Prompt of example pic：

keai, <lora:SSS_style-000018:0.6>

Negative prompt: bad-picture-chill-75v, negativeembed

Steps: 30, Sampler: Euler a, CFG scale: 7, Seed: 4231914231, Size: 560x700,

Attention!

By downloading this lora, you certify your voluntary compliance with the following terms:

Terms of Use:

• My Lora is only for personal learning and communication.Any form of reproduction and dissemination is prohibited; any commercial use is prohibited.

• Use for illegal activities is prohibited, and please observe the laws and regulations of your location when using Lora. I am not responsible for and strongly oppose any illegal actions by Lora users.

• it's discouraged for generating NSFW content.

Add pointy ears to negatives to prevent elven ears. You may try these tags: mechanical arms, mechanical legs, arms behind back, bound arms, reverse pray, extra arms, four arms, extra legs, four legs but they may wont work as intended.

Fake mechanical limbs while the real ones are immobilized.

A very early version that I tried over a month ago, but failed due to a lack of images or improperly spelled tags. A total of 16 attempts and over 120,000 training steps were made. I'm posting some more or less successful models that didn't turn faces into mess. I gave up and won't be able to go back to it anytime soon, so I decided to post it as is.

I guess the only way to make such a model is to generate it yourself in a 3d editor (daz, mmd, blender, even sfm), enough material, and I do not have the next few months such an opportunity

敬告：之前的障壕因溪芳的郑智正确而被广理反复疯浸，故而开新号重新上床。内容暂时没有改动，诸位无需重复下载。

T2版本是考虑到不同人群需要做的。相对于T1版本，T2版本的生成的人物面相更成熟，体态更丰满，在其他地方并无太大区别。个人还是更推荐T1版本。

同样的，不需要过多prompt就能得到优秀成品，荤素兼备，大道至简。触发词也一如既往可爱（keai）。

• 喜欢的话请给个五星好评作为鼓励!

• 你们的积极反馈不仅能帮助我改进我的作品，同时也是对我创作的价值的认可。所以，有相关问题、或者渲染出来的优秀作品，欢迎在文字区或图片区回复。

BTW:由于我之前贴出历史资料来反驳寒郭人的纹画稻妾行为，并且鄙夷溪芳那套虚萎的郑智正确，我已遭到馆里园多次翩毯式的疯浸，当最糟糕的情况发生时，你也可以在huggingface找到我的作品。

我的抱脸账号：https://huggingface.co/BigXiang

看板娘样图示范（sample prompts）:

masterpiece,highres,(photorealistic:1.3),keai, <lora:SSS_style-000004:0.6>,nsfw,small breasts,
Negative prompt: bad-picture-chill-75v, negativeembed, (bad anatomy),(broken hand),oiled, abs,
Steps: 30, Sampler: Euler a, CFG scale: 7.5, Seed: 3476265398, Size: 560x700, Model hash: fc2511737a, Model: 真人_chilloutmix_NiPrunedFp32Fix, Denoising strength: 0.3, ENSD: 31337, Hires upscale: 1.5, Hires upscaler: Lanczos

使用注意：

一旦下载这个lora，即证明你自愿遵守以下使用事项：

• 本Lora仅供个人学习交流，禁止任何形式的转载、传播；禁止用于任何商业用途。

• Lora禁止用于从事非法活动，使用时请遵守所在地的法律法规。对于Lora使用者的非法行为，本人概不负责并坚决反对。

• 不鼓励将Lora用于生成NSFW内容。

EN

The T2 version is made for the needs of different users. Compared to the T1 version, the T2 version of the character more face more mature, more full-bodied. Personally, I still recommend the T1 version.

Attention!

By downloading this lora, you certify your voluntary compliance with the following terms:

Terms of Use:

• My Lora is only for personal learning and communication.Any form of reproduction and dissemination is prohibited; any commercial use is prohibited.

• Use for illegal activities is prohibited, and please observe the laws and regulations of your location when using Lora. I am not responsible for and strongly oppose any illegal actions by Lora users.

• it's discouraged for generating NSFW content.

General Info

This LoRA is very experimental and exists due to my desire to create a capable, gundam-focused model. Due to not having the necessary equipment to create a whole gundam checkpoint i tried this on a LoRA. Another reason i attempted this is, that the Aurora model from Luna (https://civitai.com/models/40199/aurora) is pretty capable in general so it encouraged me to try training on that one. The LoRA is simply a test so it should be reviewed with that in mind. Though i am happy for everyone who tries it out, tries to improve it or can offer what i can not to the community.

Usage

The model has been trained using Kohya so you probably want to use it in the same way as you use other Kohya models. If you are Using A1111's Webui you want to move the file into the models>Lora directory. Feel free to play around with settings etc. to eventually draw any use out of this

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

RIGHT NOW IMAGES METADATA ARE MISSING AS I'M USING COMFYUI TO SIMPLIFY THE IMAGE GENERATION. I'M WORKING ON A WAY TO FIX THIS USING A CUSTOM NODE.

What's this LoRA

This is a LoRA trained on AI generated hentai picture with a specific body type and style. My models are not trained to achieve the exact style of their source, but to change the general vibe of the image. They are also fine-tuned on anime models, so they may not work perfectly with realistic ones.

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.

The LoRA is trained on NSFW pictures.

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

Warning: In this case the program generates well if you ask it well, otherwise it generates badly. And to top it off there is the incredible problem of the hands, long objects apparently generate several problems and it is difficult to train them if they must be held by the hands, in any case I leave the Lora in case anyone needs it or it works. Use it with patience

To access the full uncensored version of my Loras visit:
 +) https://huggingface.co/Pepomariuomddddd

This is not a necessary, but If you like my Lora, please consider to donate or even Comissioning your own requested Lora on My Ko-fi

https://ko-fi.com/warrensky

https://ko-fi.com/warrensky/commissions

Please give 5 star reviews by sharing your picture creation if you like it,

• Use any checkpoint based on SD1.5

• its 768,768 image

• turn on Hires.fix (or not, please experiment with it yourself)

• Hires steps: 10-17, denoising:04-65 ish.. just try it yourself

• CGE: 4-7

• Use ERSGAN 4X+Anime 6B, DPM++ 2M Karras or DPM++ SDE Karras

• Use 20-50 sampling steps

• Strength: 0.4-1

• I suggest turn off the restore faces check mark for better results of this Lora

She might look like someone from real life, but she is not what you think. She is a totally fictional character.

Legal & Risk

It is prohibited to use this model for commercial purposes and any scenarios of illegal acts and purposes.

本人原则上不支持使用此模型生成任何成人内容。用户在使用本模型时，请务必遵守当地法律法规，勿侵犯他人的名誉权、隐私权或肖像权。

I generally do not support the use of this model to generate any adult content. When using this model, users should abide by local laws and regulations and shall not violate any other's right to reputation, privacy or portrait.

私は一般的に、このモデルを使用して成人コンテンツを生成することをサポートしません。 このモデルを使用する場合は、ユーザーは当地の法律や規制を守り、他人の名誉権、プライバシー権または肖像権を侵害しないようにしなければなりません。

原画师主页：https://www.pixiv.net/users/3921595/

A Character Lora for Jiro Yamada (山田二郎) 【Hypnosis mic】

Weight:0.6～1

Hires. fix

trigger words: yamada jiro\(hypnosis mic\)

Sub words:black hair, yellow eyes, heterochromia, green eyes, cap, hat, baseball cap

※If many moles【mole, mole under mouth, mole under eye, 】 are drawn, set it to negative. "tongue, tongue out" Even if tongue out is not specified, the appearance rate is high. If you don't want to stick your tongue out, "close mouth", etc...

Normal costume words: long sleeves, hat, jacket, pants, blue jacket, clothes around waist, multicolored jacket, black shirt, ring

※Hats usually appear even if you don't write anything. If you don't want it, make 【cap, hat, baseball cap】 negative

Such a microphone and speaker appearance words: microphone, Mc.M.B microphone\(hypnosis mic\)

1) Please use this Lora Data at your own risk. I can not be held liable or responsible for any damages caused by using this data.

2) This Lora Data may not be used for any works which contain negative content toward the original product, individual, &/or group.

3) Do not use this Lora to create pornographic and/or violent materials.(There is no problem for personal enjoyment. Please refrain from posting pornographic and/or violent depictions using this Lora on SNS or other public places.)

I am using google translate.Sorry for the details in Japanese

※以下日本語で失礼します。

運要素が絡みますがぽい衣装を出現させることは確認しています。

【注意】

共通呪文 : yamada jiro\(hypnosis mic\)

認識率を高める呪文：black hair, yellow eyes, heterochromia, green eyes, cap, hat, baseball cap

１）AIがほくろを認識するのは現状かなり厳しい気がします。意図しない位置に出現したり数が多かったりと…気になる場合は【mole, mole under mouth, mole under eye, 】カッコ内をネガティブにセットして下さい。

２）指定をしなくても舌を出してしまうことが多々あります。してほしくない場合は歯で隠すか口を閉じるなどお願いします

デフォ衣装：long sleeves, hat, jacket, pants, blue jacket, clothes around waist, multicolored jacket, black shirt, ring

※帽子は指定しなくてもかぶった状態が多いです。要らない場合は【cap, hat, baseball cap】 をネガティブに書いて下さい。

Mc.M.Bマイクと出現：microphone, Mc.M.B microphone\(hypnosis mic\)

（マイクのM.B部分細部まで学習できていません。あくまでも、ぽいもの…青いマイクが出現する程度としてご使用下さい…）

1) このLora データは、自己責任で使用してください。 このデータを使用して生じたいかなる損害についても、私は責任を負いません。

2）本Loraデータは、原作者、個人、団体等に対する否定的な内容を含む作品への使用はできません。

3) このLoraを使用してポルノや暴力的な素材を作成しないでください。(個人で楽しむ分には問題ありません。このLoraを使用してのポルノや暴力的描写をSNSなど公の場に触れるのはお控え下さい。)

To avoid the problem of improper commercial or illegal use disputes, when you start using the model, you are fully aware of the declaration document. I do not support the use of this model to generate adult content. The above model is for learning purposes only. Users must abide by local laws and regulations when using the model, and must not violate others' honor, privacy and portrait rights. I solemnly declare that the final right of interpretation of this model belongs to expired myself. If you have commercial needs, please directly contact with business through the official Weibo account of the owner of portrait rights. Without authorization, no commercial use of this model is allowed, otherwise all consequences will be taken by yourself. This declaration document does not represent the official, nor does it represent itself, but is only used for disclaimer by software author himself.

为避免不当商业或非法使用纠纷的问题，当您开始使用模型时，您已充分了解声明文件。 我不支持使用此模型生成成人内容。 以上模型仅供学习使用。 用户在使用模型时必须遵守当地法律法规，不得侵犯他人名誉权、隐私权和肖像权。 本人郑重声明，本模型最终解释权归本人所有。 如有商业需求，请通过肖像权人官方微博直接联系商业。 未经授权，不得将本模型用于商业用途，否则一切后果自负。 本声明文件不代表官方，也不代表其自身，仅供软件作者本人作免责声明之用。

不適切な商業的または違法な使用に関する紛争の問題を回避するために、モデルの使用を開始するときは、宣言文書を十分に認識してください。 このモデルを使用してアダルト コンテンツを生成することはサポートしていません。 上記のモデルは、学習のみを目的としています。 ユーザーは、モデルを使用する際に現地の法律および規制を遵守する必要があり、他人の名誉、プライバシー、および肖像権を侵害してはなりません。 私は、このモデルの最終的な解釈権が失効した私自身に属することを厳粛に宣言します。 商業的なニーズがある場合は、肖像権の所有者の公式 Weibo アカウントを通じてビジネスに直接連絡してください。 許可がなければ、このモデルの商用利用は許可されません。 この宣言文書は公式を表すものでも、それ自体を表すものでもありませんが、ソフトウェア作者自身による免責事項としてのみ使用されます。

부적절한 상업적 또는 불법 사용 분쟁의 문제를 피하기 위해 모델 사용을 시작할 때 선언 문서를 완전히 알고 있습니다. 성인용 콘텐츠를 생성하기 위해 이 모델을 사용하는 것을 지원하지 않습니다. 위의 모델은 학습 목적으로만 사용됩니다. 사용자는 모델을 사용할 때 현지 법률 및 규정을 준수해야 하며 타인의 명예, 사생활 및 초상권을 침해해서는 안 됩니다. 본인은 이 모델에 대한 최종 해석권이 만료된 본인에게 속함을 엄숙히 선언합니다. 상업적인 필요가 있는 경우 초상권 소유자의 공식 Weibo 계정을 통해 비즈니스에 직접 문의하십시오. 승인 없이 이 모델을 상업적으로 사용할 수 없습니다. 그렇지 않으면 모든 결과는 본인이 져야 합니다. 이 선언 문서는 공식을 대표하거나 그 자체를 대표하지 않으며 소프트웨어 작성자 자신의 면책 조항으로만 사용됩니다.

Virtual youtuber：Mononobe alice

Weights： 0.6~0.75

trigger word：1girl，alice

This is Mint Bramanche from Galaxy Angel

Freya from JKT48 LoRA trained on ChilloutMix. The recommended model is NeverEnding Dream (NED) as you can mix it with anime LoRA to create cosplay shot.

Trigger Word : freya jkt48 (optional for more similar face : realistic, black hair, black eyes, lips, smile/closed mouth)

Please do not use it to generate NSFW

The Lora is doing quite well, but it is something constant.

Basically it correctly generates two people in a closet, also one, although it was designed to put two people.

To access the full uncensored version of my Loras visit:
 +) https://huggingface.co/Pepomariuomddddd

Lora for making obvious boobjobs.

Also makes them more european and mature.

The model was created from 22 pictures from the video game Fire Emblem Heroes, mostly in-game pictures and artist's pictures. And was trained on AnythingV4 model from Andite.

I used the Lora Training Colab from Hollowstrawberry and my own Dataset to train the model.

The model is not designed for NSFW pictures, but if you use a NSWF-capable model, it may create NSFW pictures.

Here is a recommended sample prompt for Triandra with her standard outfit :

1girl, fairy_wings, butterfly_wings, ( head wreath vines single red_rose hair_ornament ) , hair_over_one_eye, red_hair, medium_hair, (blue eyes, facial_mark), looking_at_viewer, smile, ( hair tips green vines thorns) ,

support the artist: https://twitter.com/hikinito0902

anything v4.5 checkpoint, anything v4.0 vae, clip skip 2

let me know what you think!

LoRA for generating MLP styled brushable pony toys.

Main tags include:
G0, G1, G2, G3, G3.5, G4, G5, My Pretty Pony, Pony Life, accessories, advertisement, alicorn, brush, brushable, earth pony, filly, horn, male, packaging, pegasus, photo, pony, seapony, toy, unicorn, wings

Trained on NAI with 48 pictures (mix cosplay, screencaps,fanart). Versatile lora (anime, semi-realistic, 3D, realistic)
Best result with weight between : 0.8-1.

Prompt activator : batgirl
Other prompt : mask, green eyes, orange hair, bodysuit, cape, gloves
Wounded prompt : dirty, bruise, blood

Commissioned by ocelot1300 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.

V2版本说明。

在V1的基础上融合了

koreanDollLikeness,

aespaKarina,

aliceNikke,

JKuniform,

LORADilraba

Use with XXMix_9 or majicMIX realistic, other Base models have not been tested.

与XXMix_9或majicMIX realistic搭配使用,其他大模型没有测试过。

Cover photo using controlnet --openpose+hed封面照片使用了controlnet--openpose+hed
推荐设置Recommended settings:

• VAE: vae-ft-mse-840000-ema-pruned (For realistic models)

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

• Steps: 20 to 35

• Clipskip: 2

• Upscaler : 4x-UltraSharp

• CFG Scale : 6 to 8

ティーカップの中にキャラクターを配置します。

以下のタグの併用を推奨します。

orversized object,minigirl,cup,saucer

また、lora blockweightを使用しないと正常に出力されません。

以下の値で効果を確認しています。

ITC1:1,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0

私の作成したlora "on the desk"を使うとさらに魅力的になります。

on the desk - v1.0 | Stable Diffusion LoRA | Civitai

Place the character inside the teacup.

I recommend using the following tags together

orversized object,minigirl,cup,saucer

Also, lora blockweight must be used for proper output.

The effect is confirmed with the following values.

ITC1:1,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0

The lora "on the desk" I created makes it even more attractive.

on the desk - v1.0 | Stable Diffusion LoRA | Civitai

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.

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 MAKE SURE THAT YOUR LORA FILENAME MATCHES LORA NAME IN PROMPT.

This lora trained on remasterd and upscaled Skills Icons from WoW.
1st version probably will be a kinda buggy, but i did my best with current dataset, so it maybe will be updated in the future. (Or not, i have no idea really)

22.04.23 Update: in V2 borderless icons is default, so if you want add border just add border in the end of your prompt.

So, this lora is aimed to generate square icons in metal frame with vivid light effects.

How to use:

Your prompt should be not too complicated. Shorter = better. Trigger words is wowicon of for obects and props, wowicon portrait of for portraits. Also i highly recommed to use this structure:
wowicon of YOUR OBJECT, YOUR PROMPT, COLOR glow

Additional features (Optional):
COLOR outline, COLOR trail, COLOR round token, particles, bokeh

Enjoy and share your feedback.

Here is the Lora for POV from behind !

The trigger word is ass focus, 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

Model already understand the "bodysuit" word, but not sexy enough for me. So this is created, have fun.

===============================================================

Please rate it if you like it, this is great help.

If you like this LoRA consider buy-me-a-coffe, it will encourage and support me to create more models like this.

https://buymeacoffee.com/xyzdali

===============================================================

Usage: bodysuit, <anything, anyone, any role> you like

Trained and tested on https://civitai.com/models/6424/chilloutmix

May or may not work on other model

Just for fun, if you like this, please help to rate it, and share your results pictures with others~~

Example:

bodysuit, wonder woman
Negative prompt: illustration, 3d, sepia, painting, cartoons, sketch, (worst quality:2), (low quality:2), (normal quality:2), lowres, bad anatomy, bad hands, normal quality, ((monochrome)), ((grayscale:1.2)), futanari, full-package_futanari, penis_from_girl, newhalf, collapsed eyeshadow, multiple eyebrows, vaginas in breasts,holes on breasts, fleckles, stretched nipples, nipples on buttocks, analog, analogphoto, signatre, logo,2 faces

Steps: 30, Sampler: DPM++ 2S a Karras, CFG scale: 7, Seed: 3931114707, Face restoration: CodeFormer, Size: 512x768, Model hash: fc2511737a, Model: chilloutmix_NiPrunedFp32Fix, AddNet Enabled: True, AddNet Module 1: LoRA, AddNet Model 1: body_suit.v1.0(b9a4a2ed1e09), AddNet Weight A 1: 0.8, AddNet Weight B 1: 0.8

2023.04.22：

• 重新校对文本集，提示词的准确有所上升。

• Reviewed the text dataset again, and the accuracy of prompt words has improved.

• 增加训练步数，提升拟合程度。

• Increased the number of training steps to improve the fitting level.

2023.04.13：

这是我第一次进行LoRA训练，出现问题的地方我未必知道如何解决。

This is my first time training with LoRA, and there may be issues that I am not sure how to solve.

大就是好。

bigger is better.

🥵

Training Set：

• number of images = 100

  • close-up = 10

  • portrait = 10

  • other poses = 10

  • upper body = 10

  • cowboy shot = 10

  • full body = 10

  • other clothing = 10

  • nsfw = 30

• number of repeats = 2

• number of epochs = 17

!/bin/bash
# LoRA train script by @Akegarasu
​
# Train data path | 设置训练用模型、图片
pretrained_model="./sd-models/model.ckpt"  # base model path | 底模路径
is_v2_model=0                              # SD2.0 model | SD2.0模型 2.0模型下 clip_skip 默认无效
parameterization=0                         # parameterization | 参数化 本参数需要和 V2 参数同步使用 实验性功能
train_data_dir="./train/formidable-mix"    # train dataset path | 训练数据集路径
reg_data_dir=""                            # directory for regularization images | 正则化数据集路径，默认不使用正则化图像。
​
# Network settings | 网络设置
network_module="networks.lora" # 在这里将会设置训练的网络种类，默认为 networks.lora 也就是 LoRA 训练。如果你想训练 LyCORIS（LoCon、LoHa） 等，则修改这个值为 lycoris.kohya
network_weights=""             # pretrained weights for LoRA network | 若需要从已有的 LoRA 模型上继续训练，请填写 LoRA 模型路径。
network_dim=32                 # network dim | 常用 4~128，不是越大越好
network_alpha=16               # network alpha | 常用与 network_dim 相同的值或者采用较小的值，如 network_dim的一半 防止下溢。默认值为 1，使用较小的 alpha 需要提升学习率。
​
# Train related params | 训练相关参数
resolution="768,768"  # image resolution w,h. 图片分辨率，宽,高。支持非正方形，但必须是 64 倍数。
batch_size=2          # batch size
max_train_epoches=10  # max train epoches | 最大训练 epoch
save_every_n_epochs=1 # save every n epochs | 每 N 个 epoch 保存一次
​
train_unet_only=0            # train U-Net only | 仅训练 U-Net，开启这个会牺牲效果大幅减少显存使用。6G显存可以开启
train_text_encoder_only=0    # train Text Encoder only | 仅训练 文本编码器
stop_text_encoder_training=0 # stop text encoder training | 在第N步时停止训练文本编码器
​
noise_offset="0" # noise offset | 在训练中添加噪声偏移来改良生成非常暗或者非常亮的图像，如果启用，推荐参数为0.1
keep_tokens=2    # keep heading N tokens when shuffling caption tokens | 在随机打乱 tokens 时，保留前 N 个不变。
min_snr_gamma=0  # minimum signal-to-noise ratio (SNR) value for gamma-ray | 伽马射线事件的最小信噪比（SNR）值  默认为 0
​
# Learning rate | 学习率
lr="1e-4"
unet_lr="0.35e-4"
text_encoder_lr="0.3e-5"
lr_scheduler="cosine_with_restarts" # "linear", "cosine", "cosine_with_restarts", "polynomial", "constant", "constant_with_warmup", "adafactor"
lr_warmup_steps=0                   # warmup steps | 学习率预热步数，lr_scheduler 为 constant 或 adafactor 时该值需要设为0。
lr_restart_cycles=1                 # cosine_with_restarts restart cycles | 余弦退火重启次数，仅在 lr_scheduler 为 cosine_with_restarts 时起效。
​
# Output settings | 输出设置
output_name="formidable-mix"     # output model name | 模型保存名称
save_model_as="safetensors"      # model save ext | 模型保存格式 ckpt, pt, safetensors
​
# Resume training state | 恢复训练设置
save_state=0 # save state | 保存训练状态 名称类似于 <output_name>-??????-state ?????? 表示 epoch 数
resume=""    # resume from state | 从某个状态文件夹中恢复训练 需配合上方参数同时使用 由于规范文件限制 epoch 数和全局步数不会保存 即使恢复时它们也从 1 开始 与 network_weights 的具体实现操作并不一致
​
# 其他设置
min_bucket_reso=256              # arb min resolution | arb 最小分辨率
max_bucket_reso=1024             # arb max resolution | arb 最大分辨率
persistent_data_loader_workers=0 # persistent dataloader workers | 容易爆内存，保留加载训练集的worker，减少每个 epoch 之间的停顿
clip_skip=2                      # clip skip | 玄学 一般用 2
​
# 优化器设置
optimizer_type="Lion" # Optimizer type | 优化器类型 默认为 8bitadam，可选：AdamW AdamW8bit Lion SGDNesterov SGDNesterov8bit DAdaptation AdaFactor
​
# LyCORIS 训练设置
algo="lora"  # LyCORIS network algo | LyCORIS 网络算法 可选 lora、loha、lokr、ia3、dylora。lora即为locon
conv_dim=4   # conv dim | 类似于 network_dim，推荐为 4
conv_alpha=4 # conv alpha | 类似于 network_alpha，可以采用与 conv_dim 一致或者更小的值
dropout="0"  # dropout | dropout 概率, 0 为不使用 dropout, 越大则 dropout 越多，推荐 0~0.5， LoHa/LoKr/(IA)^3暂时不支持

ポケモンのカガリです　ちょい不安定気味　エンゲイジしたい人向け

ファイル名が置き換わってしまうためzipで投稿します　トレーニングファイルになっていますが問題ありません

出力する際は添付されているテキストファイルを参照してください

I will post it as a zip because the file name will be replaced It is a training file, but there is no problem　

Please refer to the attached text file when outputting.

FGOの牛若ちゃんです　ぽんぽこしたい人向け　パンツはちょっと不安定

Kumbhira from Granblue Fantasy, trained on the animefull-final model for 8 epochs at a network dim of 64, which then got resized to one of 16.

The following tags were used during training:

Character: kumbhira

Default outfit: white coat, camisole, shorts

Summer: halterneck bikini

Event: open jacket, white sweater

The event outfit isn't particularly accurate due to a small dataset but it will do. If you want the fur coat to go over the chest it helps to put emphasis on the "white coat" part of the prompt.

萧薰儿漫画版

底层模型时 yesmix_16，vae: vae-ft-mse-840000 , 权重 0.6 - 0.8

Get early access to build and test build, be able to try all epochs and test them by yourself on Patreon or contact me for support on Discord

To support me you can buy me a coffe at https://ko-fi.com/antonioriolo

Training settings:

33 images, 25 steps, 9 epochs, you find many epochs (with test images using plot xy script) on my Patreon while supporting me.

Recommended settings:

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

• Sampler: DPM++ SDE Karras (Recommended for best quality, you may try other samplers)

• CFG Scale : 5 to 9

• LORA weight for txt2img: reccomended 0.5-0.8

• Keys: wearing sexy cop uniform

I'm not very happy about this model. It doesn't work as i expected.

First advice i give you it's to put a big weight on the keyword, like "(wearing sexy cop uniform:1.5)" and use always the model with atleast a 0.5-0.6 weight or you will see a bikini girl with a cop cap hat.

You can prompt something like "blue uniform" or "black uniform" to have different types of cop uniform (other then different colors)

One other thing that have worked is to start the prompt with "a woman next to a police car". As you can see from some example images.

I also tryied to had handcuffs and cop glasses as keywords but they do not seem to work, atleast i didn't manage to let them work. Handcuffs keyword produce nonsense, cop glasses produce normal glasses, even if there are many training images with cop glasses.

My mistake might be that i had to do more steps to train something like this, but i'm not sure that it could give a better result.

Anyway, problems aside, if you follow the weight instructions, you will end up with some good sexy cop images, Enjoy!

If you'd like to support my efforts and have access to early and bleeding edge builds, please think about joining my Patreon: https://www.patreon.com/betterCuteAsians

Note: this is ealiy version, not very good on pussy shape/colors!!!

If you care that details, try multiple times with different weights.

===============================================================

Please rate it if you like it, this is great help.

If you like this LoRA consider buy-me-a-coffe, it will encourage and support me to create more models like this.

https://buymeacoffee.com/xyzdali

===============================================================

Usage:

ridding a dildo, <any girl, any style, any role you like>

Recommend weights: 0.8~1

Recommend model: chilloutmix

Trained and tested on https://civitai.com/models/6424/chilloutmix, not sure how other models perform.

Example:

rilding a dildo, young white woman, big breasts
Negative prompt: illustration, 3d, sepia, painting, cartoons, sketch, (worst quality:2), (low quality:2), (normal quality:2), lowres, bad anatomy, bad hands, normal quality, ((monochrome)), ((grayscale:1.2)), futanari, full-package_futanari, penis_from_girl, newhalf, collapsed eyeshadow, multiple eyebrows, vaginas in breasts,holes on breasts, fleckles, stretched nipples, nipples on buttocks, analog, analogphoto, signatre, logo,2 faces

Steps: 30, Sampler: DPM++ 2S a Karras, CFG scale: 7, Seed: 3915039207, Face restoration: CodeFormer, Size: 512x768, Model hash: fc2511737a, Model: chilloutmix_NiPrunedFp32Fix

Generalization to item other than dildo :

Change the prompts to: riding a <some object you like>, ...

Weights should be lower, about 0.3~0.5 ish to ride other items than dildo, because the trainining data is mostly dildo.

Byproducts of the learning process

The style is relatively closer to that of human painting

犯大吴疆土者，盛必击而破之！

这是“三国杀"中稀释珍宝界徐盛的专属lora，建议权重0.8

由于大宝原画实在不多，模型效果一般，但绝对值得一试哈哈

可以试试隐藏的词条例如：smile, teeth, open_mouth，收获阳光开朗大男孩儿

强烈建议融合下面模型：https://civitai.com/models/16142/ow-hanzov10（权重0.3）

添加词条hanzo, frowning

另外可融合模型：https://civitai.com/models/16346/chinese-armor-zhajia（权重0.15）

其它训练细节可点击gallery图片查看

下载完记得去给√卡打差评！

Please give a negative review on The Killers of Three Kingdoms!

Hello everyone, this is going to be my first character pack, in this case I have decided to make each and every one of the mothers of Pokemon, both from video games and anime, if you like the pack don't forget to support me on patreon or donating me via crypto: patreon.com/TecnoIA bc1qcka5rhskts5ap8l72sfw96a5vzcv9mk4jg6ggg (BTC) 0x7C6BE02BF18027b46032f20e6d1bF7Ca5541932B (ETH)

欢迎使用武墨模型，这是一个武侠与墨风交融的Lora

==========

Welcome to the Wumo model, which is a Lora that blends martial arts and ink style

注意！大部分预览图使用了( https://github.com/ototadana/sd-face-editor.git )插件对脸部进行了自动修复( 插件的介绍与使用：https://www.bilibili.com/video/BV1RN411w7LQ/ )

1代武墨推荐权重：0.7-1.3，适用于2.5D与2D模型
特别推荐国风3.3( https://civitai.com/models/10415/3-guofeng3 )

==========

1st generation Wumo recommended weight: 0.7-1.3, suitable for 2.5D and 2D models

Special Recommendation: National Wind 3.3（ https://civitai.com/models/10415/3-guofeng3 )

Hello everyone, this is going to be my first character pack, in this case I have decided to make each and every one of the mothers of Pokemon, both from video games and anime, if you like the pack don't forget to support me on patreon or donating me via crypto: patreon.com/TecnoIA bc1qcka5rhskts5ap8l72sfw96a5vzcv9mk4jg6ggg (BTC) 0x7C6BE02BF18027b46032f20e6d1bF7Ca5541932B (ETH)

Recommended settings is inside the example pictures, recommended vae is sd-840k. Should work fine with AOM2 and Based64 mixes.

Use "chibi" inside the prompt to get more miniatured character if you want.

Hello everyone, this is going to be my first character pack, in this case I have decided to make each and every one of the mothers of Pokemon, both from video games and anime, if you like the pack don't forget to support me on patreon or donating me via crypto: patreon.com/TecnoIA bc1qcka5rhskts5ap8l72sfw96a5vzcv9mk4jg6ggg (BTC) 0x7C6BE02BF18027b46032f20e6d1bF7Ca5541932B (ETH)

weight of 0.6~1.0

Prompt：clementine, 1girl, solo, blonde hair, short hair, armor, smile, gauntlets, (symmetrical clear eyes red eyes:1.2)

Sometimes you may get a weapon, so try to describe the weapon in the negativeprompt.

もうちょっと煮詰めました。マックの顔ばかり見すぎて似てるか似てないかよくわからなくなってきました…ひとますこれで完成とします

Updated to version 3.5. mayibe it is last version...

Main-model is "meinamix_meinaV8.safetensors".

Prompt is "mejiromac, 1girl, <lora:mejiromac3_5:(1~0.6)>,......"

maybe "1" is too strong effort, "0.7~0.6" is better.

Hello everyone, this is going to be my first character pack, in this case I have decided to make each and every one of the mothers of Pokemon, both from video games and anime, if you like the pack don't forget to support me on patreon or donating me via crypto: patreon.com/TecnoIA bc1qcka5rhskts5ap8l72sfw96a5vzcv9mk4jg6ggg (BTC) 0x7C6BE02BF18027b46032f20e6d1bF7Ca5541932B (ETH)

Misako Amamiya 雨宮美沙子 | Enbo 艶母

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：puffy_short_sleeves,White_shirt,yellow_apron ,
角色：<lora:MisakoAmamiya-V1:0.8>,housewife,<lora:odaNonStyleLORA_v2:0.2>, oda Non,
头发：Black_hair,Brown_eyes,single_braid,
角色鎖定：
1 girl, Manga,25yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,
眼神表情：
looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),
畫質:
official art,extremely detailed CG unity 8k wallpaper, perfect lighting,Colorful, Bright_Front_face_Lighting,
(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,
photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2)
身材:
(Beautiful,large_Breasts:1.4), (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: 2561000958, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original, Denoising strength: 0.5, Hires upscale: 2, Hires steps: 18, Hires upscaler: R-ESRGAN 4x+ Anime6B

This LoRA is trained on 167 high-quality flat-style images of Twilight Sparkle. All pictures were filtered and edited for the best results. This LoRA is perfect for solo Twilight image generation. Use it with the base model Pony Diffusion v4.
You can find the dataset here.

This is a Kinomoto Sakura lora model trained from official arts. All training images were detaily tagged so you could control the costumes.

Trigger words

kinomoto sakura is the primary trigger word. There are some optional trigger words as below, please increase their weights if the effect is not obvious:

• tomoeda primary school uniform: wear tomoeda primary school uniform.

• battle costume: wear battle costume.

• middle school student: use middle school student body proportion as in Clear Card Arc.

• sealing wand: hold sealing wand

Hairstyle

Generally sakura in generated images will NOT have any hair ornament, to add them, or remove them, please add one of the following prompts to the positive prompts or negative prompts:

• two side up

• hairband

Notes

If sakura in generated images is too mature, please add female child to povitive prompts, and adjust its weight gradually.

需要带上 twintails 词条，否则会生成很长的马尾，带上才能生成合适塞拉长度的马尾（可以调整权重来控制马尾长度）

You can get an idea about the specifics of this LoRA by looking at example images.

The base model is Suzumehachi.

JAV Model Tsubomi. Works best with cyberrealistic_v13.

You don't need a trigger word to get this model working, however I used "Tsubomi" in the training.

Hope you like it, and I am waiting for your feedback

V2 Description

This version is in my experience much superior to V1. When I trained V1 I haven't put much effort, and since my dataset was small anyway I was content with meh result. Here I've used same set on training images but I've put much more effort into it.

Important thing to note is that this model outputs can differ in style quite a lot even with same prompt and different seeds. They can range from semi-realistic to old school. Personally I love the old school style and I found out you can control it fairly consistently by using retro artstyle tag. My preview images have all been made with weight 0.9 and with retro artstyle put either in prompt or negative prompt, so you can see what a big different in makes. Whichever you prefer, I advice to put it in one of two prompts for more consistent results.

Also retro artstyle have more accurate dark skin, without it it can be a bit lackluster.

Apart from that, prompt is pretty much the same - urd \(aa megami-sama\), facial mark, forehead mark, elbow gloves, choker, bracelet, cape, purple dress, high heels

Cape often seems to need more weight, but that's about it.

V1 Description

Lora model of Urd from Aa Megami-Sama. Previews were made with weight 0.85.

I've been spending too much time on SD so now I wanted to make something simplier and low effort and this is it. There's one outfit and have to struggle a bit for more unusual poses. Details are not very accurate because training set was small. Still for all it's faults, can make nice images also big booba important.

Prompt for right outfit is urd \(aa megami-sama\), facial mark, forehead mark, elbow gloves, choker, bracelet, cape, purple dress, high heels

Pretty thick with big breasts woman Lora model,

Enjoy, have fun.

A LoRA that generates girls in positions for twister game

The dataset contains mostly 1girl with various angles, with a few 2girls ones too. So it's unlikely to generate good results when you prompt for 3+ girls.

Yura Onoda 小野田 優良 | Futari Ecchi ふたりエッチ 夫妻成長日記

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：Pink_maid,Pink_maid_apron, maid headdress,
角色：<lora:YuraOnoda-V1:0.85>,maid,<lora:odaNonStyleLORA_v2:0.2>, oda Non,
头发：Brown_long_hair,wavy_hair,Brown_eyes,(Bare_forehead:1.2),(center_parted_Bangs:1.2),
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),
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: 3610504128, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original, Denoising strength: 0.5, Hires upscale: 2, Hires steps: 18, Hires upscaler: Latent

底模是原版7G,每个图片的提示词都写好了,点击图片就能查看

Click on the image to view prompt

Lora的触发词是：pandora \(mega man\)

Lora tag is:pandora \(mega man\)

A request from @cold4949494939391

It's my mid-term exam tomorrow bruh

Anyway I just left my computer working and made it

No time for making better example pictures sry

Please leave comments if you want to request a model (well , to be honest, I just want to create models of characters from《BanG Dream》and 《Project Sekai》now)

在使用c站一段时间之后我发现简介还是要用英文写的（）

三幻甘宁模型，因为训练素材过于少，所以会导致邪魅张嘴和绑定头巾，头巾可以通过分层控制去除，张嘴目前只能抽卡，权重不应高于0.75。触发词请务必打上！

For the Sanhuan Ganning model, because there are too few training materials, it will cause Xiemei to open her mouth and bind her turban. The turban can be removed through layered control. Opening her mouth can only draw cards at present, and the weight should not be higher than 0.75.Be sure to type in the trigger words!

Sanhuan Ganning モデルの場合、トレーニング用の資料が少なすぎるため、Xiemei が口を開いてターバンを結びます.ターバンはレイヤード コントロールを使用して取り外すことができます.彼女の口を開くと、現在カードを引くことしかできません。 0.75 を超えてはなりません。必ずトリガーワードを入力してください！

Sanhuan Ganning 모델의 경우 교육 자료가 너무 적기 때문에 Xiemei가 입을 벌리고 터번을 묶을 수 있습니다.터번은 레이어드 제어를 통해 제거할 수 있습니다.입을 열면 현재 카드를 뽑을 수만 있으며 무게는 0.75보다 높지 않아야 합니다.트리거 단어를 입력하십시오!

Honestly, this model was and is a pain. There is yet again not as many images in the dataset as i'd have liked. It will prompt and work but the quality can be less than great at times. Her hair ribbons have a habit of messing up and her school uniform likes to have a blue collar from time to time.

It obviously is a hentai, so NSFW.

I might revisit this model in future. You can use visual_novel in the prompt to push it be in that style, which honestly usally looks better than the screens from the actual hentai.

Art style lora of the video game 'Blue Archive'.

recommend strength:0.5~0.7

recommended weight: 0.7-0.9

trigger words: long pony tail, up pony tail, white cape, black short pants

略

Form for requests.

I also sometimes do polls on my Patreon for the next LoRA (no payment required).

A weight of ~1.0 is recommended.
Feel free to experiment with other weights and share your results if you want to!

The main prompt is "Otonashi Kotori"

Use "short hair" for her hair style.
You can try to get her mole with "mole" or "mole under mouth", but honestly I wouldn't bother... The ai is terrible with moles and you are probably better off just adding it yourself in photoshop.

For her outfit you can try the following prompt:
"Otonashi Kotori, short hair, hairband, thighhighs, black skirt, pencil skirt, yellow bowtie, white shirt, green vest"

Let me know if you have any issues.

Used embeddings for sample pictures:

EasyNegative

bad_prompt_version2

If you have any feedback or questions just write in the comments or hit me up on reddit or discord.

better with: <lora:charlotte3:0.95>, two-tone hair, white hair, black hair, twintails, orange eyes, fake horns, pointy ears

loli, small breasts by taste

my twitter: @AdskijDrocer

The following settings are recommended
Checkpoint : MinaiMix
Positive prompt: (masterpiece), best quality,1girl,seran,dress,twintails ,seiran (touhou),frilled dress,blue hair,Rabbit ears,red eyes,
Negative prompt: EyasNegative
weight: 0.7~0.8, CFG Scale: 4~8

just for test

I want to try to train her to fit my image.

我想试着训练一个符合我心中形象的她

"masterpieces,detailed,detailed eyes,mature female,milf, ((white skin)),(pale skin:1,.3),tall female,semiramis \(fate\),black dress, black fur trim, off shoulders,bare shoulders,cleavages,bridal gauntlets, ((masterpiece)) ,<lora:semi-08:1>,(smirk:1.2),black hair,long hair, yellow eyes,pointy ears, (slit pupils) "

For Test Beta！

train：use 78pics Zen Card train style. 使用78张禅卡手绘牌训练风格，如果你有了解过禅卡

advice：use 2D or 2.5D model 建议使用2d或2.5d模型

lora weight: advice 建议0.4-0.8 ，过高会影响面部细节和眼睛，因为原画风眼睛都是闭着较多。0.9-1画面会很“神圣”，当然还有很多奇怪的画面，原版特有的画风也会显现出来，可以使用负面词控制，或者使用一些其他lora去对冲这种效果，个人喜欢用哈尔移动城堡。

使用以后画面会给人心灵更宁静安详的感觉。建议搭配剪纸lora增强画面立体感。挑选了几个不同的模型和权重对比，下面分别是0，0.4，0.8 ，1以及叠加剪纸lora效果

Get early access to build and test build, be able to try all epochs and test them by yourself on Patreon or contact me for support on Discord

To support me you can buy me a coffe at https://ko-fi.com/antonioriolo

Training settings:

42 images, 15 steps, 10 epochs, you find many epochs (with test images using plot xy script) on my Patreon while supporting me.

Recommended settings:

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

• Sampler: DPM++ SDE Karras (Recommended for best quality, you may try other samplers)

• CFG Scale : 5 to 9

• LORA weight for txt2img: reccomended 0.3-0.8 (best 0.5)

• Keys: wearing sexy maid dress

This model is very easy to use, just use the keyword and if necessary give to it a weight.

If you don't specify the color it will be black and white as the majority of the images of the training. but you can start your prompt with something like "a woman, (wearing sexy maid dress:1.2), red dress" to choice a particular color.

You might also want to try different weights for the lora, i used 0.5 and it worked great. with higher weight it may be more accurate in the dress but you risk to get artifacts in your images or wierd faces.

Enjoy!

If you'd like to support my efforts and have access to early and bleeding edge builds, please think about joining my Patreon: https://www.patreon.com/betterCuteAsians

weight 0.6 is best.

I created this LoRA in response to a request.

akali, 1girl, solo, k/da \(league of legends\), (black hair:1.1), blonde hair, (multicolored hair:1.1), jewelry, fingerless gloves, midriff, earrings, two-tone hair, open jacket, black pants, crop top, blue eyes, belt buckle, ponytail

Depending on the differences in your model, you may need to modify the weight of the hair color.

weight 0.6 is best.

I created this LoRA in response to a request.

chl1, 1girl, braided_ponytail, (ahoge:1.1), school uniform, white dress, collared dress, collarbone, eyelashes, short sleeves

Add a twintails to the negative prompt.

When creating her image, her hair is produced in twintails with a low probability.

这是一个用于生成胶片风格AI写真照片的LoRA。与MoonMIX或Chilloumix搭配使用，可以生成逼真的胶片风格照片。

This is a LoRA for generating film style AI portrait photos. when used with MoonMIX or Chilloumix, it can generate realistic film style photos.

我常用的CFG范围是5~7，Lora权重通常在0.5~0.7。但也不一定这个参数组合效果就最好，大家可以摸索摸索多试试。

My usual CFG range is 5~7, and Lora weight is usually 0.5~0.7.

有时我也会搭配一些其他人像模型一起使用，效果也蛮不错。

Sometimes I also use it together with some other image models, the effect is also quite good.

by 豹仙人Leopa-sama

榛名は大丈夫です！

Read this before using the model I trained.

My one and only WAIFU. Trained on NAI.

Yeah, I know you can generate Haruna without using LoRA, but FFS that's very unstable and won't work on any mixed model newer than AOM2.

This is a test on expmix:

Now you get what I mean, right?

Trigger Words:

• Weight: 0.8-1

• Base:

  • Haruna \(Kancolle\), black hair, hairclips, gold eyes,

    • brown eyes maybe, but I like gold ones. If the elements mess up, try to lower Haruna \(Kancolle\) weight to 0.5-0.8 first.

• OGoutfit:

  • HarunaOG, headgear, hairband, nontraditional miko, bare shoulders,

    wide sleeves, detached sleeves,

    waist bow, red skirt, hakama short skirt, pleated skirt, thighhigh, thighboots,

    • other words like sarashi, sideboob, zettai ryouiki can be used for typical angle.

• Rain Season Outfit:

  • HarunaRain, white hairband,

    (green shirt:1.5), short sleeves,

    white skirt, pleated skirt,

    • bit overfitted, will generate umbrella sometime even put it in negative prompts, if it always happen, then change your base model to those prompt-based model like Anyv3/v5 or ExpMix. (Yeah, maybe those prompt-free model is easy to use, they are usually not good for LoRA.)

• Causual Outfit (Mitsukoshi Collabo):

  • HarunaCausual, white hairband, anchor necklace, jewelry,

    (white jacket, open jacket:1.2), black shirt,

    long grey skirt, black pantyhose, black boots,

    • in some model it may mess up with the outfit below, you should try add (hoodie, open hoodie:1.5) in negative prompts.

• Summer Outfit:

  • HarunaSummer, headgear, hairband, cleavage, collarbone,

    (white bikini, front-tie bikini:1.3),

    (white sarong:1.3),bare legs,

    • if you want the hoodie (the in-game illust) , replace the second line withwhite hoodie, if you keep both of them for half-open hoodie with bikini, it will mess up very often. I don't know how to fix it, but seems like topless will help.

• Christmas Outfit:

  • HarunaSanta, headgear, hariband, santa costume, breasts, cleavage, red capelet, red dress, fur-trimmed dress, fur-trimmed capelet,

    fingerless gloves, red gloves,

    belt, red leg wear, bare legs,

    • not very stable. The lower dress will sometime mess up with OGoutfit's pleated skirt. Add pleated skirt, skirt, maybe help.

Shiona from Harmonia

Don't even try getting the outfit, despite being thoroughly trained on the outfit, it just doesn't work. I tried. The rest of her is fine though.

Character: a-chan あーちゃん先輩 from game/anime Little Busters! and Kud Wafter

weight:0.65 use hires fix:0.5~0.58

tag "dress" for another outfit

用 "dress" tag有另一套衣服

A LoRA of Tojo Kirumi from Danganronpa v3. Trained on NAI.

See samples for good prompts.

If you like this LoRA, please let a review! I'm planning to do every female Danganronpa characters so tell me who you wants first!

All of my LoRAs will always be free. If you paid for this or an image using this LoRA, you got scamed! If you wants to supports me, feel free to buy me a Ko-Fi.

No.1xx 是一些我完全不认识的人，个别知道的我或许也因为懒得改名就放在这个系列里。

No.2xx 大概会是我个人喜欢的明星或者博主。

No.4xx 是一些概念。

lora强度推荐0.6-0.7。chilloutmix出图，触发词为1girl，推荐配合close-up和upper body词条使用，或者分层控制。

词条看示意图，负面提示词我用了EasyNegative：EasyNegative | Stable Diffusion TextualInversion | Civitai

素材来自网络，仅有五张，拆解为十张训练，非常极限。

请及时下载保存，模型随时可能被删除。

你的每一个五星都是我更新的动力。

Every 5-star review of yours motivates me to update.

公共频道：t.me/dawn66666666

这是一款基于《虚构推理》动漫中的女主角岩永琴子设计的LoRA模型。底模采用了sd1.5。

在本次训练中，我们使用了大约40张左右的图片。我们建议在0.6-0.7的权重范围内进行训练以达到最佳效果。使用dpep2 LoRA可以给模型增添一些特色效果。

您可以在auto1111中选择此模型，并将其复制到stable-diffusion-webui/models/lora中使用。请确保更改权重，因为默认值为1，这通常会导致过度拟合。

我们会继续改进和优化该模型，如果有更多的人使用它，我们会不断加入更多的特征和功能。

最后，感谢您对该模型的支持和使用。

This is a LoRA model based on the main character Kiki Iwanaga from the anime "Fictional Detective". The model uses sd1.5 as the base and has been trained on approximately 40 images.

We recommend training the model at a weight of 0.6-0.7 for optimal results. Adding dpep2 LoRA can give the model some unique effects.

To use the model, you can select it in auto1111 and copy it to stable-diffusion-webui/models/lora. Make sure to adjust the weight, as the default value of 1 often leads to overfitting.

We will continue to improve and optimize the model, and more features and functions will be added if more people use it.

Thank you for your support and use of this model.

No.1xx 是一些我完全不认识的人，个别知道的我或许也因为懒得改名就放在这个系列里。

No.2xx 大概会是我个人喜欢的明星或者博主。

No.4xx 是一些概念。

lora强度推荐0.8-1.0，我使用1.0。chilloutmix出图，触发词为kigurumi，这是第一版，之后还会继续更新。

词条看示意图，负面提示词我用了EasyNegative：EasyNegative | Stable Diffusion TextualInversion | Civitai

素材来自网络。

请及时下载保存，模型随时可能被删除。

你的每一个五星都是我更新的动力。

Every 5-star review of yours motivates me to update.

公共频道：t.me/dawn66666666

No.1xx 是一些我完全不认识的人，个别知道的我或许也因为懒得改名就放在这个系列里。

No.2xx 大概会是我个人喜欢的明星或者博主。

No.4xx 是一些概念。

lora强度推荐0.6-0.8。chilloutmix出图，词条看示意图，负面提示词我用了EasyNegative：EasyNegative | Stable Diffusion TextualInversion | Civitai

素材来自网络视频截图。

请及时下载保存，模型随时可能被删除。

你的每一个五星都是我更新的动力。

Every 5-star review of yours motivates me to update.

公共频道：t.me/dawn66666666

简单的训练了一下，背景可能不是很好。推荐0.4-0.6

常服：sundress

泳装：bikini，bikini under clothes

校服：black serafuku,school uniform

睡衣：pajamas

I broke my Gundam again, can i sit in your cockpit instead?

General Info

This LoRA is the first that i trained on the Aurora model from Luna (https://civitai.com/models/40199/aurora) and the results are stuunning! It does an amazing job of generating different style images of Christina Mackenzie from Mobile Suit Gundam 0080: War in the Pocket. Most of the sample images have been generated with Aurora too, a few have been generated using Fidelity too.

Usage

The model has been trained using Kohya so you probably want to use it in the same way as you use other Kohya models. If you are Using A1111's Webui you want to move the file into the models>Lora directory. Make sure to use the corresponding trigger if you want her to have the correct hair color and get optimal results.

你可以在这个网站赞助我，你的赞助就是我创作的激励

You can sponsor me on this website, and your sponsorship is my motivation for creating

https://afdian.net/a/NovitiateAlchemist

• Drawing NSFW images is better，It needs to be paired with other lora

• 画瑟图效果更好，需要搭配其它lora

• 两套服饰：meidusa，nude（需要搭配meidusa）

• Two outfits: Meidusa, Nude (needs to be paired with Meidusa)

Update on v0.5/更新版本v0.5:

Rearrange the tag priority to improve the training output, now training different outfits in separate files to improve the uniformity of the single image. The generalization has greatly increased and can be used with other lora. Trained with Anythingv5, works fine with AnythingV5 and setusMix, please try for yourself for other models.

将tag的优先度进行了调整以提升训练输出的效果，并将不同的衣服分开训练以提高单个文件的统一度。目前版本的泛化程度大大提高，可以与其他lora更好的同时使用。基于AnythingV5训练，对AnythingV5以及setusMix使用的效果良好，您可以自由尝试和其他的模型组合。

Suggest weights/建议权重: 0.3~0.5

Trigger Words/触发词：hiseki erio, Erio_casual/Erio_yoshiwara/Erio_show (Depending on the model/第二个触发词基于模型会有所不同）

Suggested Prompts/建议咒语:

Casual Clothes （经典服）:

solo, 1girl, hiseki erio, Erio_casual, green_eyes, red_hair, single thighhigh, garter straps, uneven legwear, collarbone, breasts, cleavage, hair_ornament, black_tank_top, denim_shorts, off_shoulder, black thighhighs, navel, stomach, striped, striped_thighhighs, short_hair, green jacket

Japanese Clothes (和服）：

solo, 1girl, hiseki erio, Erio_yoshiwara, green_eyes, red_hair, collarbone, breasts, japanese_clothes, sarashi, bondages, cleavage, hair_ornament, hair_flower, flower, earrings, skirt, floral print, off_shoulder, black thighhighs, green skirt, breast bondage, obi, sash, kimono, sarashi,zettai ryouiki, yellow flower,

Show Clothes （演出服）：

solo, 1girl, hiseki erio, Erio_show, green_eyes,red_hair, collarbone, breasts, cleavage, gloves, black gloves, fingerless gloves, green_skirt, shorts, navel, thighhighs, high_ponytails, ponytail, hair ornament, gold_striped thighhighs, lace-trimmed legwear, white belt, belt

Update on v0.4 / 更新版本v0.4:

Trained with better tagger and improve the details of different clothes, setting weight at 0.5~0.7 would be good. Check the trigger words from each clothes on the sample images.

使用了更好的标签工具并提升了每一件衣服的细节，0.5~0.7的权重较为适合，具体的每一个触发词可以看对应衣服的范例。

This lora model is trained based on the original authors of the OrangeMix and AbyssOrangeMix, by downloading and using this model, you are agreeing to the CreativeML OpenRAIL-M and any further licenses applied from the OrangeMix. Please read the full license here.

This work is only trained with anime figures, and no real person data is included during the training process. Please respect the people and local laws that apply when generating or publishing your work based on this model on the internet.

这个lora模型是基于OrangeMix 中的AbyssOrangeMix 进行训练的。下载并使用此模型，意味着您认同 CreativeML OpenRAIL-M 协议，以及其他在OrangeMix作者追加的使用协议。完整的协议，请阅读此处。

这个模型的训练仅使用了二次元图像，没有任何真人相关的数据在训练中被使用。使用此模型生成或发表图像时，请尊重他人，并遵守相关的法律法规。

(中文在下方）

English brief:

The model is based on the avatars of the Vup (Vtuber) on the Bilibili/Youtube, Hiseki Erio. Six L2D clothes are trained and embedded in the current version. Please feel free to follow/subscribe her account if you like this model.

The main model training is based on the AOM2, and it should work fine with SD1.5, Dreambox, AOM2, and AOM3 of any type.

The suggested baking weight for the Lora is 0.5~0.8.

The keywords of training come with one main keyword “Erio” and six other keywords depending on the selection of the L2D style you want to produce: “casual_clothes”, “japanese_clothes”, “show_clothes”, “loli_clothes”, “leather_clothes” and “china_dress” (for V0.2). The keywords may change in different versions.

Some of the example prompts are provided in the figures. I did not modify the prompts in the figures so that might be some difference in the model versions or keywords in some figures. But it should be fine as a starting hint :)

Please enjoy.

中文简介：

这个模型是基于在Bilibili和Youtube活跃的Vup/Vtuber绯赤艾莉欧的二次元形象生成的。6件L2D的服装被整合在当前的版本中。如果您喜欢这个模型，欢迎关注/订阅她的账号。

模型的训练的底模基于AOM2，实际测试中，应当支持SD1.5, Dreambox, AOM2和AOM3相关底模的图片生成。

建议的模型lora权重为0.5~0.8。

主要的关键词为：”Erio”。

可以使用另外6个关键词来强化想要生成的服装风格：“casual_clothes”, “japanese_clothes”, “show_clothes”, “loli_clothes”, “leather_clothes” and “china_dress” (对于 V0.2版本)。未来关键词可能会发生变动。

可以通过范例图片查看每件衣服的生成咒语。实例内可能有些图片的关键词和版本号可能与V0.2不匹配，不过作为起步点应当是足够了:)

希望你喜欢。

Philia from Stella of the End/Tsui no Stella

Lower the weight if you want different outfits!

Puyopuyo Series - Intelligent Heroine - Ando Ringo LoRA

Note: If you like it, please give it a STAR MARKS !

,

This is the heroine of Puyo Puyo 7.

The appearance has been reproduced reasonably well, but the drill hair may be a little unstable.

Please draw the Puyo-puyo hairpin by yourself after refining...

,

If it doesn't work, the only way to deal with it is to just do the numbers.

And don't stick to the sample prompts.

Thank you for your patience, it is much appreciated.

,

,

ぷよぷよシリーズ 頭脳明晰なヒロイン あんどうりんご LoRA

※ 気に入っていただけたならぜひ★をつけてください！

,

ぷよぷよ７のヒロインです。

外見はそれなりに再現できていますが、ドリルヘアーがやや不安定かも。

ぷよヘアピンは錬成後に自力で描いてください・・・

,

うまくいかない場合の対処方法はひたすら数をこなすこと。

そしてサンプルプロンプトにこだわらないことです。

あなたの忍耐に感謝いたします。

,

,

SamplePrompt (サンプルプロンプト)

,

<lora:ringo:0.7> kawaii ringo petite cute 1girl, green eyes, character hair ornament, red hair, ahoge, twin drill, wristband white skirt, cleamyellow sweater vest, red skirt, purple necktie, school uniform, navyblue thighhighs, red shoes, zettai ryouiki kind smile

,

Negative prompt

badhandv4 bad_prompt EasyNegative, 2girls NSFW

,

Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 4, Seed: 136328499, Size: 384x512, Model hash: 55b3fdcfc6, Model: caulkinumV2_AR, Denoising strength: 0.6, Clip skip: 2, ENSD: 31337, Hires upscale: 2, Hires upscaler: R-ESRGAN 4x+ Anime6B

,

,

,

,

Use Checkpoint (使用したチェックポイント)

caulkinumV2_AR

https://civitai.com/models/32537/caulkinumv2ar-for-anime-style

You can use it to highquality illustrations.

クオリティの高いイラストを錬成できるので使ってみてください。

,

,

Donation (寄付)

Coffee will increase work efficiency and motivation.

コーヒーは作業効率を高めモチベーションをアップさせるでしょう。

https://ko-fi.com/489489

,

Thank you for reading to the end.

I hope one of you will use it.

,

最後まで読んでくれてありがとうございます。

使用してくれるか方が一人でもいることを祈っています。

Suggested weighting: 0.9

Trigger words: kaneki ken

Also suggest to use more accurate prompts: white hair, red and black eye, blood eyes, black clothes, scorpio tentacles

✅How to use

To use it, you need to register <lora:mid_buildings_yiu_v1:0.8> in the prompt field. I recommend exactly this value of the LoRA weight, as it gives more variability.

It is recommended to write the following tokens: structure, city, house, blackout, slate gray, epic lighting, glow, complex details, vegetation, unusual colors, raw, hdr, professional photo.

Copy the prompta and parameters from the preview to this model — you will not lose.

✅What the model is suitable for

This model is designed to generate buildings of various styles: from Venetian classics to modern neo-modern and cyberpunk.

I trained LoRA on the Deliberate_v2 model, so it's better to work with her in pairs.

It is possible to generate at high resolutions and get a complete image, since the dataset had sources with a panorama and a wide shooting angle.

✅Which model works best with

It is best to use the Deliberate_v2 model.

You will find more examples and help in using them in my telegram chat.
I'm on Dprofile.
My Telegram channel.

Ai Miyashita from Love Live! Nijigasaki High School Idol Club

Best to use with weight 0.6 - 0.8 with proper tags.

miyashita ai, blonde hair, yellow eyes, bangs, clothes around waist, nijigasaki academy uniform, ponytail

Have fun!

Rina x Ai for life! <3

Puyopuyo Series: Love solves everything? Heroine Ally LoRA

Note: If you like it, please give it a STAR MARKS !

,

She is a new heroine introduced in Puyo Puyo Chronicle.

Many of you may have been captivated by her love.

The details of the modeling are subtle, but I think the figure has been generally well reproduced, including the coloring.

,

If it doesn't work, the only way to deal with it is to just do the numbers.

And don't stick to the sample prompts.

Thank you for your patience, it is much appreciated.

,

,

ぷよぷよシリーズ 愛で全て解決？なヒロイン アリィ LoRA

※ 気に入っていただけたならぜひ★をつけてください！

,

ぷよぷよクロニクルで登場した新ヒロインです。

彼女の愛に包まれて虜になってしまった方も多いでしょう。

細かい造形は微妙ですが、色合い含めておおむね再現できていると思います。

,

うまくいかない場合の対処方法はひたすら数をこなすこと。

そしてサンプルプロンプトにこだわらないことです。

あなたの忍耐に感謝いたします。

,

,

SamplePrompt (サンプルプロンプト)

,

<lora:ally:0.8> kawaii ally cute 1girl, green eyes, purple pendant, light brown hair, medium hair, yellow pantyhose, green dress, lightgreen shirt, green boots naughty smile style parody

,

Negative prompt

badhandv4 bad_prompt EasyNegative, 2girls NSFW

,

Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 4, Seed: 3117124254, Size: 384x512, Model hash: 55b3fdcfc6, Model: caulkinumV2_AR, Denoising strength: 0.6, Clip skip: 2, ENSD: 31337, Hires upscale: 2, Hires upscaler: R-ESRGAN 4x+ Anime6B

,

,

,

,

Use Checkpoint (使用したチェックポイント)

caulkinumV2_AR

https://civitai.com/models/32537/caulkinumv2ar-for-anime-style

You can use it to highquality illustrations.

クオリティの高いイラストを錬成できるので使ってみてください。

,

,

Donation (寄付)

Coffee will increase work efficiency and motivation.

コーヒーは作業効率を高めモチベーションをアップさせるでしょう。

https://ko-fi.com/489489

,

Thank you for reading to the end.

I hope one of you will use it.

,

最後まで読んでくれてありがとうございます。

使用してくれるか方が一人でもいることを祈っています。

LoRA of Lila from Atelier Ryza, works around 0.8-1.0

莉拉的模型，推荐权重0.8-1.0

-

support 3 outfits, default, swimsuit from Ryza1 and costume of Azur Lane Collab

支持三套外观，默认，一代泳装和碧蓝航线联动皮肤

-

Default默认=OfficialOutfit, Default Clothing, bodysuit

Swimsuit-Cool Selenite泳装-清凉水晶=OfficialClothing2, Lila Swimsuit, one-piece swimsuit

Azur Lane-Moonlit Encounter碧蓝联动-月下的邂逅=MoonlitEncounter, See-Through Nightgown, see-through dress

-

Feel free to post any idea or character model request in the Discussion

有模型的需求可以B站私信或者在这里留言给我

id：不吃青蛙的阿龙

有空就会做

If you like my work, feel free to support me on patreon.com/RiuKiMK1

or RiuKi_MK1 is Creating AI artworks, LoRa models (buymeacoffee.com)

Works good at the weight of 1

This is something I've been meaning to do since long time. What if you could make a model that has dark souls vibe but anime? Well this model aims to be that. I'm sure I could improve but I figured it can be shared in the current state anyways as it's already quite good and fun to play with.

Puyo Puyo Series - Vivacious and Cute Heroine Amiti LoRA

Note: If you like it, please give it a STAR MARKS !

,

This is the heroine of Puyo Puyo Fever.

Her appearance is fairly well reproduced, but retouching is required as she tends to have mysterious Puyo Puyo creatures around her.

The parody style is more like the original, but it looks more modern and cute without it!

,

If it doesn't work, the only way to deal with it is to just do the numbers.

And don't stick to the sample prompts.

Thank you for your patience, it is much appreciated.

,

,

ぷよぷよシリーズ 快活でキュートなヒロイン アミティ LoRA

※ 気に入っていただけたならぜひ★をつけてください！

,

ぷよぷよフィーバーのヒロインです。

外見はまあまあ再現できていますが、周囲になぞのぷよ状生物が発生しがちなので要レタッチ。

パロディスタイルを入れた方が原作寄りですが、入れない方が現代風でかわいいですね！

,

うまくいかない場合の対処方法はひたすら数をこなすこと。

そしてサンプルプロンプトにこだわらないことです。

あなたの忍耐に感謝いたします。

,

,

SamplePrompt (サンプルプロンプト)

,

<lora:amitie:0.8> kawaii amitie petite cute 1girl, green eyes, blonde hair, short hair, bracelet, jewelry hat, skyblue shirt sleeveless, red belt, red shoes, yellow shorts, pink heart whiskers

Negative prompt: badhandv4 bad_prompt EasyNegative NSFW

,

Negative prompt

badhandv4 bad_prompt EasyNegative, 2girls NSFW

,

Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 4, Seed: 1162829878, Size: 384x512, Model hash: 55b3fdcfc6, Model: caulkinumV2_AR, Denoising strength: 0.6, Clip skip: 2, ENSD: 31337, Hires upscale: 2, Hires upscaler: R-ESRGAN 4x+ Anime6B

,

,

,

,

Use Checkpoint (使用したチェックポイント)

caulkinumV2_AR

https://civitai.com/models/32537/caulkinumv2ar-for-anime-style

You can use it to highquality illustrations.

クオリティの高いイラストを錬成できるので使ってみてください。

,

,

Donation (寄付)

Coffee will increase work efficiency and motivation.

コーヒーは作業効率を高めモチベーションをアップさせるでしょう。

https://ko-fi.com/489489

,

Thank you for reading to the end.

I hope one of you will use it.

,

最後まで読んでくれてありがとうございます。

使用してくれるか方が一人でもいることを祈っています。

就纯纯画个fumo用的，画⑨效果比较好，别的人物包括oc也可以，

trigger word: in wall

It can generate frontal view and side view, but not stable, suggest use controlnet

🏴‍☠️Ahoy! This Lora is divided into 2 halves in order to maximize the quality of each outfit. Check the second image for prompt comparisons.

• Unison edition: The 4 classic outfits, Pirate + Gothic + Kimono + Beach

• Special edition: New or unique outfits, Officer + Beret + Nun + Idol

Both Loras work at weight 1.

Made with my Lora guide.

原始服装：

Flametail,origen,(aged down),armor,medium tail,(ear ornament:1.4),ear covers,thighhighs,armored boots,low ponytail,empty handed,standing,full body

小健将：

Flametail,jianjiang,(aged down),orange shirt,medium tail,white shorts,suspender shorts,grey striped socks,black footwear,head-mounted display,thigh strap,crop top,full body,standing

权重(weigh):0.8~1 cfg:6~8

模型的尾巴和耳饰可能有点问题，不知道是不是tag冲突了，是的话之后改改，不知道方舟什么时候才能多出一些可爱一点的角色了，不然我只能接着去找以前那些没人气角色接着弄了

There may be some problems with the tail and earring of the model, I don't know if there is a conflict between the tags, if so, I will change it later. I don't know when the arknights will have some more cute characters, or I can only go to those unpopular characters before

Puyopuyo series, atrocious and immovable heroine Arle Nadja LoRA

Note: If you like it, please give it a STAR MARKS !

,

It is based on the costume that was redesigned for Puyo Puyo after the Magic Tale.

The pose tends to be fixed, so please use controlNet, etc.

The parody style is more like the original, but it is more modern and cute without it!

,

If it doesn't work, the only way to deal with it is to just do the numbers.

And don't stick to the sample prompts.

Thank you for your patience, it is much appreciated.

,

,

ぷよぷよシリーズ 非道で不動のヒロイン アルル・ナジャ LoRA

※ 気に入っていただけたならぜひ★をつけてください！

,

魔導物語の後、ぷよぷよでデザイン変更となった衣装がベースです。

ポーズが固定されがちなのでコントロールネット等を活用してください。

パロディスタイルを入れた方が原作寄りですが、入れない方が現代風でかわいいですね！

,

うまくいかない場合の対処方法はひたすら数をこなすこと。

そしてサンプルプロンプトにこだわらないことです。

あなたの忍耐に感謝いたします。

,

,

SamplePrompt (サンプルプロンプト)

,

<lora:arle:0.7> arle nadja kawaii petit cute girl yellow eyes brown hair,half updo short topknot blue skirt, blue short sleeves, blue breast armor, white shirt, wrist cuffs blue capelet blue boots :> style parody

,

Negative prompt

badhandv4 bad_prompt EasyNegative, 2girls NSFW

,

Steps: 20, Sampler: DPM++ SDE Karras, CFG scale: 4, Seed: 382064522, Size: 384x512, Model hash: 55b3fdcfc6, Model: caulkinumV2_AR, Denoising strength: 0.6, Clip skip: 2, ENSD: 31337, Hires upscale: 2, Hires upscaler: R-ESRGAN 4x+ Anime6B

,

,

,

,

Use Checkpoint (使用したチェックポイント)

caulkinumV2_AR

https://civitai.com/models/32537/caulkinumv2ar-for-anime-style

You can use it to highquality illustrations.

クオリティの高いイラストを錬成できるので使ってみてください。

,

,

Donation (寄付)

Coffee will increase work efficiency and motivation.

コーヒーは作業効率を高めモチベーションをアップさせるでしょう。

https://ko-fi.com/489489

,

Thank you for reading to the end.

I hope one of you will use it.

,

最後まで読んでくれてありがとうございます。

使用してくれるか方が一人でもいることを祈っています。

Created by GeomanticArts

There are already some Bowsers here on the site, but something always just felt lacking so here's another LoRA taking a crack at it. Trying some new training settings that should hopefully work at 64 DIM, feedback is definitely appreciated.

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

'Bowser' - Example prompt

Recommend adding 'full body' to get his greatness in full view.

LoRA metadata has all the tags used.

My (nsfw) Twitter

Trained on NAI.

刘浩存

The eyes are the window to the soul.

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

权重约0.7，可以自由更改权重来获得不同效果，依旧是用ai图生成的素材

WEIGHT 0.7 You can freely change the weight to achieve different effects.It is still made using materials generated from AI images

大概是一种风格，服装，人物三要素都沾一点的lora

It's probably a Lora with three elements of style, clothing, and characters

可以通过自由搭配tag来获得不同风格的服饰，外表

You can freely mix and match tags to obtain different styles of clothing and appearance

Trigger word: koyori_togashi, japanese_clothes, magatama, hakama, hakama_skirt, goggles_on_head, ofuda, miko, tabi, red_hakama

weight = 0.7

Named "Pure Innocent Girl" because it can produce a youthful and beautiful innocent face, and is a more natural and realistic oval face (with less plastic surgery sensation); The recommended weight is between 0.2 and 0.5. The higher the weight, the more likely it is to produce childish features that look too childish. The maximum weight should not exceed 0.7, otherwise it will cause deviation and affect the style. Suitable for semi realistic and realistic effects, but not tested in anime styles.

If you like this LoRA, please leave a comment or image!

取名叫“純真女孩”，是因為它能產生年輕美麗的清純臉，並且是更自然真實（整形手術感覺更少）的鵝蛋臉； 權重建議在0.2～0.5之間，越高越容易產生看起來太孩子氣的幼稚特徵，最高不要超過0.7，否則會產生偏差影響風格。適合半寫實和逼真效果，但並沒有在動漫風格中測試。

如果你喜歡這個LoRA，請留下評論或圖片！

On April 22nd,

The V2 version is the final version of this LoRA, with the following adjustments made:

The model was trained with "stable_diffusion 15 pruned" and "vae ft mse-840000" to improve the universality of LoRA;

Trim tags, remove some tags and add new tags such as clothing and posture, while simplifying the abbreviation keyword to "PI_girl". Of course, as long as LoRA is loaded or the entry is written<lora: Pure_ Innocent_ Girl2:0.4>, can be used without triggering words;

Updated the dataset, removed some potentially harmful images, added some high-quality high-definition new images, trained with about 300 images, and also improved the training resolution;

Increase the dim value to 64, increase the number of training steps, and adjust the learning rate.

Known issue: Using excessive weights (exceeding 0.7) can still affect the character's clothing and background style.

Example of dataset labels:

1girl，best quality，masterpiece，realistic，solo，looking at viewer，face，eyes，nose，mouth，lips，smile，makeup，

Pose: suggestive_ spread_ legs_ pose，lying，on back，Standing，sitting，kneeling，Squatting，

Clothing: Ice princess dress,

4月22日，

V2版本是這個LoRA的最終版本，做了以下調整：

模型改用“stable_diffusion_1_5-pruned”與“vae-ft-mse-840000”訓練，以提高LoRA的泛用性；

修剪標籤，删除了一些標籤和新增了一些服飾、姿勢等新標籤，同時簡化縮寫關鍵字為“PI_girl”，當然，只要加載了LoRA或者詞條寫上<lora:Pure_ Innocent_ girl2:0.4>，無需觸發詞也可使用；

更新數据集，删除了一些可能導致不良的影像，添加了一些高品質的高清新影像，用了約300張圖片訓練，同時也提高了訓練分辯率；

增大dim值調整為64，新增訓練步數，以及調整了學習率。

已知問題：使用過高權重（超過0.7）依然會影響人物服裝和背景風格。

數据集標籤示例：

1girl，best quality，masterpiece，realistic，solo，looking at viewer，face，eyes，nose，mouth，lips，smile，makeup，

姿勢：suggestive_ spread_ legs_ pose，lying，on back，Standing，sitting，kneeling，Squatting，

服飾：Ice princess dress，

On April 10th,

The V1 version is an experimental LORA model that used approximately 200 images to train in "ChiloutMix" and generate "vae-ft-mse-840000" VAE. This is a very featureless LoRA, which did not meet my original expectations. It still looks like a mixed race with many Korean faces.

4月10日，

V1版本是一個實驗性的LORA模型，用了約200張圖片在“ChilloutMix”中訓練和“vae-ft-mse-840000”VAE生成。 這是一個非常沒有特點的LoRA，並不符合我原本的預期，看起來還是混合了很多韓國面孔的混血。

This model was trained on SD1.5.

I suggest using CFG 3-8, strength 1.2, and anything above 15 sampling steps.

Model used: realisticVisionV20_v20

Didn't see best girl from FF14 so I decided to do something about it. Getting the mole right is some hardcore gacha and I recommend weighting mole_under_mouth to less than full unless you want too many moles. Not sure why that happens but eh.

0.8 best

learn from 4044 majic his telegram group is: https://t.me/majicplayground

my channel is https://t.me/dollsgirlswifes

Based in US, Uchi, Data Scientist

dm here https://t.me/happygoat1946

Workstation: NVIDIA® A40 (from lab)

#ai #stablediffusion #chilloutmix #lora

A LoRa to show that special moment when you are at the right place at the right time

It's seem a bit difficult for me to create this pose but now it's work.
try some keyword like
(2boys grabbing 1girl breast:1.1),2boys,grabbing breast, girl, (bald male, muscle male, black skin male:1.1).
It's better mix with other LoRA to create your character pose like this.
Feel free to upload your work and give me idea!
If it's help dont forget to buy me a coffee!

Note:

To achieve better results, it is recommended to use lora block weights.

A simple method to set the weights is as follows: CHARACTER:1,1,1,1,1,0,0,0,1,1,1,1,1,1,1,0,0

triggle words :

Hoshino Ai, long hair, purple hair, streaked hair ,purple eyes, star-shaped pupils, hair ornament,

model: most of anime-like models

V8.3 includes improvements from Version 5, offering slightly enhanced brightness; however, its performance remains subpar when it comes to hair ornament details.

V9 employs the V1 dataset and has been re-tagged to eliminate the yellow cloth around the neck, resulting in a more satisfactory output.

V1.1 has issues with red pollution, frequently displays yellow cloth on the neck, and its posture is relatively static.

V5 addresses the issues found in V1.1, but it introduces some new problems, making it not necessarily superior to V1.1. After making significant changes to the dataset, it appears that there are too many images with poor lighting, which may occasionally compromise the image quality.

V5 appears very dark, and it seems challenging to increase its brightness, and if you use easynegative, it will somehow look like 'deep-frying' without highres fix. I'm trying to solve the problem.

So choose the model you like.

It seems civitai can't read the pnginfo from my picture due to Dynamic Prompts.

要流眼泪需要添加tears

要官图手扶柰子的姿势需要添加hand on breasts

Short bio:

Sasha Banks, now under the name Mercedes Moné to New Japan Pro-Wrestling, is a female professional wrestler who was best known for being a popular figure in WWE in 2012 to 2022.

Banks began her wrestling career in 2010, training under the likes of Brian Kendrick and Sara Del Rey. She signed with WWE in 2012 and made her debut in NXT in 2013.

She quickly became a fan favorite in NXT, winning the NXT Women's Championship and having memorable matches with the likes of Bayley and Charlotte Flair. Banks was eventually called up to the main roster in 2015 and has since become one of the most recognizable and popular wrestlers in the world.

Throughout her WWE career, Banks has won multiple championships, including the Raw Women's Championship, SmackDown Women's Championship, and WWE Women's Tag Team Championship. She is known for her high-flying, innovative wrestling style and her magnetic personality.

Outside of wrestling, Banks has appeared in a number of TV shows and films, including The Mandalorian and The Big Show Show.

Trained this on Stable Diffusion 1.5, and can also work in other styles such as Deliberate and Realistic Vision 2.0!

Recommended CFG scale: 5 - 7

Recommended denoising strength: 0.5 - 0.7

Recommended sampler(s): Euler A or DPM++ 2M Karras

Make sure to add the trigger word "Sasha Banks" (minus the quotations)

Negative prompt template that works well:

paintings, sketches, (worst quality:2), (low quality:2), (normal quality:2), dot, mole, lowers, normal quality, monochrome, grayscale, lowers, text, error, cropped, worst quality, low quality, jpeg artifacts, ugly, duplicate, morbid, mutilated, out of frame, extra fingers, mutated hands, poorly drawn hands, poorly drawn face, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, cloned face, disfigured, gross proportions, malformed limbs, missing arms, missing legs, extra arms, extra legs, fused fingers, too many fingers, long neck, username, watermark, signature, less than five fingers

This is my first ever LoRA, and I would love to hear feedback on how my LoRA is like and how to improve on future LoRAs!

It's the style of my favourite mangaka, she draws Yuru Yuri.

-WARNING-

I'm not satisfied with this, its my first style lora. now that I know what I'm doing (kind of) I realize that this has plenty of room for improvements so there will be an update in the future. I'm uploading this anyways because SOMETIMES it works and there could be people interested in it since it was a popular series back in the day.
also, if the hands look good without ControlNet consider playing the lottery.

no trigger words but these help with the aesthetic: retro artstyle, 1990s (style), anime screencap , screencap

Lora strength: 0.7-1 (with 1 sometimes you will have add weight to some of your prompts)

use Hires. fix for better results

you can summon(most of the time) :

Aika: blonde, ponytail, long hair, blue eyes

Rion: short hair, red hair, blown eyes, glasses

models tested: yesmix_v15, corneos7thHeavenMix_v2, abyssorangemix2_Hard, anything-v4.5, walnutcreamBlend_herbmixV1

Lora trained on 974 images of Shana per request.

I release all my models for free, if you feel like supporting me and my work you can support me here and buy me a drink.

Prompts:

shanaschlblack, shanaschlred, shanared, shanablack

Trained on NAI, gen'd on AOM2 hard

阪神甲子園球場…を学習させたはずなんだけど…海外の球場っぽくなっちゃうんですよ…なんでやねん…野球場を学習させたはずなのに…グラウンドめちゃくちゃなんですよ…なんでやねん…

・外野席からグラウンドを見る

koshienstadium, scenery, outdoors, grass, night, too many, crowd, 6+boys, multiple boys, multiple girls, 6+girls, sportswear, real world location, baseball stadium

・外野席スタンド

koshienstadium, crowd, black hair, outdoors, shirt, male focus, 6+boys, short hair, short sleeves, multiple boys, cloud, yellow shirt, scenery, sky, from behind, people, 6+others, solo focus, sportswear, 1boy, too many, multiple girls, 6+girls, real world location, baseball stadium

・内野席アイビーシート

koshienstadiumIB, 6+boys, crowd, 6+girls, multiple boys, multiple girls, real world location, photo background, black hair, sportswear, audience, baseball stadium

・内野席グリーンシート

koshienstadiumGS, 6+boys, multiple boys, real world location, sky, outdoors, cloud, flag, scenery, baseball stadium

・ジェット風船

koshienstadiumJB, audience, crowd, balloon, solo focus, english text, wristband, 6+boys, multiple girls, multiple boys, 6+girls, sportswear, real world location, baseball stadium

・トラッキー

tolucky, 1boy, mascot, male focus, stadium, solo, standing, baseball mitt, full body,

PLASTIC MEMORIES Isla

use anylora

0.7 best

learn from 4044 majic his telegram group is: https://t.me/majicplayground

my channel is https://t.me/dollsgirlswifes

Based in US, Uchi, Data Scientist

dm here https://t.me/happygoat1946

Workstation: NVIDIA® A40 (from lab)

#ai #stablediffusion #chilloutmix #lora

Model of the beautiful Juno Temple.

It is friendly with all models and styles, use with 1.7 strength.

use anylora

If you'd like to support me and the LoRAs that I create consider throwing a tip my way @ https://ko-fi.com/kouyasimp

LoRA trained to replicate the meme/art trend "Cover Them Up Slut". Trained on 43 images. Compatible with character LoRAs.

Recommended weight is 1. Recommended increasing weight if using with a character lora as otherwise the word bubble likes to disappear.

Trigger word is meme. I may update it to have the word bubble and side heart/profile picture to be triggerable.

Recommended Weight: 0.6~1.0

Trigger prompts: red blindfold, twin braids, armlet, fingerless gloves, headdress

For full body, add: toeless heels

A interesting prompt you could try (and may not work properly): floating ring

https://www.buymeacoffee.com/kittykangz

ChilloutNight

vae-ft-mse-840000-ema-pruned

What? Do you think the name of this model refers to some ship name? No, no...

Look, some people have Asuna as their Waifu, others on the other hand prefer something more like GGO Kirito, especially after what Sword Art Online Code Register did. So... Why not both in a single body? Technically that would be SAO ultimate waifu, and deep down you know it's true regardless of your preferences.

I'm not the first to come up with this idea and you can trace similar ideas on the Internet years ago, but it's always short lived. I don't know what people are so afraid of. Do you think Kirito's vengeful spirit will hunt you down? I take the risk for you then, if he comes I'm sure I can one shot him.

Anyway, let's get down to business. This model was trained with several official images of Kirito and Asuna and therefore it understands some physical characteristics from regular Kirito, GGO Kirito and Asuna.

• Kirito's main prompts are: black eyes, black hair, short hair, hair between eyes

• GGO Kirito's are: purple eyes, black hair, long hair, hair between eyes

• Asuna's: brown eyes, brown hair, long hair, braid, french braid, crown braid

The idea of the model is to mix these characteristics as you prefer. I usually consider past ideas that Kirisuna would have long black hair with crown/french braid and brown eyes, but you can do it your way. The model will generate the character as female, which is its intention. But if you prefer the result to be something more "exotic" there are ways out there if you look for it.

It's possible to generate Kirito's and Asuna's default clothes, as seen in the examples. But Asuna's is easier than Kirito's due to the more detailed descriptions. Of course, it's also possible to make just Kirito or just Asuna in this model. But why would you want to do that? There are other models of the characters, separated, and with much better quality than this one.

I'm still struggling to make my models to work with Weight 1, but that won't be the case here again, so you will need to scale it down a bit to make it decent.

Have fun!

斯库拉的lora，训练数据来源于p站。官方数据不多，同人图的质量也一般，所以模型可能还有点问题，比如十字瞳有点难弄。

分为原皮和制服皮肤两个概念，大致上来看，质量是原皮会更好一点。不过制服的丝袜质感不错（需要调高权重）。

An LoRA of scylla|shira(azur lane).

Trigger words can be set as origin or uniform to change her clothes.

西宮北高校近辺の風景を学習させたLoRA

北高と苦楽園小の間の坂道（通称ハルヒ坂）

haruhizaka, kitakoumae, scenery, outdoors, road, power lines, city, tree, building, sign, cityscape, road sign, utility pole, ground vehicle, motor vehicle, street, car, sky, day, real world location, realistic,

北高・苦楽園小・中の間の交差点

kousaten1, outdoors, scenery, tree, road, sky, street, fence, cloud, building, lamppost, utility pole, power lines, road sign, day, crosswalk, sign, cloudy sky, bush, blue sky, chain-link fence, real world location,

北高西門

KitakouNishimon, Kitakou, tree, outdoors, scenery, sky, building, road, day, street, cloud, window, blue sky, plant, car,

北高正門

kitakouseimon, tree, outdoors, scenery, cloud, sky, road, building, utility pole, power lines, cloudy sky, day, sign, house, no humans, fence, 1boy, grass, bush,

Male/Female SFW

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.8 - 1.0

Trigger: marianne von edmund

Default Outfit: garreg mach monastery uniform, long sleeves, pantyhose

Post Timeskip: official alternate hairstyle, official alternate costume, blue dress, capelet

Marianne von Edmund from Fire Emblem: Three Houses. Trained on NAI.

Tested at 0.6-0.9 strength.

Contains her academy, Three Hopes, and timeskip outfits.

• Trigger word is "marianne_academy" for her academy appearance. Add "uniform, black skirt, garreg mach monastery uniform, blue pantyhose, boots" for her typical look.

• Trigger word is "marianne_hopes" for her Three Hopes appearance. Add "long hair, hairband, blue dress, capelet" for her typical look.

• Trigger word is "marianne_timeskip" for her timeskip appearance. Add "blue dress, capelet" for her typical look.

甲陽園駅近辺の風景を学習させたLoRAです。

現実世界の甲陽園の様々な特徴をごった煮にして雑に再構成したような異世界の風景が描けます。どうしてこうなった。再現度を求めるなら現実世界の写真にimg2imgとかPhotoshopで合成したり3Dモデリングとかでアセット作った方がはるかに良いじゃない？……これはごちゃごちゃ言うより出来るか出来へんかまずは一度やってみて出来たものを見て考えようやないかという興味本位な動機で始まった雑な実験の成果報告である。

1:駅構内プラットホーム

platfrom, koyoenst, scenery, outdoors, tree, power lines, building, road, day, ground vehicle, train station, street, real world location, utility pole, sky,

2:駅改札口

kaisatsu, koyoenst, scenery, train station, multiple girls, multiple boys, bag, real world location, 6+girls, 6+boys, brown hair, scenery, train, solo focus, black hair, short hair,

3:甲陽園駅外観

koyoenst, exterior, train station, scenery, sign, outdoors, railroad tracks, ground vehicle, real world location, train, scenery

4:駅前風景

koyoen, scenery, tree, outdoors, road, power lines, lamppost, street, sign, sky, road sign, building, real world location, cloud, utility pole, day, fence, crosswalk, ground vehicle, realistic, photo background, photo (medium), photorealistic, photo background

AI先生の特徴を捉える能力はすごい。色々なものを学習して描ける。

既存モデルの海外っぽい風景ではなくて、電柱が立ち並ぶなんちゃって日本の市街地みたいな雰囲気を学んで再構成して絵を描けている。これはこれで素晴らしい。

これまでいくつも風景を学習させてきたけど、作る度に現実世界の場所を「正確に再現する」のはとても困難だと思い知る。

based on Acertainthing.ckpt，you can try with other model

I am the Reindeer Czar, so I created my first LORA, of Dolph from Fortnite. This depicts him really well I think.

Latest Description:

Version 2 improves the model very much. No more issue with background, overall more expressive and high quality. 0.5 - 0.7 weight for LORA is good. Euler a and DPM++ seem to be best sampling. Sample around 50 seems good too, more can be better too.

You can add tags like skydiving, sitting, open mouth, smiling, holding weapon, lightsaber to use for some dynamic situations. You can also add Widow's Promise, Sword to get that back bling on him.

If you like my models, please consider a tip, to help me train more models!

https://ko-fi.com/reindeerczar

Deprecated Version 1 Description:

Strength for LORA is say, 0.4 to 0.65 to get the good result, if you want a b1ackground.

It was trained on 5600 iterations, so quite much. Add the tags blue gradient background and simple background in the negative tag to get rid of the background better.

中文版：

非常感谢Lapin的支持，C站的政策貌似也改了，于是思考了很久最后还是决定回来。因为我是真的没有那么多时间和精力去管理一个频道，目前也没有那么多流量去额外上传Lora和模型。

回来总不能啥都没有，所以就炼了一个素晴的芸芸

提示词：yunyun，yunyun's Style

想要换衣服的话依然是只保留yunyun，然后删除其他的词

但是由于Lora训练方式的问题，很可能一些涉及到训练集里的tag的时候就会无法彻底替换，这是没法避免的，除非欠拟合，但是这样的话很多模型跑出来的效果基本上约等于没有

English vision：

Thank you very much Lapin for your support and it seems that the policy on Civitai has changed, so after thinking about it for a long time I finally decided to come back. Because I really don't have that much time and energy to manage a channel, and I don't have that much traffic to upload extra Lora and models at the moment.

I can't come back with nothing, so I train a Yunyun's Lora

tigger words: yunyun, yunyun's Style

If you want to change clothes, you can still just keep yunyun and delete the rest of the words

However, due to the problem of Lora's training method, it is likely that some of the tags in the training set will not be replaced completely, which cannot be avoided unless they are underfitted, but then many of the models will run out with approximately the same effect as nothing

神戸市東灘区にある石田鶏卵さんの風景

設置されている「うどん自販機」と「カップヌードル」自販機

そして「うどん自販機」のおうどん

うどん自販機

UVM1, vending machine, scenery, outdoors, sign, tile floor, tiles,

うどん自販機のおうどん

IKUdon, food, food focus,

※石田鶏卵さん推奨の「たまご載せ」ができます。

IKUdon, food, noodles, food focus, still life, bowl, realistic, egg, meat, chopsticks,

カップヌードル自販機

CNVM, vending machine, shop, scenery, close-up, food, noodles, cup ramen, food focus, traditional media, still life, scenery, outdoors,

石田鶏卵さん外観

ishidakeiran, outdoors, power lines, scenery, utility pole, sky, road, ground vehicle, vending machine, building, day, street, motor vehicle, blue sky, shop, sign, car, tree, cloud, exterior

XSarchitectural-41AnimeGirl For 2D3D

Suggestions for use:

• If you want to get the parameters, download the image and use the stable diffusion feature.

If you think it's good, please remember to give me a five-star review

If you have any questions or suggestions, please contact us at xsarchitectrualstudio@gmail.com

First LORA I've ever created. The subject is the internet model Imogen Dyer.

A simple reverse cowgirl LoRA. Trained at 768x768 on high resolution images.

Once in a while you'll encounter floating cocks, or missing limbs. I think you'll get the best generations at 512x768, but there is a lot of potential using the "wide shot" tag with 768x768 images!

Caucasian/black cock and shallow/deep penetration don't totally work as expected. I'll be able to fix that in the next version.

As usual, modify the LoRA weight itself if needed.

Note: make sure to add "open mouth" to the negative if needed. Tons of the training images displayed the woman in ecstasy/surprise, so their mouth was always open.

Prompt template

revcowgirl [wide shot] [woman description] [breast size] breasts [hair style/color] hair [shaved|trimmed|hairy] pussy [black|caucasian] cock [shallow|deep] penetration [setting]

Prompt examples

• revcowgirl 8k uhd photo naked caucasian woman small breasts long blonde hair shaved pussy caucasian cock deep penetration on a couch in a living room

• revcowgirl 8k uhd photo chubby naked polynesian woman large breasts short pink hair hairy pussy black cock deep penetration on a bed in a bedroom

• revcowgirl 8k uhd photo petite woman small breasts short blue tattoos hair shaved pussy (caucasian cock deep penetration) on a bed in a bedroom

• revcowgirl wide shot 8k uhd photo (muscular naked woman) large breasts purple pixie hair (black cock shallow penetration) outdoors pool

• revcowgirl wide shot 8k uhd photo (muscular naked Argentinian woman) eyes closed smiling caucasian cock shallow penetration on a bed in a bedroom

• revcowgirl wide shot petite naked caucasian woman medium breasts blonde hair shaved pussy (black cock shallow penetration) on a bed in a bedroom

• revcowgirl wide shot (muscular naked caucasian woman:1.2) (red hair) (large breasts:1.2) (fair skin) (freckles) (hairy pussy) Caucasian cock deep penetration on the edge of a bed

秋葉原にあった居酒屋「村役場」の思い出 LoRA has memories of the izakaya "Murayakuba" in Akihabara, Tokyo.

murayakuba, izakaya, scenery, indoors, lamp, table, realistic, photo (medium), photo background, poster (object)

Illustration, chibi, toon Style

This is a model that does not need much to give good results.

This v2 should finally work better with full body shots, but keep in mind this model is mainly thought for medium shots, portraits and closeups. JoJo faces have way too many expression lines, so SD will mess up when trying to simplify them. With high resolutions, it should work fine, though.

Trigger: 'Jolyne_Anime'

Optional tags: 'coat', 'pants', 'face focus', '(((pink hair)))' -> (for an alternate pallet effect)

Weight: [0.6 - 0.9]. Go high for a stronger art style (with less flexibility), or low for more creative uses, poses or realism. With closeups/portraits, faces keep the style solid under 0.65, so don't hesitate to go down if needed.

Suggestion: Try to experiment with lights, colors and visual effects. Some results can be quite interesting!

______

Example prompt:

Jolyne_anime, confident, ready to fight, creative shot, green eyes, green lips, perfect quality, (absurdres), 4k, masterpiece, lens flare, light rays, (psychedelic background), illustration, ((anime screenshot)), professional lighting, posing for viewer, ((psychedelia, face focus, wind, detailed face)) + [LoRA set to 0.65]

LoRA trained with images of character Miyamoto Rei (High School of the Dead)

v1.5 (NEW)

Increased dataset, modified tags, cleaned up tags, used new and improved training parameters. This results in a both more consistent and flexible LoRA. It works with more models and can be combined with other LoRAs. The example images combine with another character LoRA.

-- previous versions --

v1.5

New training parameters, tags and dataset. Trained on NAI. Tested on AOM2, Perfect World, NeverEnding Dream, and other models. Works with 1.0 weight and 0.3-0.7 CFG, but depends on the model.

You don't need a trigger word to get the character. For the basic uniform you can simply use the trigger "serafuku". You can also add to the positive or negative prompt to get variations from the tagged training data: shirt, skirt, thighhighs, elbow pads, knee pads, sneakers, rifle, stick, blood, torn clothes, shoulder holster.

See the information in the image examples, including for tested models.

Works with multiple models and LoRAs.

v1.0

Better results with 0.5-0.7 weight and 0.3-0.6 CFG (v1.0), but it may depend on the model used.

Works with realistic and anime models and to work with other characters and styles (see examples on how to achieve this). Clothes can be changed by adding, removing or reordering these tags:

"green sailor collar, black bow, green skirt, black thighhighs, long sleeves, brown hair, brown eyes, hair between eyes, ponytail, sneakers, white footwear"

(tip: don't describe the shoes if you don't want a full body image)

See the version information for more on the LoRA training. Please post your images and let me know your feedback.

Want to support a banana? Feel free to donate ありがとう 🍌https://www.buymeacoffee.com/colakoos

READ VERSION DESCRIPTIONS

All character(s) shown belong to their rightful owners.

日本のモダンなおうちの玄関 Modern Genkan in Japanese Houses

集合住宅の玄関（家の中から）Entrance of an apartment building (from inside the house)

genkan, JMG2, door, wooden floor, shoes removed, shoes, indoors, scenery,

genkan, JMG2, door, scenery, indoors, tiles, tile floor, open door, slippers, towel, box, wooden floor,

集合住宅の玄関（家の外から）Entrance to an apartment building (from outside the house)

JMG3, genkan, door, indoors, scenery, hallway,

一軒家の玄関（家の中から）Entrance of a house (from inside)

genkan, JMG1, door, hallway, scenery, indoors, tiles, tile floor, open door, window, light, ceiling light,

一軒家の玄関（家の外から）Entrance of a house (from outside)

genkan, JMG0, door, scenery, window, tiles, tile floor, outdoors, day, open door, exterior

genkan, JMG0, scenery, door, tree, plant, window, outdoors, building, exterior

genkan, JMG0, scenery, outdoors, sky, tree, cloud, day, plant, building, door, blue sky, window, shadow, wall, house, stairs, exterior

realisticを足すと描画がリアル風味になります　Adding "realistic" makes the drawing more realistic.

※どうやら玄関で靴を脱ぐ概念の学習は難しかったようです。

*Learning the concept of taking off shoes at the entrance was difficult.

这是一款基于《Fate/Grand Order》/FGO 中的Abigail Williams(阿比盖尔·威廉姆斯) 设计的LoRA模型。

该模型使用SD1.5作为底模，建议将权重设置在0.8左右。过低的参数可能会导致样本特征不明显。采用的采样方法是DPM++ SDE Karras。

经过测试，我们推荐使用dalcefoPainting_3rd、anything等二次元大模型，并配合moxin等画风LoRA使用，以获得更好的效果。如果需要激活NSFW功能，您可以使用"NSFW"作为关键词，并与其他NSFW LoRA一起使用。

此外，如果该模型被广泛下载，我们将继续加入更多的特征和功能，并持续进行更新和改进。

我们衷心感谢所有使用和支持该模型的人。如果您有任何反馈或建议，请随时与我们联系。

This is a LoRA model based on Abigail Williams, from "Fate/Grand Order".

The model uses SD1.5 as the base and is recommended to be trained at around 0.8 weight. Setting the weight too low may result in unclear features. The sampling method used is DPM++ SDE Karras.

For the choice of large models, after testing, it is recommended to use dalcefoPainting_3rd, anything, and other 2D large models, and to use moxin and other style LoRAs in conjunction for better results. If you want to activate the NSFW function, you can use "NSFW" as the keyword and use it together with other NSFW LoRAs.

In addition, this model will continue to be updated and improved, and more features and functions will be added if there are many downloads.

We would like to express our sincere gratitude to all the people who use and support this model. If you have any feedback or suggestions, please feel free to contact us.

I forgot to prune so you must prompt everything.

suggested prompt is: claire, furry catgirl, white fur, blue long hair, parted bangs, fangs, black eyes, slit pupils, :3, pentagram forehead mark, black collared dress, short sleeves, pink animal slippers, cat ears, cat nose

Someday will fix this and train at higher resolution too.

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.8

Trigger + Appearance: charlie morningstar, long hair

Default Outfit: red suit, red pants

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

This was quite the request.

Weight: 0.5 - 1.0 (default), 0.8 - 1.0 (Timeless Classics)

Trigger + Appearance: formidable \(azur lane\), red eyes, large breasts

Default Outfit: frilled dress, white thighhighs, cleavage

Timeless Classics: formidable \(timeless classics\) \(azur lane\), long hair, twintails

Do not put formidable \(azur lane\) and formidable \(timeless classics\) \(azur lane\) together.

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.7 - 1.0

Trigger: lacus clyne

Sample Outfit: kimono, bridal gauntlets

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.7 - 1.0

Trigger: sena \(xenoblade\)

Default Outfit: jacket, crop top, shorts, thighhighs, elbow gloves

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.8 - 1.0

Trigger: aizawa ema

Default Outfit: hair ornament, dress, gloves, black choker, jewelry, earrings

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.8

Trigger + Appearance: carly nagisa, long black hair, coke-bottle glasses

Default Outfit: brown sleeveless jacket, open jacket, striped shirt, jeans, wristband

Negative: (eyes:1.2), bare shoulders

This one's weird. Her eyes can be seen through her glasses sometimes and her clothing might have shoulder cutouts.

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.8

Trigger: komori met, virtual youtuber

Sample Outfit: black headwear, jacket, long sleeves

You may replace black headwear with a visor cap if you'd like.

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.7

Trigger + Appearance: vegetto, spiked hair, jewelry, earrings, tail

Sample Outfit: dougi, white boots, white gloves, pectorals

Super Saiyan: super saiyan, blonde hair, blue eyes

April Nottingham

April is an American Instagram famous person.

This is a LoRA . If you'd rather have a textual inversion of April, please go HERE.

Like my work? Buy me a beer!

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.5 - 1.0

Trigger + Appearance: super sonico, headphones, red eyes

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.7 - 1.0

Trigger: alena \(dq4\)

Default Outfit: hat, dress, cape, pantyhose, gloves, belt, jewelry, earrings

Resource used: https://civitai.com/models/22530

This was requested through Google Forms

Weight: 0.5 - 1.0 (default), 0.7 - 1.0 (armor)

Trigger + Appearance: mash kyrielight, pink hair, hair over one eye

Default Outfit: glasses, jacket, dress, necktie, pantyhose

Armor: mash kyrielight \(armor\)

Do not use mash kyrielight and mash kyrielight \(armor\) together. Use the normal trigger for her default outfit.

Delightfulhug (Nikki)

THE BEST WAY TO GET ACCURATE RESULTS

Make a regular 512x image and manually upscale it through image to image. If you need to use hires fix I used Uniscale balanced and UltraSharpX4 they looked okay but changed the face a bit.

trained on "a woman" I would recommend not using any photo jargon starting prompt with "a photo of" just start every prompt with a woman and go into details, but do whatever you want blud.

Photos are very prone to have slight watermark since every trained photo contained one

To avoid the problem of improper commercial or illegal use disputes, when you start using the model, you are fully aware of the declaration document. I do not support the use of this model to generate adult content. The above model is for learning purposes only. Users must abide by local laws and regulations when using the model, and must not violate others' honor, privacy and portrait rights. I solemnly declare that the final right of interpretation of this model belongs to expired myself. If you have commercial needs, please directly contact with business through the official Weibo account of the owner of portrait rights. Without authorization, no commercial use of this model is allowed, otherwise all consequences will be taken by yourself. This declaration document does not represent the official, nor does it represent itself, but is only used for disclaimer by software author himself.

为避免不当商业或非法使用纠纷的问题，当您开始使用模型时，您已充分了解声明文件。 我不支持使用此模型生成成人内容。 以上模型仅供学习使用。 用户在使用模型时必须遵守当地法律法规，不得侵犯他人名誉权、隐私权和肖像权。 本人郑重声明，本模型最终解释权归本人所有。 如有商业需求，请通过肖像权人官方微博直接联系商业。 未经授权，不得将本模型用于商业用途，否则一切后果自负。 本声明文件不代表官方，也不代表其自身，仅供软件作者本人作免责声明之用。

不適切な商業的または違法な使用に関する紛争の問題を回避するために、モデルの使用を開始するときは、宣言文書を十分に認識してください。 このモデルを使用してアダルト コンテンツを生成することはサポートしていません。 上記のモデルは、学習のみを目的としています。 ユーザーは、モデルを使用する際に現地の法律および規制を遵守する必要があり、他人の名誉、プライバシー、および肖像権を侵害してはなりません。 私は、このモデルの最終的な解釈権が失効した私自身に属することを厳粛に宣言します。 商業的なニーズがある場合は、肖像権の所有者の公式 Weibo アカウントを通じてビジネスに直接連絡してください。 許可がなければ、このモデルの商用利用は許可されません。 この宣言文書は公式を表すものでも、それ自体を表すものでもありませんが、ソフトウェア作者自身による免責事項としてのみ使用されます。

부적절한 상업적 또는 불법 사용 분쟁의 문제를 피하기 위해 모델 사용을 시작할 때 선언 문서를 완전히 알고 있습니다. 성인용 콘텐츠를 생성하기 위해 이 모델을 사용하는 것을 지원하지 않습니다. 위의 모델은 학습 목적으로만 사용됩니다. 사용자는 모델을 사용할 때 현지 법률 및 규정을 준수해야 하며 타인의 명예, 사생활 및 초상권을 침해해서는 안 됩니다. 본인은 이 모델에 대한 최종 해석권이 만료된 본인에게 속함을 엄숙히 선언합니다. 상업적인 필요가 있는 경우 초상권 소유자의 공식 Weibo 계정을 통해 비즈니스에 직접 문의하십시오. 승인 없이 이 모델을 상업적으로 사용할 수 없습니다. 그렇지 않으면 모든 결과는 본인이 져야 합니다. 이 선언 문서는 공식을 대표하거나 그 자체를 대표하지 않으며 소프트웨어 작성자 자신의 면책 조항으로만 사용됩니다.

Yuusuke Mogi (Final Fantasy XIV) 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 official artists of Final Fantasy XIV. Works perfectly on AnyLoRA at weight 1. Increase accuracy with mogi yuusuke. Please 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)

trigger words:girl ,oni horns, bodystocking(optional),glasses (optional)

Character model

Character model

Lora trained on 165 images of Tilarna

I release all my models for free, if you feel like supporting me and my work you can support me here and buy me a drink.

May not do the default robe cloth right / May not do her ear piece right on armored ver. Due to limited dataset.

Prompts:

tilarnadef, tilarnaarm, tilarnarnd,

Trained on NAI, gen'd on AOM2 hard

use at a level of 0.8 - 1

actually, you just need to type "dora" to get results like the picture. however, you can also type specific commands to get stronger results.

i hope you like it

Bigger, better creampies for non-anime models. Trained on a 150 image dataset. If you weight it highly (eg 0.8 to 1) it tries to make the position missionary but it will get positions correct with lower weights, and it is fantastic with inpainting on nearly anything combined with URPM in any position. A big additional benefit with this model is that the amount and volume of cum is directly related to the weighting eg 0.9 for large amounts 0.4 for more subtle effects, and it is much more realistic appearing than most of the other models.

Meiko anime prison school work with 0.6 weight enjoy it

Trigger words : meiko

follow me on Instagram @creativehotia

Asaka Yashima from the Konbini Shoujo Z eroge

• Trained with Anime (full-final-pruned) model.

• 5 versions; 9-1:0 (9 epochs), 10-1:0 (epochs), 6-1:2 (6 epochs, activation tag with 1:2 weights), 7-1:2 (7 epochs, 1:2 weights), 10-1:2 (10 epochs, 1:2 weights).

• gives good results with 0.5-0.7 weights

Works pretty well at 0.85 weight. I tried a variety of checkpoints and it seemed fine.
Upscale it from 512x512 1.5x or 2x while adding some sunny or good lighting effects and it will look great.
Any feedback would be welcomed!

DO NOT ENTER private property

Ruin exploration may cause you injured

and today, RUINS are MONITORED by sensors

Versions

• v1.1b For Real models? Or maybe good for collab with another LoRAs (Layered training model)

• v1.1a For Anime models? Structures looks good than v1.1b

• v1.0 First release. Most apocalyptic

Features

It was traind with photographs of actual Japanese ruins. These were taken 20 years ago, now those ruins had been vanished.

This model will be improved for your support. See Donations:

How to Use

Recommended LoRA weight is 0.4 to 0.7.

Watch out for your downloded file name. Write same name into prompts.

<lora:factoryRuins_v10:0.5>, (factory ruins:1.1)

Prompts Act

Location

• indoors, outdoors

Scenenary

• train station : impressive broken roofs

• railroad tracks, ground vehicle

• industrial pipe : chemical plants

• indoors, window : inside a room

• machine : equipments

• steel : rusted structures

• building, stone wall

More Apocalyptic

• rubble, grass