If you like my LoRAs please consider to support me by buying me a coffee here !

Model : chilloutmix_NiPrunedFp32Fix

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

Please do not post NSFW or sexual pictures.

Majestic Containers

Containers with Majestic Lights And Contents.

Important -

• I'll Appreciate your likes and reviews.

• You can use different type of container keywords.

• No human, black background is a must.

• if you are getting girls, put 1girl, human in negative prompts.

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

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

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

https://forms.gle/h9iuLCCnaEs6W9ZB8

Furukawa Sanae from Clannad.

Best Key milf coming through. Use the triple ahoge and diamond sweater vest prompts for best results.

Spent quite a lot of time on this model with stitching, editing and photoshopping frames from the hentai: Bokura No Sex for the character Natsu Kaori, a large breasted tanned life guard/swimming instructor/teacher? who gets up to some rather naughty things during a rain storm. It's obviously NSFW so expect it in this LORA.

Use the tag "natsu_kaori" and tags like tan, blue eyes to get a good representation of her and feel free to check out the metadata for other tags.

This model suffers from some over training, which is a result from lack of training material. She only appears in one episode and has almost zero fan made images. I've made use of the OVA frames and DVD cover images as well as some manga images as much as I can to make it work. However, it will suffer from the "box problem" since a lot of the episode takes place in a pool changing room, with a wood floor, box shelves and grey ceiling. The boxes will turn up a lot, i've tried to counter this with AI created images to add some variety but it's caused other issues. I'm working on another version with the boxes shopped out to white backgrounds but I still need to train that, so expect another version in the future.

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

RIGHT NOW IMAGES METADATA ARE MISSING AS I'M USING COMFYUI TO SIMPLIFY THE IMAGE GENERATION. I'M TRYING TO FIND A WAY TO FIX THIS, IF SOMEONE KNOWS HOW TO DO IT, I'D LIKE TO KNOW.

What's this LoRA

This is a LoRA trained on illustrations and painting inspired by the Gond cultural heritage. 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

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

The LoRA works also on NSFW (check out the examples).

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 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

Mizushima Asa from Sora no Iro, Mizu no Iro.

Using image data from both the hentai, visual novel and other artwork.

Eyes in the hentai are blue, visual novel uses grey eyes so pick which you want. You might have to play with weights with this one but it can give some very good results.

Momoko Hanasaki 花咲桃子 | Wedding Peach

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

地点：(Night:1.7),
动作：sitting at attention,
服饰：red dress with a heart on her chest and a gold ring around her neck,bracelet,choker,white_wings, cherry_blossoms,red_footwear, skirt,yellow_hairband, yellow_ribbon
角色：<lora:MomokoHanasaki-V2:0.9>,
头发：Pink_long_hair,blue_eyes,glint, hair_between_eyes, hairband,collarbone,
角色鎖定：
1 girl,Solo, Manga,20yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,
眼神表情：
looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),(light_smile:0.6),
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, 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: 3375655376, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original, Denoising strength: 0.5, Hires upscale: 2, Hires steps: 18, Hires upscaler: R-ESRGAN 4x+ Anime6B

The Bombchu Bowling lady from The Legend of Zelda: Ocarina of Time. Trained on NAI.

Tested at 0.7-1.0 strength.

Trigger phrase is "Bombchu Lady".

• Add "striped shirt" for her typical look.

• We have no idea what she looks like under the counter, but "pink pants" was used for some of the example images.

trained with anything V3

trigger word:silver hair，long hair，very long hair，ahoge，purple eyea

Miyazawa Yukine from Clannad.

Goes well with the ahoge prompt and a weight of 0.7-1.0

Ibuki Fuko from Clannad

1.0 weight is best, cause prompting in her hair ribbon on your own is very difficult to do.

Fujibayashi Kyou from Clannad.

The prompts "clannad uniform" "blue skirt" work with her regular school outfit. The prompts "white t-shirt" and "red gym bloomers" work with her gym outfit. And of course, don't forget white stockings.

Trained on NAI.

Recommended weigth: 0.6 - 1

Improvement on recreating the uniform, try using RoyalAcademy to get more consistant results.

Feel free to share your result on the comments.

Sakagami Tomoyo from Clannad.

Best used with a weight of 1.0 but with emphasized prompts to avoid excessive stylizing.

I trained this on some ~200 odd AI art images that I personally found interesting and visually impressive.

weigh 0.5~0.7 use hires fix 0.5~0.58

Historical European Uniforms use the trigger word "HistorischeUniformen".

The training has been performed on a selection of 130 historical pencil and watercolour illustrations depicting different European uniform designs from the late 1600s to the and of the1900s

see https://civitai.com/models/43766 for a full checkpoint

Kudou Yukiko LORA

Set weight 0.8

해당 LoRA는 원저작자(본인) 에게 허가 받지 않은 모든 상업적 사용을 금지합니다

This LoRA prohibits any commercial use not authorized by the original author (me)

contact: j36127@gmail.com

The description below is machine translated.

LoRA reduces the frequency of hand distortion and draws pretty hands.

Trigger word: QuickHands

V2 Recommended weight: 0.3~0.6

V1 Recommended weight: 0.4~0.6

Recommende Prompt

positive: 5 fingers, detailed hand

If you want a specific hand shape: open hand, close hand, heart sign, V sign

one hand, two hands

Negative: 3d, realistic(When used for animated models), multiple arm, bad hands, only hand, missing finger

bolded text are those that cause a significant difference in results

Recommended for use with animated models. I haven't tried it on models with other art styles

Difference between V1 and V2

V1 has a more realistic weathering when used on animated models. I haven't tested it, but I think it works well for photorealistic models.

When used with high weights(~0.7), only hand drawn.

V2 is great for animation style models. It almost not changes of the original model's art style.

there have been reviews that it distorts the screen when used on photorealistic models.

I recommend using V2

Art generated after applying QuickHands V2 LoRA with a weight of 0.6

Model: Anything v3

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。

lora强度推荐0.6-0.8，有时候觉得不够像的话还可以试一下0.9或者分层控制1.0。chilloutmix出图，词条看示意图，负面提示词我用了EasyNegative.pt：EasyNegative | Stable Diffusion TextualInversion | Civitai

素材来自网络，用的是老图，年代久远分辨率不高。

请及时下载保存，真人模型随时可能被删除。

喜欢请给五星。

公共频道：t.me/dawn66666666

Kotonoha Katsura 桂言葉 | School Days

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

我很冒險地將檔案壓到很細

所以需要用其他STYLE的LORA去協助提升畫質

原因是我希望製作更多LORA

我也未掌握應該用多少權重令畫質提升

I took the risk of compressing the file down to very small

So you need to use other STYLE LORA to help improve the picture quality

The reason is that I wish to make more LORA

I also don't know how much weight should be used to improve the image quality

地点：(Night:1.7), Cyberpunk_CityView, Before Window,
动作：sitting at attention,sitting on couch,
服饰：red_bowtie, school_uniform, shirt,long_sleeves, pleated_skirt,Black_jacket,thighhighs, black_legwear,
角色：<lora:aiueOkaStyle_v10:0.1>, <lora:saiminSeishidou:0.3>,<lora:Kotonoha-V1:0.85>,<lora:odaNonStyleLORA_v2:0.3>, oda Non
头发：Black_long_hair,bangs, Black_eyes
角色鎖定：
1 girl,Solo, Manga,20yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,
眼神表情：
looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),
畫質:
official art,extremely detailed CG unity 8k wallpaper, perfect lighting,Colorful, Bright_Front_face_Lighting,
(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,
photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2)
身材:
(Beautiful,large_Breasts:1.3), (beautiful_face:1.5),(narrow_waist),
Negative prompt: EasyNegative, 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: DPM++ SDE Karras, CFG scale: 7, Seed: 1928053053, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original, Denoising strength: 0.5, Hires upscale: 2, Hires steps: 18, Hires upscaler: Latent (bicubic antialiased)

Here is the Lora for gang bang!

Here is the trigger word: gangbang

You may try with these prompt: vaginal, fellatio, hand job, anal

Hope you enjoy my work!

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

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

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

Here is the Lora for footjob!

The trigger words is footjob

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

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

Ajijoy model

https://t.me/stdif_kos

Subaru introduced the "New Age" Impreza to Japan in August 2000. Larger in size compared to the previous iteration, the sedan increased its width by 40 millimetres (1.6 in), while the wagon notably increased by just 5 millimetres (0.2 in)—placing the two variants in different Japanese classification categories. The coupe body style from the first generation did not reappear for the new series, and the off-road appearance package that included contrasting-colored bumpers did carry over forward. Marketed as a separate model line, this North America-only variant was, as before, badged the Outback Sport.

Naturally aspirated flat-four (boxer) engines comprised the 1.5-liter EJ15, the 1.6-liter EJ16, the 2.0-liter EJ20, and the 2.5-liter EJ25. Turbocharged versions of the 2.0- and 2.5-liter engines were offered in the WRX and WRX STI models. STI models featured a more powerful 2.0-liter (2.5-liter outside of the Japanese market) turbocharged engine. WRX models featured a 2.0-liter turbocharged boxer engine until 2005, after which they switched to the 2.5-liter turbocharged engine.

At the Australian Wheels Car of the Year, the car won the overall COTY award in 2000. At the American MotorWeek Drivers' Choice Awards, the Impreza won the Best Sport Sedan in 2002 and the Best Performance Car in 2004.

⌚ Time traveling detective! Now with all 4 of her outfits and all accessories. Check the second image for basic prompts and the other images or the right side for extra tags.

Works at weight 1 but might need hires fix sometimes.

Made using my Lora guide.

The Pineapple Toppin' Gal based off of the Pineapple Toppin' from Pizza Tower.

Probably gonna redo this one cause it's not very good in many ways.

Recommend using DreamShaper, with weight of 0.7.

This model has two characters and supports the presence of two people in one screen.

We need to define the roles in this format: "a {token} boy" or "a {token} girl"

To prevent conflicts with the semantics of the original model, I defined separate tokens:

For kirito, the token is tongren

For asuna, the token is yasina

For example:

a yasina girl, sunshine, wearing hat

a tongren boy, and a yasina girl, chibi

感谢评论区kikuwithaurora470网友提供了Inkyung相关图片资料，本模型仅供测试，请不要用于商业！

之前发布的陈潇LORA模型链接 欢迎下载！

陆续会更新更多角色的LORA

喜欢什么角色欢迎评论，提供优质图片资源的，优先采纳!

Miranda Vanderbilt from the game series Monster Prom

Helpful positives: pink skin, crown, head fins, jewelry, dress

Vivid and elegant, niji expressive style lora

Do you suffer from sd style rigidity?

Do you envy niji v5 for its action and camera?

Now, you can experience the same exquisite style on sd

Trained by 500+ quality niji pictures

no trigger words

Do well on Novel ai original model

weight 0.9 ? or other, have fun!!

本LORA以技术测试作为主要目的，目标是整理出一套可复制、可持续使用的方法，目前已经测试过33次不同的方式，欢迎大家一齐研究，加入频道：https://discord.gg/CxynduC63t

I will display this model publicly for academic research and discussion. I hope to discuss working methods with you. I strongly advise against uploading any NSFW generated by this model. You are warmly welcome to communicate with me here on how to improve the working method and make the model more realistic. If you have any suggestions, such as choosing a different base model, training with more images, increasing the number of training steps, or changing the training size for a more realistic overall effect, please feel free to leave a comment below.

我會把這個模型公開展示出來，目的是學術研究與討論用，希望和大家一起探討工作方法。我強烈建議不要上傳任何由此模型生成的 NSFW。熱烈歡迎您在這裡與我交流如何改進工作方法，使模型更逼真。如果您有任何建議，例如選擇不同的基礎模型、使用更多圖像進行訓練、增加訓練步數或更改訓練大小以獲得更逼真的整體效果，請隨時在下面發表評論。

本人原則上不支持使用此模型生成任何成人內容。用戶在使用本模型時，請務必遵守當地法律法規，勿侵犯他人的名譽權、隱私權或肖像權。

如本lora侵犯了您的权利，请联系我，将会下架处理。

底模：chilloutmix_NiPrunedFp32Fix

V1-test：未经完全测试，upperbody 0.7权重较像，大头可能会崩

Lora of actress Radha Mitchell

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

Menat from Street Fighter 5

There is already a great LoRA of Menat made by Murky, but his didn't had the holiday costume which is my favorite, so i trained her as well.

This Lora supports 3 of Menat's costumes, classic, holiday and mummy

• for classic costume use "menat_classic" and add to your prompt

  • bodysuit, shorts, thighhighs, circlet, navel cutout, shawl

• for mummy costume use "menat_mummy" and add to your prompt

  • hood, bell, red dress, bare shoulders, paw gloves, tail

• for holiday costume use "menat_holiday" and add to your prompt

  • hood, bell, red dress, bare shoulders, paw gloves, tail

You may also want to add "dark skin, green eyes, egyptian, bob cut", to your Prompt.

Trained on Nai

Can be used at 1.0 weight

All images were generated using yabalMix (https://civitai.com/models/28648/yabalmix) with xformers on and using kl-f8-anime2 vae.

images were generated at 512x768 and upscaled using highres fix by 1.95 and using 4x_foolhardy_Remacri upscaler

模型融合了不少韩游画风，这版本主要会强化画面的风格质感，对各种装饰品，挂件，武器等产生风格影响，在写实类画面中，各种质感关键词添加越多画面越有趣,

推荐强度0.6-0.9

推荐搭配CKPT：revanimated

样图关键词分享：

(masterpiece), (best quality:1.2), Egyptian clothing, ancient Egypt, pyramids, deserts, ancient Egypt,outdoors,Anubis head, Pharaoh's scepter,upper body,(best quality:1.2),masterpiece:1.3,HD:1.32,space reflection,ultra-detailed,best illustration,colorful,(Master works),8K wall-paper,highly realistic,huge_filesize

perfect lighting:1.3,cinematic lighting,Lens Flare,vivid colours,lmaginative,atmospheric,intricated details:1.4,Perfect detail:1.2,movie poster, pov hands, full-body tattoo, mourning,

Fantasy:1.4,(Medieval battlefield),beautiful background,Perfect detail,Flying spark,(__color__ smoke),Grand battle,epic,cinematic lighting

(1girl:1) enchanting gaze, captivating pose, delicate wings, otherworldly charm, mystical sky, (Luis Royo:1.2), (Yoshitaka Amano:1.1), moonlit night, soft colors, (detailed cloudscape:1.3), (high-resolution:1.2),(perfect body),(perfect face),__Emotions__,__eyes color__. __hair color__,__hair__,__hairstyle__,__iris color__,

, silver scales, sharp claws, dragon eyes glowing red,Fierce battle with dragon:1.3,One pair of wings, four claws,1dragon,epic fantasy, holding weapon, epic weapon

Shika Eiseishi-san | Tawawa on Monday

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

https://anidb.net/character/87727

地点：(day:1.7), dentist clinic,

动作：sitting in a chair with a computer on her desk,

服饰：pink_sweater,mouth_mask,long_sleeves,face mask,(White_pencil_skirt),

角色：<lora:ShikaEiseishi-V1:0.7>,dentisit,<lora:KisekiHimuraLocon_kisekiHimura4Locon:0.3>,

头发：bangs, brown_eyes, brown_hair,hair_ornament, hair_over_shoulder, hair_scrunchie, hairclip,

角色鎖定：

1 girl,Solo, Manga,20yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,

眼神表情：

looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),

畫質:

official art,extremely detailed CG unity 8k wallpaper, perfect lighting,Colorful, Bright_Front_face_Lighting,

(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,

photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2)

身材:

(Beautiful,large_Breasts:1.4), (beautiful_face:1.5),(narrow_waist),

Negative prompt: EasyNegative, 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: 913484534, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original

A brief intro to the iconic Amsterdam Windows: https://thingstodoinamsterdam.com/red-light-district/

Weight: <= 0.8; see samples for the prompt

Tested dimensions:

• W600 * H900

Note:

• To avoid generating white boxes on a character's face, add ((white box)), ((face cover:1.2)), to the negative prompt

• If you end up generating too many window frames, add ((window)) to the negative prompt to mitigate

Character LoRAs used in the samples (credits to the authors for their excellent works):

• Akane Tsunemori (created by https://civitai.com/user/Deto15): https://civitai.com/models/16485/tsunemori-akane-psycho-pass

• Firekeeper (created by https://civitai.com/user/Akito): https://civitai.com/models/8214/firekeeper-lora-from-darksouls3

• Konan (created by https://civitai.com/user/Rindo192): https://civitai.com/models/15520/konan-naruto-lora

• Kurisu Makise (created by https://civitai.com/user/Imoto4155): https://civitai.com/models/18364/makisekurisuchristinasteins-gate

• Malenia (created by https://civitai.com/user/C_zZz_K): https://civitai.com/models/7588/character-malenia

• Mashiro Shiina (created by https://civitai.com/user/wizzlol): https://civitai.com/models/11279/mashiro-shiina-sakurasou-pet-na-kanojo

• Megumi Katou (created by https://civitai.com/user/Deto15): https://civitai.com/models/19371/katou-megumi-saekano

• Pecorine (created by https://civitai.com/user/leirbag421): https://civitai.com/models/21271

• Rin Tohsaka (created by https://civitai.com/user/kozue): https://civitai.com/models/11798/tohsaka-rin-fate

• Suzuha Amane (created by https://civitai.com/user/sunismeszp336): https://civitai.com/models/13366/amane-suzuha-lora

• Chisa Yukizome: https://civitai.com/models/32285/chisa-yukizome-danganronpa-3

• Dextra: https://civitai.com/models/41872/dextra-arena-of-valor

This is a LORA of the character Brighid from the Monolith Soft JRPG Xenoblade Chronicles 2.

To prompt this character, enter "<lora:Brighidv1:strength>Brighid, purple long dress, jeweled collar, accessory belt, closed eyes, closed mouth,". To prompt her holding her whipswords, enter "dual wielding, holding swords," Can be prompted with a 3D aesthetic using "3D Model".

This LORA looks the best at 0.6 ~ 0.8 strength in my opinion. Brighid was trained on AnythingElsev4.5 at Clip Skip 2, regardless of what this site might tell you otherwise. Really like the outputs this LoRA makes.

Mainly Mikecat or tiger cat will be available.

My second favorite Sekirei the first being Tsukiumi which i have already made. All in all the LORA seems to work nicely.

I was able to more or less reliably make it spit her veiled sekirei guise and her normal outfit. I have the feeling that it is just a tad overcooked which might be due to the limited amount of images but seems to be able to go up to 1.15 before going horribly wrong.

I had good results with: <lora:SekireiUzume:0.7>, SekireiUzume, side_ponytail, large_breasts, brown_hair

Her available outfits are:

• VeiledSekirei

• UzumeOutfit

Her hair color tends to get a bit reddish so you might want to set it as brown_hair. If you omit side_ponytail and add long_hair you might get her loose haired version.

I have added the tags summary inside the image dataset if you want to experiment with the available tags.

Sophie XI Sophie Elisabeth Irail | Isekai Harem Monogatari

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

地点：(Night:1.7), in the Castle,medieval ,

动作：sitting at attention,sitting on the throne,

服饰：medieval costumes,(white_bridal_veil:1.1), cleavage,

角色：<lora:aiueOkaStyle_v10:0.1>, <lora:saiminSeishidou:0.1>,<lora:SophieXI-V1:0.7>,Queen,<lora:odaNonStyleLORA_v2:0.2>, oda Non,

头发：blonde_long_hair,braid,jewelry,

角色鎖定：

1 girl,Solo, Manga,20yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,

眼神表情：

looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),

畫質:

official art,extremely detailed CG unity 8k wallpaper, perfect lighting,Colorful, Bright_Front_face_Lighting,

(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,

photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2)

身材:

(Beautiful,large_Breasts:1.4), (beautiful_face:1.5),(narrow_waist),

Negative prompt: EasyNegative, 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: 2944401018, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original

Trained on 220 tagged images from e621

clip skip 2 was used. All characters belong to their rightful owners.

Some tags this was trained on:
dipstick antennae - prompts antennae (optional)

plasma blaster (lilo and stitch) - prompts weapon

spacesuit - prompts space suit

Tesla Model X trained on 768x768 images

0.6 - 0.8

Tag - Photo_comparison from Sankaku

Important -

• This is the BETA Model. It can produce good results but here are the settings to get those. I'll appreciate your Reviews and Likes, but please don't downrate it too much its beta.

• You need a Wide Shot image, like Width=>Height, Width should be at least 768.

• There are two separate prompts necessary. One for Background Waifu, Second for Waifu in Picture.

• Put tag - Photo_Comparison, NSFW first, then describe BG waifu, then put phone, cellphone, smartphone, holding, Picture (you can also use other similar prompts), After that put Picture Waifu prompt, ex - Sex, Vaginal, Nude, NSFW etc.

• Attached a Picture of My Generation Folder, it shows the chances of getting the pics.

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

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

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

https://forms.gle/h9iuLCCnaEs6W9ZB8

LoRA based on artworks from artist sl8-all.

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

There is a LoRA of her already.

Weight: 0.7 - 0.8

Trigger + Outfit: clementine \(overlord\), armor, garter belt, gauntlets

Suitable for use with BB95 Furry Mix v3

Hires fix will sometimes cause oddities. This Lora works well with lower CFG scales.

I suggest adding the following at the beginning of the prompt:
(((cigarette, (cigarette in mouth:1), smoking, smoke cloud, cigarette smoke))) <lora:quentins_smoking_LoRA_v1:0.6>

All image were generated with Anything-4.5,

but it should work with most models

Maybe it should be called GreenhousePunk.

I've noticed that many creators tend to delay uploading their models, waiting for download counters and likes to be fixed. However, I don't wait (in fact, I upload even more now). I'm not chasing the numbers; I'm simply delivering my best work to you all. Enjoy!

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

Also, don't hesitate to send me suggestions. I tend to alternate between creating what I want and what I'm asked for. Rest assured, nothing goes unnoticed.

A LoRA based on Latin streamer Didiwinx

I could improve I accept tips

I publish it to receive suggestions and improve the model together with the community. Please let me know your considerations.

It was made with the default configuration of the colab of

Hollowstrawberry Guide

LoRA based on artworks from artist Atte Nanakusa.

The example resolution is 512x1024 and clip skip is 2.

black or white pantyhose

LoRA based on the artstyle from artist Hijiri Rei.

Hs.50 from Girls' Frontline
Features 2 outfits

Keyword:
Hs50
Hs50Dress

Recommend 0.8-1 strength

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

Lora trained from Iahfy's art. I'd recommend no face restore, DPM++ SDE Karras, and probably >30 sampling steps, strength weight of 0.9 was used for photos.

LoRA based on the artstyle from artist Omodaka (nitera1041).

Lora of actress Linda Fiorentino

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

This LoRa is trained on studio photos of cars from the 80s and 90s, as well as a little bit of the 70s.

You can get some nice concepts for inspiration. It can also be used with people and landscapes, but the results are not guaranteed.

It is better to add AutoStudio at the beginning of the prompt, this is the keyword for the LoRa.

To create unusual colours and lighting, it's recommended to use prompts like [colour1:colour2:step]. For example [red:green:15], which means that the first 15 steps will use red and after that green. This also applies to body types, car brands and other things.

In general, it works about the same on all models. For sure you can get variety of styles on different models as well.

Recommended parameters:

Resolution: 768 x 512

Samplers: DPM++ 2 and DPM SDE series. You may try others on different models.

25-100 steps.

Hires.fix is recommended, especially 4x-UltraSharp.

CFG Scale Fix: CFG Scale 14-20, mimic scale 7-10. For more information, read the extension description here.

You can see examples of prompts in the gallery.

55tan is the mascot from former brazilian website 55chan.

Gym Leader Valerie (Mache in JP) from pokemon x and pokemon y

Use valerie_mache as trigger word.

Trained on Nai

You may want to add "sidelocks, pantyhose, sleeves past wrists", to your Prompt

Can be used at 1.0 weight but you should lower it if you want to change her clothes

All images were generated using yabalMix (https://civitai.com/models/28648/yabalmix) with xformers on and using kl-f8-anime2 vae.

images were generated at 512x768 and upscaled using highres fix by 1.95 and using 4x_foolhardy_Remacri upscaler

This is a LoRA for generating images of Steins;Gate character Tennouji Nae.

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

Recommended weights:

• Recommended weights: 0.5~0.7

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

這是一個用於生成Steins;Gate角色天王寺綯的圖像的LoRA

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

建議權重：

• 建議權重：0.5〜0.7

  （注意：如果權重低於0.4，則LoRA可能無法生效。）

Trained for 2000 steps on the phrase:
#:0.04|~fluffy:0.2

https://github.com/ntc-ai/conceptmod

movies created with the lora applied from 0 to 3.5 .

animation with:

https://github.com/ntc-ai/conceptmod/blob/main/lora_anim.py

Enome anime Futoku no guild work with 0.6 weight enjoy it

Triggers word: enome

follow me on Instagram @creativehotia

However, that's another story altogether

A lora based on the character Enju Aihara from Black Bullet

Use at 0.8~1.0 weights

• Use triggerword "enjudefault" for her default yellow outfit

• Use triggerword "enjucasual" to summon her in her hoodie + short/skirt outfit

• Use triggerword "enjudress" for the nightdress outfit

The dataset for both the dress and casual wear were really small (80% screencap) and
might be restrained in terms of poses and environnement

The casual outfit was trained without the color tags, you can influence its colors by adding a color tag (not perfect tho)

Some output from the casual outfit might have blur effects.
Using highres.fix at minimum 1.3 should completly negate it.

Tested on :

• Greymix

• MidnightMelt

• AOM3

• Kotosmix

• Sweetmix

• Henssaku

• Mandarine
  

If you have any questions feel free to ask.
Kaïros#4463

Enjoy ♥

A request from a Kofi donator
R93 from Girls' Frontline
Features 2 outfits

Keywords:
R93
R93Swimsuit

Not to be confused with a boat
Recommend 0.8-1 strength

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

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

The tall woman from Made in Abyss.

April 17, 2023

Fellas, I bring good news: she now has thousand-men pins in her arms (and maybe legs)! I tried giving her the hat, but I guess I should've put those images into its own dataset. Regardless, I did learn something from this experience and I can't wait to do more like this.

You'll have to take her top off (I asked her for permission, don't worry) by inputting ozentmp instead of ozen, then adding either topless or nude. You may add no gloves, but I don't know if that's necessary.

April 14, 2023

Weight: 0.7 - 0.8

Trigger + Default Outfit: ozen, black jacket, cape, black pants, black gloves

This does not have her limbs full of thousand-men pins.

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

A quickly baked LoRA.

Weight: 0.7 - 0.8

Trigger: gfl-m1911

Add ahoge to the positive prompt if it's not showing.

Default Outfit: jacket, open jacket, short sleeves, white collared shirt, striped necktie, white skirt, white thighhighs, white gloves, brown footwear, boots, belt

Breaker of the Sky: blue hair bow, blue frilled dress, single thighhigh, white elbow gloves, blue footwear, high heels

The dress may not be fully blue.

Role of Girls' Frontline

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

TRIGGER:

1.BASE: original outfit;

2.OG outfit: dress, asymmetrical legwear, mismatched legwear

ᶘ ᵒᴥᵒᶅ
ᶘ ᵒᴥᵒᶅ
ᶘ ᵒᴥᵒᶅ

The Zastava M21 is a selective fire assault rifle designed by the Zastava Arms company in 2004. The rifle is essentially a Kalashnikov pattern rifle chambered for the 5.56x45mm NATO cartridge, leveraging Zastava's lengthy experience with manufacturing the AK platform. The M21 is the current standard issue service rifle for the Serbian Army.

The Zastava M21 is a gas-operated rifle with a rotating bolt, and utilizes the legendary long-stroke gas piston system that the Kalashnikov is known for. The M21 is chambered for the 5.56x45mm NATO cartridge, and is fed from proprietary 30-round magazines. The rifle is able to achieve a rate of fire of 680 rounds per minute, with an effective range of 450 meters when using iron sights. During the development process for the M21, Zastava Arms experimented with different barrel forging processes. As a result, the M21 has the option of using barrels with either traditional rifling or polygonal rifling. Regardless of what it chosen, all M21 barrels are hard chromium-plated to extend service life. On top of the operating mechanism, the M21 also has many similar design features of a Kalashnikov rifle, such as the charging handle on the right side, a behind-the-well magazine catch/release, and the trigger group/fire control lever style. This safety lever toggles the rifle between full, semi auto, and safe positions.

Like most other modern assault rifle designs, the M21 is designed to be modular, allowing the operator to attach different combat accessories for different situations, such as bipods and vertical fore-grips. A derived model of the Soviet GP-25 underslung 40mm grenade launcher unit was designed for and issued for M21s, and 22mm rifle grenades can also be launched from the muzzle flash hider. To improve compatibility with NATO-issued accessories, M21s can also undergo modifications and mount Picatinny accessory rails. The M21 also features a foldable stock made of rigid polymer and rifle forearm guard made of heavy-duty synthetic material. The rifle has conventional iron sights modeled after the ones on the Finnish RK 62 and the Israeli Galil, with the front and rear sight being relocated to the gas block and top cover respectably. It also borrowed the Galil's thumb safety, and the top cover is hinged like the AKS-74U.

Besides the main battle rifle variant, the M21 A, the rifle also has two short barrel variants: the M21 S (an assault rifle with a shorter barrel) and the M21 C (a carbine). The variants differ mostly in terms of barrel length, overall weapon size/dimension/weight, and firing performance in terms of accuracy/effective range

All image were generated using Anything-V4.5 , except the realistic one that was made using Deliberate-v2

Ghost Sweeper Mikami

It is my first parrot and the truth gave good results.

The keyword is "mikami"

If you like it, put photos of your tests and suggestions are appreciated!

lora experimental pixelart style works with 0.8 you can also try it at 1 you can vary

Triggers words: pixelart

model recomended: Revanimated (you can try other models and notify me in which ones it works better thanks)

follow me on Instagram @creativehotia

v2
14k more steps.
based on 1,5 normal for bigger compatibility with other mothers:

Data available in lora file

ss_mixed_precision:  "bf16",ss_network_alpha:  "512.0",ss_network_args:  "{"conv_dim": "16", "conv_alpha": "16"}",ss_network_dim:  "512"

More info on the Fusion project is soon to be announced ;)

🤟 🥃

Easter Lora experiment / full info +training data embedded on the lora itself.
based on a few(~6000) Midjourney handpicked photos

prompting is not done right / fixing for next easter :D

try clip skip 2-12

"ss_sd_model_name": "runwayml/stable-diffusion-v1-5", "ss_resolution": "(768, 768)", "ss_clip_skip": "2", "ss_num_train_images": "6836", "ss_tag_frequency": { "2_easter easter": { "there are three eggs that are cracked and one is still in the process of being eaten": 1, "eggs": 411, "egg yolk": 128, "an egg": 469, "hyper - realistic render": 373, "hyper-realistic render": 363, "hyperrealistic render": 361, "photo still": 28, "high quality food photography": 13, "hyper realistic render": 290, "art render": 29, "high speed photography": 6, "raw egg yolks": 10, "super high speed photography": 5, "portrait shot": 217, "simple background": 280, "no humans": 782, "bug": 18, "black background": 158, "butterfly": 10, "brightly lit easter egg with butterflies and swirls on a dark background": 1, "ornate egg": 975, "beautiful digital artwork": 154, "detailed digital 3d art": 70, "high detailed digital art": 72, "cute detailed digital art": 152, "4k highly detailed digital art": 43, "highly detailed digital art": 141, "overdetailed digital art": 81, "stylized digital art": 28, "very detailed digital art": 38, "hyper detailed digital art": 38, "intricate artwork. octane render": 103, "there is a large egg sitting on a rock in the dark": 1, "daily render": 24, "created in octane render": 92, "rendered in octane render": 56, "super rendered in octane render": 29, "rendered in octane render 32k": 125, "(octane render": 1, "rendered in cinema 4 d": 401, "rendered in cinema4d": 398, "3 d render even lit": 3, "( octane render )": 1, "rendered in cinema 4 d octane": 126, "space": 14, "planet": 48, "colorful": 53, "earth (planet)": 12,

Do you want to support me? Consider being a patron on Patreon to vote on train polls and access stuff early. ❤️

Trained on 80 image and 768x768 resolution. For better quality, use hires fix to generate images.

Trigger Word is KokonoeMS

Tags you'll need for her looks are

animal ears, cat ears, pince-nez, tail , cat tail, multiple tails, solo, pants, navel, belt, capri pants,

Use the LoRA weight around 0.8 and 1 for the best results

Discord server feel free to drop and say hi ^^

Do you want to support me? Consider being a patron on Patreon to vote on train polls and access stuff early. ❤️

Trained on 120 image and 768x768 resolution. For better quality, use hires fix to generate images.

Trigger Word is RachelMS

Tags you'll need for her looks are

red eyes, bow, ribbon, gothic lolita, lolita fashion, dress, frills,

Use the LoRA weight around 0.8 and 1 for the best results

Discord server feel free to drop and say hi ^^

Leonie Pinelli from Fire Emblem: Three Houses. Trained on NAI.

Tested at 0.8-1.0 strength.

Contains both academy and timeskip outfits. Trigger phrase is "Leonie Pinelli" for both.

• Add "short hair, white shirt, garreg mach monastery uniform, black corset, black skirt, clothes around waist" for her academy look.

• Add "hair over shoulder, orange shirt, shoulder armor, fingerless gloves, black shorts" for her timeskip look.

Do you want to support me? Consider being a patron on Patreon to vote on train polls and access stuff early. ❤️

Trained on 200 image and 768x768 resolution. For better quality, use hires fix to generate images.

Trigger Word is MaiNatsumeMS

Tags you'll need for her looks are

bow, ribbon, sidelocks, bare shoulders, halterneck, backless outfit, sideboob, pants, lowleg pants,

Use the LoRA weight around 0.8 and 1 for the best results

Discord server feel free to drop and say hi ^^

Due to the limited training data, I couldn't really make it perfect, so I split it up.

Fix looks better but is stiffer. Flex is easier to change clothes with.

Trigger word is Ochette.

Weight ~1.

Use "white tank top", "purple loincloth" and "armlet" for her clothes if you use the flex version. Also "green eyes" for both versions.

Trained on NAI.

Do you want to support me? Consider being a patron on Patreon to vote on train polls and access stuff early. ❤️

Trained on 150 image and 768x768 resolution. For better quality, use hires fix to generate images.

Trigger Word is MakotoMS

Tags you'll need for her looks are

animal ears , squirrel girl, squirrel tail, squirrel ears , antenna hair, tail , two-tone hair, brown eyes, gloves,skirt, underboob, orange skirt, midriff, bare shoulders thighhighs, highleg panties, tonfa

Use the LoRA weight around 0.8 and 1 for the best results

Discord server feel free to drop and say hi ^^

💖🖤 Homura Kougetsu (Edens Zero) 🖤💖

This model was trained on 168 images, with 3 repetitions, a batch size of 2, and 10 epochs, equivalent to 2520 steps.

The model in question that was used was: animefull-final-pruned-fp16.safetensors

I don't recommend a specific weight, try out whichever you like best. In my case, I like to leave it at a weight of 0.7.

Activation Tag:

homurakougetsu

Keywords that you can use for this Lora:

thighhighs, detached sleeves, japanese clothes, black thighhighs, kimono, sash, obi, sandals, hair ribbon, ribbon, yellow ribbon, red eyes, hair ornament, bare shoulders, flower, choker, hair flower, playboy bunny, leotard. (see example images for details).

Installation and model:

Intalallation:

Option 1: To install it, download it and add it to your "Lora" folder in Stable Diffusion (stable-diffusion-webui\models\Lora).

Option 2: If you are in Collab you can add a code box place this command:

#Add LORA in the Cloud 
!wget https://civitai.com/api/download/models/48471 -O /content/stable-diffusion-webui/models/Lora/homuraKougetsuEdens_v10.safetensors
 
#Add LORA in the Drive
!wget https://civitai.com/api/download/models/48471 -O /content/drive/MyDrive/sd/stable-diffusion-webui/models/Lora/homuraKougetsuEdens_v10.safetensors

Model:

Regarding recommended models, I really like Anything-v4.5 with the VAE anything-v4.0.vae.pt

Finally, you can leave me any suggestions in the comments, they are greatly appreciated 😊💜💚

LoRA for rolled short sleeves.

It may also help in drawing tied shirt.

This model is trained as Kokkoro, a character from Princess Connect! with her original outfit. It's based on SD1.5 model, it would competiable with other checkpoints based on SD1.5. Safe to blend with other LoRA, LoCon based on SD1.5 as well.

Triggered fine with "kkr, origin"

Optional to be reinforced with "white hair, short hair, sandles, bridal gauntlets, white flower, collarbone", for full body images and better lock down details of the character.

Add "pointy ears, elf" to match with the elf looking.

Suggested weight 0.8-1.0, works fine with 1.0.

Tested with Chilloutmix Ni, Anything 4.5, AOM3A3, CounterfeitV25, meinamix_meinaV8, perfectWorld_v2Baked, neverendingDreamNED_bakedVae, check the samples for details (Hires Fix suggested with realistic models).

It's not trained to change to other outfits, still it's possible, lower the weights to 0.8 or below while enhance the alternative outfit description, see the samples for ref.

Have fun!

**It's already been iterated for seveal version before publishing, should be stable enough to use. :)

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

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

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.

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

https://ko-fi.com/reindeerczar

Lora trained on 209 images of Sistine Fibel.

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:

sistineFibeldef, sistineFibelrnd,

Trained on NAI, gen'd on AOM2 hard.

The Juice is in the description below

Tips to use:

• Use LoRA weight between 0.6-0.8

• To summon her use the trigger Tags plus following tags: solo, long hair, blue footwear, blonde hair, armor, yellow eyes, very long hair, thighhighs, breastplate, gloves, thigh boots, white dress, hair between eyes, zettai ryouiki, shoulder armor,

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

If you have any questions feel free to ask!

Hope you guys like it! ❤️

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

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

The base model is Suzumehachi.

If you like my LoRAs please consider to support me by buying me a coffee here !

Model : chilloutmix_NiPrunedFp32Fix

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

Please do not post NSFW or sexual pictures.

Kiana-Herrscher if Finality

As before, this lora was based on Meinaalter——

https://civitai.com/models/20945/meinaalter

if your outputs are undesirable, try this ckpt or the series of Meina XD

Notice：The characters are copyrighted by miHoYo, this lora for exchange leaning only, do not use for commercial purposes.

琪亚娜，终焉之律者。

和以前一样，这款lora是基于Meinaalter——
https://civitai.com/models/20945/meinaalter

如果你的输出结果不理想，请试着使用Meinaalter或者Meina系列ckpt

注意：角色版权归米哈游公司所有，本lora仅供交流学习，请勿用于商业用途。

weight 0.7 is best.

shionne1, ponytail, 1girl, solo, shoulder armor, dress, very long hair, thigh strap, white dress, cleavage, pauldrons, white thighhighs, medium breasts, bangs, detached sleeves

weight 0.7 is best.

mikuru1, 1girl, asahina mikuru, long hair, kita high school uniform, solo, blue sailor collar, serafuku, blue skirt, short sleeves, red ribbon, large breasts

mikuru1, 1girl, asahina mikuru, twintails, solo, large breasts, apron, puffy short sleeves, (name tag:0.9), long hair, corset, waitress, white apron, wrist cuffs, pink dress, frills, bangs, maid, ribbon

mikuru1, 1girl, asahina mikuru, long hair, solo, maid, long sleeves, puffy sleeves, blue dress, juliet sleeves, white apron, dress, maid apron, frilled apron, mary janes, bangs, large breasts

mikuru1, 1girl, asahina mikuru, long hair, solo, playboy bunny, animal ears, solo, rabbit ears, pantyhose, wrist cuffs, bowtie, detached collar, cleavage, red leotard, large breasts

weight 0.7 is best.

lilith1, 1girl, solo, braid, hat, beret, long hair, elbow gloves, black thighhighs, sleeveless, black gloves, black necktie, sleeveless shirt, large breasts, bare shoulders, pleated skirt, black skirt, white shirt, miniskirt

weight 0.7 is best.

sui1, 1girl, solo, side ponytail, hoshimachi suisei, fingerless gloves, single thighhigh, jewelry, single sock, thigh strap, bracelet, blue socks, buttons, single kneehigh, plaid dress, blue choker, blue belt, plaid skirt, mini crown, grey skirt, blue ascot, long sleeves, plaid jacket

sui2, 1girl, solo, one side up, hoshimachi suisei, orange pantyhose, green ribbon, black skirt, star \(symbol\), off shoulder, blue jacket, jewelry, short sleeves, thigh strap, black shirt, lace choker, miniskirt, neck ribbon, green bow, black choker, belt

sui3, 1girl, hood up, hoshimachi suisei, solo, white jacket, animal hood, white socks, open jacket, pleated skirt, black skirt, black sailor collar, black choker, school uniform, black shirt, puffy long sleeves

"ai" does not express the shape of the hand wearing fingerless gloves. This is an unavoidable phenomenon.

You can modify the shape of the hand using the in-paint function, or you can print out a normal hand by excluding the fingerless gloves prompt.

I apologize if the example prompts seem long. But I take costume reproduction and quality very seriously.

weight 0.7~0.8 is best.

1girl, hair over eyes, heterochromia, blue eyes, yellow eyes, solo, frankenstein's monster \(fate\), mechanical horns, veil, elbow gloves, white dress, long dress, white gloves, short hair, pink hair, bare shoulders, medium breasts, headgear

This was the hardest LoRA I've ever made.

The reason was that the dataset had too many different images of her frontal hair. One with both eyes covered, one with only one eye, and so on.

I could use the prompt example and cover both of her eyes with hair, but the difference in image quality from checkpoint to checkpoint is too great to write about in the description.

This is my second LoRA from a Queen's Blade character, Menace, the woman inspired by the egyptian queen Cleopatra. It's trained with Anything V3 checkpoint and the Kohya-ss GUI utility, and works very well with anime checkpoints and photorrealistic ones too (you can see some examples of this in the sample images).

Works well with strenghts between 0.6-1. With strenghts below 0.9, it's convenient to write in the prompt some of her caracteristic clothes if you want her with her traditional appearance. Like this:

Menace, 1woman, crown, underboobs, striped panties, bob haircut

This LoRA, like the character itself, it's very nsfw, so if you don't want her showing her nipples or in hot posses, write on the negative prompt the words 'topless', 'nipples' and 'nsfw'. For good results, use in the negative prompt (worst quality, low quality:1.4) and/or negative embeddings like EasyNegative.

Setra, the living cat-like scepter of Menace, usually doesn't spawn well almost never, so I recommend to put in the negative prompt 'cat' and 'scepter' so it doesn't show up in the images.

I recommend to use hires.fix to upscale to x1.5-x2, with 10 hires steps, 0.3 of denoise strenght, and 'none' in upscaler.

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

"I'm sugar and spice."

Weight: 0.8

Trigger: piper

Default Outfit: blonde hair, ponytail, blue dress, puffy short sleeves, fingerless gloves

Mariposa Piper Skin: orange hair, golden butterfly hair ornament, black sleeveless dress, pelvic curtain, red thigh strap, black elbow gloves, fingerless gloves

Negative: animal ears

Choco Piper and Lunar Piper are included within the dataset but hard to replicate, which is surprising considering they're older than the release of Mariposa Piper. You can get her purple hairstyle from Lunar though. Putting her in custom clothing can get unstable (e.g., animal ears popping up, double hair bun).

Wow, how I love your fallen breasts 😍

This is a lite version of full Lora that you can see in my Patreon.
Preview of the full version, with Full LoRA you gain consistency, better nipples and more details.

LoRA prompt weight: 0.6 to 0.8
Model used for training: Uber Realistic Porn Merge 1.3.

Recomendations:

• CFG ~9

• Steps >=27

• DPM Family or DDIM

Prompts that can improve your results: large_areolae, huge_breasts, stretch_marks, pale_areolae...

Tendencies of this LoRA (and resolution):

• This LoRA have a tendency of close-up on breasts. (close-up:0.3) and (breast_focus:1.3) on neg can improve results.

• The model changes faces a bit, you can inpaint or put some prompts to improve the face details.

• The model also ages people, you can put young on positive and milf in neg.

See my other LoRAs:

• Veiny Breasts (only for Longer Prompters+ on Patreon)

• Puffy Nipples

• Perfect Perky Breasts

• Tight Bikini Top

• Big Nipple Tip [Lite]

• Small Breasts [Lite]

• Big Natural Breasts Type01 [Lite]

• Big Natural Breasts Type02 [Lite]

• Ghost Nipples [Lite]

• Underwater Naked Shot

• Fake Breasts [Lite]

Testing only

Lora trained on Gabimaru from Hell’s Paradise

Master model anime: AnyLoRA, Based65, meinaunreal, AbyssOrangeMix3

2,5D: fantasticmixReal_V30, ReV Animated

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

Recommended settings:

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

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

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

• Steps: 20 to 35

• Clipskip: 1 or 2

• Upscaler : img2img in comfyui using 4x-UltraMix_Smooth

• CFG Scale : 5 to 9 (mostly 7)

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

• Denoise strength for img2img: 0.4 to 0.7

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

Trained with AnyLoRA checkpoint

for the pictures i used comfyui

The model was trained on Deliberate 2.

I advice to set it to 0.6-0.7

tag: 1girl

There is less data and more clothing, so the characteristics will be a bit messy. For specific tags, please refer to the example image information, but at least this must be an adult, so I uploaded the real image with confidence

数据较少，服装较多，所以特征会有些乱，具体的tag请参考示例图信息，不过至少这肯定是个成年人了，所以我放心的上传了真实图片

maronMaron circle Style LORA

• you can add "moronMaron" to the prompt to help anchor the style even more

• Tested on Anyhentai19 , Grapefruit mix and AOM2 H.

• Samples made with AOM2 H

Looking forward to see your creations 😉

If you like my LoRAs please consider to support me by buying me a coffee here !

Model : chilloutmix_NiPrunedFp32Fix

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

Please do not post NSFW or sexual pictures.

By using hollowstrawberry's dataset maker.

Suggested weighting at 0.6-0.9

Trigger word: ichihara yuuko

You may also play with: kimono, pipe, bangs, butterflies, smoke

Lora trained on 1378 images of Yoimiya

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:

yoimiyadef, yoimiyarnd

Trained on NAI, gen'd on AOM2 hard

Leonid Semyonoviсh Kanevsky is a Soviet, Russian and Israeli actor. He became popular with the Soviet audience after starring in The Investigation Is Conducted by ZnaToKi detective series where he appeared as major Tomin.

Model trained on Deliberate
Also fine works on other SD 1.5 based checkpoints
Best results on 0.85 LoRA weight

Edit: Let's call this an Rachnera Cospaly. Not sure if I can figure out how to properly do the spider bits....
I realized that no one had done this yet and I someone who owns a Rachnee mousepad, it bothered me. I left her hair color, eye color, and extra eyes be baked in because it felt right. What is a spider waifu without 6 red eyes to watch you? The classic outfit will need some additional work but it sometimes works.

Notes:

• Strength of 1.0 seems perfectly fine

• Number of extra eyes tends towards 4 but can be more

• The LoRA can omit her spider body but is weighted towards it because, like of course.

• Some this parts of the image look kinda oil paintingy. No idea why

Trigger word: rachnera_arachnera

Helpful words: large_breasts, multiple_legs, claws, taur
Classic(ish) Look: pelvic_curtain, underboob, short_hair, hair_between eyes, black shirt

Tiana is a well-known Disney princess. To create the Lora, a meticulously selected and refined DataSet was used to achieve a versatile and flexible model. So that it adapts perfectly to styles and models. It also works in 3D. Use the keyword “(TianaWaifu:1)” for greater precision, and I also recommend using the keywords “(ponytail), (green dress), (yellow dress, apron), (white dress) and (blue dress, tiara)” to invoke her characteristic outfits. If you want to change her outfit, I recommend lowering the weight of the Lora between “0.8 - 0.5” for greater flexibility, try it out and have fun. Remember to leave a Review with an image if you liked it and comment!! HAVE FUN :D

TIP: With Tiana I can notice that her skin color sometimes becomes lighter when Lora's weight is low, to fix this use the keyword "(very dark skin)

Don’t forget to check out my other Disney Series Loras on my profile. Currently Rapunzel, Snow White, Merida, Cinderella, Pocahontas, Fa Mulan, Elsa, Anna, Belle and Alice in Wonderland!! ;D”

add solo and isabelle_(animal_crossing) tag to make images more accurate

This LoRA was trained with a mixture of white, black, and asian references in an effort to make a mixed race model with a bit of play to it. My best results have come from keeping the CFG around 4, the steps close to 20, and the weight of the LoRA itself around 0.5.

To tweak the weight of each race further (admittedly by default the model leans towards white the most even though that's the race I used the least in training), use the following syntax in your prompt:

"photo of nude ([white::0.XX], [asian::0.XX], [black::0.XX] woman:1.0)"

Replacing "0.XX" with anything between 0.01 and 0.99; the closer to 0.99 means the more steps to which that race will be applied.

(If you are unaware, the [prompt::0.XX] syntax means "prompt" will be applied to XX% of your steps. So, for example, [blonde::0.78] and using 100 steps means the prompt "blonde" would be applied to your gen for the first 78 steps then disregarded for the final 22 steps.)

Tomoki Mikage | Saimin Seishidou 4.5

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：glasses,cleavage,shirt,
角色：<lora:KoharuV1:0.7>,Tachibana Koharu, <lora:jinMTSPSTachibana_tachibanakyoukaLora:0.3>,
头发：blonde_long_hair,green_eyes,
角色鎖定：
1 girl,Solo, Manga,20yo,mature female,Beautiful Finger,Beautiful long legs,Beautiful body,Beautiful Nose,Beautiful character design, perfect eyes, perfect face,
眼神表情：
looking at viewer, (innocent_big_eyes:1.0),(expressionless:1.3),
畫質:
official art,extremely detailed CG unity 8k wallpaper, perfect lighting,Colorful, Bright_Front_face_Lighting,
(masterpiece:1.0),(best_quality:1.0), ultra high res,4K,ultra-detailed,
photography, 8K, HDR, highres, absurdres:1.2, Kodak portra 400, film grain, blurry background, bokeh:1.2, lens flare, (vibrant_color:1.2)
身材:
(Beautiful,large_Breasts:1.4), (beautiful_face:1.5),(narrow_waist),
Negative prompt: EasyNegative, 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: DPM++ SDE Karras, CFG scale: 7, Seed: 3448137216, Size: 512x768, Model hash: 3e9211917c, Model: yesmix_v16Original

Belle is a well-known Disney princess. To create the Lora, a meticulously selected and refined DataSet was used to achieve a versatile and flexible model. So that it adapts perfectly to styles and models. It also works in 3D. Use the keyword “(BelleWaifu:1)” for greater precision, and I also recommend using the keywords “(hair bun), (yellow dress, long skirt, tiara) and (blue dress, apron)” to invoke her characteristic outfits. If you want to change her outfit, I recommend lowering the weight of the Lora between “0.8 - 0.5” for greater flexibility, try it out and have fun. Remember to leave a Review with an image if you liked it and comment!! HAVE FUN :D

Don’t forget to check out my other Disney Series Loras on my profile. Currently Rapunzel, Snow White, Merida, Cinderella, Pocahontas, Fa Mulan, Elsa, Anna and Alice in Wonderland!! ;D”

Petelgeuse Romanée-Conti from anime Re:Zero.

Trained on screencaps from the anime using AnyLora checkpoint.

Lora model of Shuten Douji from Fate Grand Order. Preview images done with weight 0.8-1.

Sadly even after I spent few days on this, it's still work in progress T_T so hopefully I'll update shortly.

I'll describe prompts better when I'm happy with it, so for now just in short:

• Standard minimum outfit - shuten douji, revealing clothes, headpiece, ankle ribbon, barefoot sandals, bare arms

• Standard kimono outfit - shuten douji, revealing clothes, headpiece, ankle ribbon, barefoot sandals, bridal gauntlets, purple kimono, obi, knee pads

• Jiangshi outfit - shuten douji, jiangshi, qing guanmao, long fingernails, china dress, hair rings, jingle bell, flats, sash, frilled sleeves, ofuda

• Halloween caster outfit - shuten douji, dudou, forehead jewel, detached sleeves, earrings, short twintails, fundoshi, shimenawa, black choker, shawl, star print, stirrup legwear, layered skirt, pelvis curtain

• Black kimono - shuten douji, black kimono, hair flower, floral print, choker, facial mark, headpiece, geta, anklet, black scarf, obi, horn ornament

• Maid outfit - shuten douji, maid, maid headdress, black thighhighs, puffy sleeves, garter straps, wrist cuffs, maid apron, high heels, bowtie, detached collar, headpiece, black dress

• Travel outfit - shuten douji, hood, black jacket, open jacket, pink hoodie, sneakers, leggings, black choker, off shoulder, clothes writing, headpiece

• Punk outfit - shuten douji, open jacket, clothes writing, bracelet, collar, wristband, leather jacket, print shirt, fishnet pantyhose, thigh strap, eyewear on head, multicolored hair, black shorts, headpiece, spiked bracelet

The last two, while not closely accurate actually work best for me.

A LORA model trained on Sergio Peinado.

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

Trained on 5490 steps.

https://t.me/stdif_kos

Introduced on 7 November 1996, the fifth generation retained an FF layout with an independent front suspension and 63/37 weight distribution. Most fifth-generation Honda Preludes came with 16-inch (410 mm) aluminum alloy wheels with all-season 205/50 R16 87V tires, featured the 11.1-inch front brakes like the 1996 VTEC model, and most Preludes also received a 5-lug hub (not the 4-lug wheel hub of older models). The Prelude was only available in three models for Canada and two models for the US (the Base and Type SH). All models came with 16-inch alloy wheels and a 200HP (203 PS) 2.2-liter DOHC VTEC 4-cylinder engine. The 2.0i and JDM Si trims came with 195/60 R15 steel wheel, and the JDM Xi came with 14-inch steel wheels. Unlike the North American market Preludes, JDM Preludes came with rear windscreen wipers, except for the Xi. Australian and JDM Preludes weigh less than American and European models: VTi-R manual weighs 1,268 kg (2,795 lb), autos weigh 1,298 kg (2,862 lb), and the ATTS weighs 1,308 kg (2,884 lb). The ATTS model received Honda's Active Torque Transfer System; the equivalent model was called the Type S in Japan, VTi-S in Europe, and Type SH ("Super Handling") in North America.[16] ATTS was designed to counteract the understeer inherent in a front-wheel drive car, but the Prelude's 63.1 percent front weight distribution was too much for the system to successfully mask.[16]

The fifth-generation Prelude marked a return to the more square bodystyle of the third generation (1987–1991), in an attempt to curb slumping sales of the fourth-generation bodystyle. All models and trim packages stayed within the BB-chassis code (BB5-BB9) and housed either an H-series or F-Series engine:

Anju from Majora's Mask, or the unnamed cuckoo lady in Kakariko in OoT. Trained with a very small dataset of only 16 images. I tried resizing the model to keep the size down since its a small dataset, but that had bad results.

Use the tags: "blue skirt, long skirt, brown vest" to get her normal outfit

Trigger word is Anju

Trained on NAI anime, 16 images

double v, single v, v

double v works best, can do multiple girls, a bunch of different poses, can do sfw and nsfw

Trained mainly on ingame images on MeinaHentai model

0.8 and bellow is good

example prompt:DownbreakerDota, glowing hair, yellow eyes, glowing eyes, blonde_hair, colored_skin, grey skin, 1girl, armor, breastplate, full_body, long_hair, breasts, dress, from_below, long_hair, no_panties, pelvic_curtain, solo, thighs, smile, forest

for the weapon try adding: war hammer, holding, holding_weapon, weapon

Tho the weapon tends to be jank, you may try adding those things in the negative prompt to remove them

This model depicts the character Thorfinn from the anime " Vinland Saga".

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

C站没找到小天城模型，2070s8G训练能跑就自给自足了。期待P站能看到更多小天城。

第一次尝试发布炼丹作品。使用Anything4.5大模型训练。

人物Tag：amagi-chan \(azur lane\), purple eyes, fox tail, fox girl, fox ears, brown hair, brown tail

衣服：kimono, japanese clothes

This model was finetuned on top of Realistic Vision 2.0 to depict light reflecting on water. 

After seeing how lighting and water could work together in a previous model, was inspired to do a version above water. Water can be difficult to depict because ripples and waves on the surface change the way light reflects, and capturing this effect can be important for realism. Though the base models can show this to some degree, after experimentation I thought this could be improved.

Around 20 HD images were used for training, selected for high resolution, as well as a variety of lighting conditions (daytime, night, dawn/sunset) and water conditions (still vs. moving water). Used a relatively higher learning rate (5e-3), and a very high contrast noise (0.30) to improve the contrast especially for night scenes. LoRA is 128 / 64 (dim / alpha).

Notably also tried to experiment with the reflections of different things (e.g. people, animals, buildings, nature). Unfortunately people and animals are quite difficult - the real instances are hard to get right, without even considering their reflections. It’s possible that the amount of detail necessary in depicting water somehow affects their realism, more experimentation may tell.

Join our Discord for any questions or feedback!

Prompting

Use “reflections” to trigger the LoRA. Append the recommended positive prompt terms from Realistic Vision. Have not tried the suggested negative prompts, but do make sure to modify them appropriately if not depicting human subjects.

The basic prompt pattern is “reflection of [object] on [body of water]”. You can add a clause to control lighting conditions (e.g. “bright lighting”) or time of day (e.g. “at dusk”) at the end. 

Example prompts:

• Reflections of cliffs over an ocean, at dawn, 8k uhd, dslr, soft lighting, high quality, film grain, Fujifilm XT3 

• Reflections of an iceberg, 8k uhd, dslr, soft lighting, high quality, film grain, Fujifilm XT3  <lora-Reflections-0.8> 2

• Blurry reflection of bridge on water, medium lighting, 8k uhd, dslr, soft lighting, high quality, film grain, Fujifilm XT3 <lora-Reflections-000400-0.8>
  

Recommend using LoRA at around 0.8. 

Settings

Resolution must be 768 x 768 for best results.

Sampler = DPM++ 2M Karras

Sample images were using CFG = 7 and 25-50 inference steps. 

https://t.me/stdif_kos

In January 1998, the 100 series Land Cruiser was introduced to replace the 8-year-old 80 series. The 100 series was previewed in October 1997 as the "Grand Cruiser" at the 32nd Tokyo Motor Show. Development began in 1991 under code name 404T, with the final design being frozen in mid-1994.[44][45]

There are two distinct versions of the 100-series, the 100 and the 105. The two versions look very similar, but there are significant differences under the bodywork. Despite these differences and official model names, both the 100 and 105 are collectively known as the 100 series.

The 105 carried over the majority of its chassis and powertrain from the 80-series with coil suspended solid axles front and rear, and straight-6 petrol and diesel engines. These models were only sold in African, Australian, Russian, and South American markets.

In 1998, a suspension system combining Active Height Control (AHC) and Skyhook TEMS Toyota Electronic Modulated Suspension on the Land Cruiser J100 was introduced.[46]

In 2002, Toyota introduced Night View, the first worldwide series production active automotive night vision system, on the Toyota Land Cruiser Cygnus or Lexus LX470. This system uses the headlight projectors emitting near infrared light aimed like the car's highbeam headlights and a CCD camera then captures that reflected radiation, this signal is then processed by a computer which produces a black-and-white image which is projected on the lower section of the windshield.[47] It was also the first Toyota vehicle with roll-over sensor and control logic.[48]

The 100 models were fitted with a slightly wider chassis,[49] independent front suspension (IFS) and two new engines. The change to IFS was a first for a Land Cruiser, and was made (in combination with rack-and-pinion steering) to improve on-road handling. However it also limited the vehicle's off-road capability and durability, hence the decision to offer the solid axle 105 models alongside the IFS 100 models in some markets was made. The table below identifies the range of 100 and 105 models and their worldwide availability.

Despite the 100 and 105 bodies being very similar, there are some exterior visual indications between both models. The most obvious is the front end of the vehicle often appearing lower than the rear on the 100 models, due to the IFS. The other indicator is the design of the wheels. The 100 models have almost flat wheel designs, while the 105 models have dished wheels. This difference allows both versions to retain similar wheel tracks, despite the 100 having a relatively wider axle track to allow for the IFS system.

The introduction of a V8 engine was also a first for a Land Cruiser, and was specifically intended to improve sales in the North-American market, where it was the only engine available. In Australia, the 100 V8 was initially only available in the range-topping GXV model, while entry and mid-range models were the 105 powered by the 1FZ-FE I6 petrol, or 1HZ diesel engines. The new 1HD-FTE turbo-diesel 100 was added to the Australian range in October 2000 after being available in Europe and the UK since the vehicle's launch in 1998. The automotive press in Australia were critical of Toyota's decision to offer the acclaimed 1HD-FTE engine only in combination with IFS. Australian 4WD Monthly magazine stated "We will never forgive Toyota for going independent at the front with the mighty 4.2 turbo-diesel".

The 100 series formed the basis for the Lexus LX 470, which was also sold in Japan as the Toyota Cygnus.

The 100 series was called the Toyota Land Cruiser Amazon in the UK and Ireland from 1998 to 2007.

In 2000, Toyota celebrated the 50th anniversary of the Land Cruiser with commemorative models offered in several countries. Total global production to date was 3.72 million vehicles.

The 100 series remained in production until late 2007, with several minor facelifts such as headlights, taillights, front grille, rear spoiler and specification changes introduced over the years.

Hidden Hatena | 隠世はてな

https://twitter.com/hatena_ny

Description is in a work in progress

I'll begin working with the thumbnail if I have sufficient enough characters.

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

Seishirou Nagi (4/17/2023)

Weight: 0.7 - 0.8

Trigger + Appearance (Optional): nagi seishirou, white hair

BLUELOCK Uniform: sportswear

Yoichi Isagi (4/17/2023)

Weight: 0.7 - 0.8

Trigger + Apperance: isagi yoichi, black hair

Outfits: He's either wearing a bodysuit or blue sportswear.

Meguru Bachira (4/13/2023)

Weight: 0.8

Trigger: bachira meguru

His short, multicolored hair and yellow eyes are pruned, but feel free to add them anyway.

BLUELOCK Uniform: The uniform will very likely not appear when generating a wider shot, but throw him some sportswear, I'm sure he'll wear it.

v0.1

+ Vengeful Spirit!

- Results can be of low quality without using Hires. fix!

⚠️I would strongly recommend the use of Hires. fix to produce good results. ⚠️

Prompt: vengeful spirit

https://t.me/stdif_kos

Announced on 22 October 1992,[5] the Impreza was released in Japan in November and offered in either front-wheel drive (FWD) or all-wheel drive (AWD) versions and as a four-door sedan or five-door wagon (Touring).[6] The car used a shortened version of the Legacy's floor pan.[7] According to a Motor Trend article written March 1992 on page 26, the name of Subaru's new compact was, initially, to be called the Loyale, displaying an official photograph of the four-door sedan. In late 1995, a two-door coupe was introduced. In Japan it was called Impreza Retna. Initial engine choices included 1.5, 1.6, 1.8, and 2.0-liter naturally aspirated engines.

Subaru chose to continue their longstanding use of the boxer engine in the Impreza. According to Subaru, their configuration of the engine inline with the transmission minimizes body roll due to the lower center of gravity compared with offset engines in most other vehicles. The boxer design provides good vibration mitigation due to the principles of a balanced engine because the movement of each piston is largely countered by a piston in the opposing cylinder bank, eliminating the need for a counter-weighted rotating crankshaft (harmonic balancer), but with some vibration from offsets. Torque steer is also reduced with this type of powertrain layout since the front drive shafts are of equal length and weight.

So....I wanted to try something here as a bit of an experiment.

I grabbed a fairly low quality bunch of images of a poser character (man, if there's something SD will kill off soon enough, it's that). I then did a LoRA using two different versions.

Auto version was me taking the pics, putting them into BIRME for 768x768, having BLIP caption them, and being done.

Manual version was same as above, but I actually manually edited the captions to add a little more intelligence to them. Cuz BLIP is pretty stupid a lot of the time.

Interestingly enough, the Realistic Vision models seem to get you closest to the original poser type aesthetic. Which isn't awesome, cuz Poser was never all that good at making good looking pictures.

So the question is...which one is better?

v1.0

A LoRA based on 1065 Battle Girl High School cards, recognizing 20+ characters to varying degrees.

Undertrained.

Usage

The style itself doesn't have a trigger word. For the characters, below are the character tags and respective top 20 related tags.

Hoshitsuki Miki,
brown hair, short hair, open mouth, smile, looking at viewer, blush, green eyes, 1girl, ahoge, hair ornament, solo, :d, skirt, multiple girls, blue eyes, bow, breasts, jewelry, thighhighs, gloves

Watagi Michelle,
blonde hair, open mouth, 1girl, smile, red eyes, blush, solo, looking at viewer, :d, hair bun, double bun, bow, skirt, short hair, breasts, thighhighs, hair ornament, bangs, long hair, multiple girls

Tokiwa Kurumi,
green eyes, 1girl, long hair, solo, green hair, looking at viewer, breasts, smile, open mouth, blush, hair ornament, cleavage, medium breasts, dress, flower, jewelry, skirt, gloves, brown hair, thighhighs

Narumi Haruka,
blue eyes, 1girl, grey hair, long hair, solo, smile, looking at viewer, blush, open mouth, braid, hair ornament, jewelry, brown hair, gloves, breasts, skirt, earrings, flower, multiple girls, dress

Kusunoki Asuha,
long hair, black hair, 1girl, looking at viewer, solo, blue hair, blush, open mouth, smile, blue eyes, bangs, skirt, green eyes, breasts, hair ornament, jewelry, gloves, flower, bow, aqua eyes

Asahina Kokomi,
blue eyes, brown hair, blush, 1girl, solo, open mouth, bangs, breasts, looking at viewer, smile, braid, medium breasts, cleavage, short hair, skirt, multiple girls, thighhighs, dress, 2girls, jewelry

Sadone,
green hair, red eyes, long hair, 1girl, open mouth, solo, bangs, looking at viewer, blush, twintails, smile, ahoge, bow, hair ornament, multiple girls, blunt bangs, dress, skirt, holding, sparkle

Sendoin Kaede,
blonde hair, long hair, purple eyes, 1girl, blush, solo, looking at viewer, smile, open mouth, drill hair, bangs, bow, dress, hair bow, sparkle, flower, skirt, jewelry, sky, holding

Tsubuzaki Anko,
long hair, 1girl, solo, twintails, brown hair, black hair, open mouth, looking at viewer, blush, breasts, hair ornament, hair between eyes, purple eyes, smile, cleavage, thighhighs, jewelry, skirt, large breasts, food

Serizawa Renge,
1girl, long hair, breasts, smile, solo, mole, looking at viewer, cleavage, blush, mole under eye, open mouth, pink hair, brown hair, pink eyes, jewelry, brown eyes, large breasts, medium breasts, gloves, skirt

Fujimiya Sakura,
brown hair, long hair, 1girl, solo, hair ornament, thick eyebrows, looking at viewer, blush, open mouth, bangs, smile, skirt, blunt bangs, hair bun, flower, single hair bun, bow, red eyes, multiple girls, purple eyes

Hasumi Urara,
red hair, long hair, yellow eyes, blush, smile, 1girl, twintails, looking at viewer, solo, open mouth, ahoge, thighhighs, ribbon, brown eyes, skirt, jewelry, multiple girls, hair ornament, hair ribbon, :d

Minami Hinata,
brown eyes, brown hair, open mouth, short hair, 1girl, smile, fang, solo, looking at viewer, blush, hair ornament, two side up, :d, skirt, black hair, outdoors, thighhighs, day, flower, bow

Wakaba Subaru,
short hair, green eyes, 1girl, open mouth, solo, green hair, smile, looking at viewer, blush, black hair, hair ornament, jewelry, ahoge, skirt, gloves, hairclip, earrings, multiple girls, brown hair, thighhighs

Kougami Kanon,
blonde hair, blue eyes, 1girl, blush, solo, looking at viewer, hair ornament, smile, open mouth, breasts, twintails, long hair, musical note, dress, short hair, gloves, flower, jewelry, musical note hair ornament, medium breasts

Kunieda Shiho,
blue hair, 1girl, blush, smile, breasts, looking at viewer, solo, hair ornament, yellow eyes, brown eyes, open mouth, short hair, large breasts, dress, cleavage, musical note, gloves, skirt, musical note hair ornament, medium breasts

Amano Nozomi,
blue eyes, 1girl, orange hair, long hair, looking at viewer, solo, smile, open mouth, jewelry, blush, bangs, breasts, necklace, holding, one eye closed, skirt, bow, gloves, medium breasts, hair ornament

Himukai Yuri,
red eyes, brown hair, long hair, 1girl, open mouth, solo, looking at viewer, smile, ponytail, blush, skirt, ribbon, :d, hair ribbon, thighhighs, jewelry, gloves, hair ornament, sparkle, multiple girls

Misaki,
brown hair, long hair, 1girl, looking at viewer, blush, green eyes, breasts, solo, open mouth, hair ornament, smile, multiple girls, bangs, very long hair, gloves, skirt, jewelry, dress, medium breasts, cleavage

Sekaide Mari,
brown hair, short hair, yellow eyes, blush, looking at viewer, breasts, smile, open mouth, multiple girls, 1girl, solo, blonde hair, medium breasts, hair ornament, skirt, thighhighs, glasses, blue hair, bangs, holding

Ameya Erika,
pink hair, pink eyes, long hair, smile, breasts, looking at viewer, 1girl, blush, solo, jewelry, open mouth, cleavage, large breasts, skirt, necklace, side ponytail, multiple girls, 2girls, earrings, thighhighs

Mitsurugi Furan,
blonde hair, smile, breasts, 1girl, braid, solo, purple eyes, long hair, blush, looking at viewer, cleavage, open mouth, medium breasts, grin, shorts, sky, sitting, skirt, sparkle, jacket

Yakumo Itsuki,
glasses, blue hair, blue eyes, long hair, red-framed eyewear, blush, open mouth, 1girl, solo, skirt, sitting, hair ornament, looking at viewer, breasts, multiple girls, sky, cloud, day, outdoors, smile

Nanashima Aoi,
brown hair, brown eyes, long hair, open mouth, smile, blush, looking at viewer, bangs, blunt bangs, flower, breasts, :d, multiple girls, 2girls, hair ornament, 1girl, solo, drill hair, skirt, gloves

Kusunoki Asuha C1,
long hair, bangs, blush, black hair, ahoge, open mouth, flower, green eyes, 1girl, solo, sparkle, blunt bangs, hair ornament, multiple girls, 2girls, blue hair, hair flower, shirt, blue eyes, smile

Training info

Training info Dataset:

• 1065 Battle Girl High School cards collected from bg-search (exact duplicates merged)

Tags generated with SmilingWolf/wd-v1-4-convnextv2-tagger-v2 with a threshold of 0.35. No balancing and tag cleaning.

Inspired by a 2022 paper, a custom LR scheduler was used:

SequentialLR(optimizer, [
    LinearLR(optimizer, 0.1, total_iters=steps_per_epoch),
    CosineAnnealingWarmRestarts(optimizer, steps_per_epoch*2, T_mult=2)
], [steps_per_epoch])

Training cost: ~4 T4-hours

Data distribution:

            character  cards
1     Hoshitsuki Miki     77
2     Watagi Michelle     58
3       Tokiwa Kurumi     58
4       Narumi Haruka     56
5      Kusunoki Asuha     56
6      Asahina Kokomi     54
7              Sadone     54
8       Sendoin Kaede     53
9      Tsubuzaki Anko     53
10     Serizawa Renge     53
11    Fujimiya Sakura     52
12       Hasumi Urara     52
13      Minami Hinata     50
14      Wakaba Subaru     50
15      Kougami Kanon     49
16      Kunieda Shiho     49
17       Amano Nozomi     49
18       Himukai Yuri     46
19             Misaki     26
20       Sekaide Mari     18
21        Ameya Erika     17
22    Mitsurugi Furan     11
23      Yakumo Itsuki     10
24      Nanashima Aoi      9
25  Kusunoki Asuha C1      5

Training config

[model_arguments]
v2 = false
v_parameterization = false
pretrained_model_name_or_path = "animefull-final-pruned.ckpt"

[additional_network_arguments]
no_metadata = false
network_module = "lycoris.kohya"
network_dim = 16
network_alpha = 1
network_args = [ "conv_dim=0", "algo=loha",]
network_train_unet_only = false
network_train_text_encoder_only = false
cache_latents = false

[optimizer_arguments]
optimizer_type = "AdamW8bit"
learning_rate = 0.001
max_grad_norm = 1.0
lr_scheduler = "custom"
#lr_warmup_steps = 100

[dataset_arguments]
debug_dataset = false
# resolution: 576x704

[training_arguments]
save_precision = "fp16"
save_every_n_epochs = 1
max_token_length = 225
mem_eff_attn = false
xformers = true
max_train_epochs = 31
max_data_loader_n_workers = 0
persistent_data_loader_workers = false
gradient_checkpointing = false
gradient_accumulation_steps = 1
mixed_precision = "fp16"
clip_skip = 2
lowram = true

[sample_prompt_arguments]

[saving_arguments]
save_model_as = "safetensors"

Other Final Fantasy VII Remake version:

Aerith_meenow

Realistic_Aerith_meenow

About tifa_ v2

This training increased the training set, optimized the tag, and Tifa became more "flexible" and easier to use with other Lora.

1. The recommended weight is around 0.75

2. This version of tifa is easy to be naked. Please adjust the tag appropriately.

关于tifa_v2

此次训练加大了训练集，优化了Tag，Tifa也变得更加"灵活"，更容易和其他Lora搭配使用。

1.建议权重 0.75附近

2.这个版本tifa比较容易裸体，请适当调整tag。

About v1:

If you also like Final Fantasy VII Remake

I believe this model will not disappoint you.

This is a Lora model about Tifa. It adopts the method of mixed training of real people and Tifa games. There are still many imperfections as a version for the time being. At the same time, I hope you can give full play to your creativity and provide further ideas for my creation.

Model related issues:

1. It is recommended to use Chilout Mix output

3. The triggering weight of Lora model is 0.9, and the effect is good

3. The model is still powerless to generate images of huge model size

4. The model is mainly good at facial modeling, and may match better with other models in terms of movement?

Thank you very much for preferring this model=)

关于V1：

如果你也很喜欢Final Fantasy VII Remake

相信这个模型不会另你失望。

这是一个关于Tifa的Lora模型，采用真人和Tifa游戏混合训练的方法，暂时作为一版本还有很多未完善的，同时我也很希望大家发挥自己创造力，提供我创作的进一步思路。

模型相关问题：

1.推荐使用ChilloutMix 输出

3.Lora模型触发权重0.9效果附近较好

3.模型对于生成巨大型号尺寸的图片暂时还很无能为力

4.模型主要擅长面部塑造，动作方面可能和其他模型搭配起来更好？

最后非常感谢大家喜欢这个模型 =)

Deman sleyar fr fr no cap

Aipii あいぴー

https://twitter.com/aipii_lp

Sio... kiss me~cupid :3

Trigger word : sio

V2

here : -> Sio cupid hihi <-

Use this wisely. I'm not responsible for anything you create with this LoRA.

icon

work with Chilloutmix, can generate natural, cute, Vietnamese girls.

Recommend: 0.4-0.7

I hated the side-view ones that were on civitai, so I made my own.

does what it says.

girl on knees, sucking weenis, you get a side-view shot, and she looks at the camera.

All of the images came from clarity2 (checkpoint), it's great at drawing caucasian women, but.. meh at others.

best prompts is: sucking a big black cock

If the face looks like it was beat with the ugly stick, go grab the vae, and make sure the image is 512x512. 30 to 60 steps, cfg from 7 to 8.5. If you want it to look better, just resize x2, and then do the 4x upscaler.

Ryusei Shido (士道 龍聖) | Blue Lock (LORA)

Ahoy me crews!

It's been awhile since I posted another LORA and well to commemorate our 6000 followers on PIXIV, I am giving this LORA for free. You can use it to make beautiful and awesome Blue Lock fanart with stable diffusion! Thank you so much for my crews and Boatswains for your love and support ^u^)/ Hopefully, we can make great LORA and artworks in the future for all people to enjoy

This Lora was trained with anime-final-pruned and 20 Epochs. This LORA is very flexible and can be used on realistic model too as you can see on my Pixiv.

乗組員の皆さん！

久しぶりにLORAを投稿しましたが、PIXIVフォロワー6000人突破を記念してこのLORAを無料で差し上げます。これを使用して、安定した拡散で美しく素晴らしいブルーロックのファンアートを作成できます!私の乗組員とボースウェインの愛とサポートに感謝します ^u^)/ 将来、すべての人が楽しめる素晴らしい LORA とアートワークを作成できることを願っています。

この Lora は、anime-final-pruned と 20 エポックでトレーニングされました。この LORA は非常に柔軟で、私の Pixiv で見られるように、現実的なモデルにも使用できます。

DITA IF YOU READ THIS: NOW YOU CAN MAKE MORE YAOI \(￣︶￣*\))

Here are the trigger words of each costumes:

各衣装のトリガーワードは以下の通りです。

1) Soccer Uniform >>> Shido, a man in a ZBK_uniform standing in a hallway, blonde hair with pink highlights, brown skin, pink eyes

2) Shirtless >>> Shido, a shirtless man sitting, blonde hair with pink highlights, brown skin, pink eyes, abs

3) Black shirt >>> Shido, a blonde haired man with pink eyes and blonde hair with pink highlights, brown skin, black long sleeves shirt

A commission by @deciweb290 to bring her back.

Kristanna Loken as the T-X in Terminator 3 is an iconic villain in cinematic history. I tried to capture her appearance as much as possible, including her clothing and accessories. Generally, the best results I've had are with a LORA weight of 0.75 to 0.85. Also, technically the LORA doesn't require a Triggerword, but I've had rare odd results without it, so I always prompt with KristannaTX to make sure.

This is the first LORA I've published, which was an interesting learning process for me. I will probably keep embeddings as my main focus, but more LORA's are likely in the future as I enjoyed the workflow, the fast training speed and the results from this project.

This LORA has been trained using SD1.5 as it's base and it's been tested to work well on most models that are capable of making realistic looking images.

If you enjoy my creations, consider supporting me on Ko-Fi ^^

LoRA extractions is not as good as the full model but it creates a compromise between size and quality and it can be used along other LoRA model with a full model. Credit goes the creator of the checkpoint in version description

Please provide request link(s) in my comments "Suggestions for a model extraction?"

04.0417.23 Name version is limited to 10 25 character, make it impossible to see full name... see below "List of all LoRA model".

How to use it in AUTOMATIC1111 WebUI?

Place the LoRA model in .\stable-diffusion-webui\models\Lora then use <lora:filename:multiplier> in the prompt.

About the comparison:

• Picture order: First = LoRA, Second = Fine-tuned Model, Third = Base Model

• DM = Network Dimension ( 120=lower quality/low size, 320=higher quality/higher size ) same for CD = Conv Dimension (Rank)
  [CD is optionally if you wish to use it, install the extension: https://github.com/KohakuBlueleaf/a1111-sd-webui-locon (How? Extensions -> Install from URL)]
  Obviously Version "1024CV1024" or in short "1024" use more VRAM
  I did a large comparison DM+CV here

• sample pictures does not use vea and LoRA model with SDv1.4

• every model from (Stable Diffusion Base Model: mostly sd-v1-4.ckpt, model base is determined by highest % affinity) was extracted using Kohya_ss (Save precision: fp16)

• Name format: <Type><DM> <CV><Name>

List of all LoRA model:

• LORA1024CV1024flat2DAnimerge_v10 - LORA320flat2DAnimerge_v10

• LORA1024CV1024realismEngine_v10 - LORA320realismEngine_v10

• LORA1024CV1024vectorArt_pulpVectorBeta - LORA320vectorArt_pulpVectorBeta

• LORA1024CV1024hasdx_hasedsdx - LORA320hasdx_hasedsdx

• LORA1024CV1024verisimilitude_V14 - LORA320verisimilitude_V14

• LORA1024CV1024rpg_V4 - LORA320rpg_V4

• LORA1024CV1024galaxytimemachinesForyou_v10 - LORA320galaxytimemachinesForyou_v10

• LORA1024CV1024galaxytimemachinesGTM_v3Pruned - LORA320galaxytimemachinesGTM_v3Pruned

• LORA1024CV1024hassakuv1m

• LyCORIS1024CV1024illuminatiDiffusionV1_v11 - LORA320illuminatiDiffusionV1_v11

• LORA1024CV1024xrikishisUnstable_alpha50 - LORA320xrikishisUnstable_alpha50

• LORA1024CV1024diivesStyle_10 - LORA320diivesStyle_10

• LORA1024CV1024elldrethsImagination_v10 - LORA320elldrethsImagination_v10

• LORA1024CV1024digitalVividMemoriesAnime_10 - LORA320digitalVividMemoriesAnime_10

• LORA1024CV1024generalPurpose - LORA320generalPurpose

• LORA1024CV1024dreamlikeDiffusion10_10 - LORA320dreamlikeDiffusion10_10

• LORA1024CV1024epicDiffusion11 - LORA320epicDiffusion11

• LORA1024CV1024protogenInfinityX86 - LORA320protogenInfinityX86

• LORA1024CV1024protogenNovaExperimentalX80 - LORA320protogenNovaExperimentalX80

• LORA1024CV1024protogenX58RebuiltScifi - LORA320protogenX58RebuiltScifi

• LORA1024CV1024modernDisney_v1 - LORA320modernDisney_v1

• LORA1024CV1024revAnimated_v11 - LORA320revAnimated_v11

• LORA1024CV1024revAnimated_v122 - LORA320revAnimated_v122

• LORA1024CV1024dosmix - LORA320dosmix

• LORA1024CV1024icomixV03 - LORA320icomixV03

• LORA1024CV1024meinaV8 - LORA320meinaV8

• LORA1024CV1024comicBabes_v1 - LORA320comicBabes_v1

• LORA320CD128perfectWorld_v2Baked - LORA120CD128perfectWorld_v2Baked -LORA1024CD1024perfectWorld_v2Baked

• LORA14DM320realisticVisionV20_v20 - LORA14DM120realisticVisionV20_v20 - LORA1024CV1024dreamshaper_4NoVae

• LORADM320realisticVisionV13_v13 - LORADM120realisticVisionV13_v13

• LoRADM320dreamshaper_4NoVae - LoRADM120dreamshaper_4NoVae

• LORA14DM320grapefruitv41 -LORA14DM120grapefruitv41

• LORA1024CV1024berrygold11 - LORA14DM120BerryGold-1-1 - LORA14DM320BerryGold-1-1

• LORA14DM120chilloutmix_NiPrunedFp32Fix - LORA14DM320chilloutmix_NiPrunedFp32Fix - LORA1024CV1024chilloutmix_NiPrunedFp32Fix

• LORA14DM120deliberate_v2 - LORA14DM320deliberate_v2 - LORA1024CV1024deliberate_v2

• LORA14DM120HSPU-1.1 - LORA14DM320HSPU-1.1 - LORA1024CV1024hspu_11

• LORA14DM120HSPUCar1-1 - LORA14DM320HSPUCar1-1 - LORA1024CV1024hspuCarV10

• LORA14DM120Maiden - LORA14DM320Maiden - LORA1024CV1024maiden

• LORA1024CV1024yiffye18Furry4Kinky

• LORA14DM320protogenX53Photorealism_10 - LORA14DM120protogenX53Photorealism_10

• LORA14DM320anything-v4.5-vae-swapped - LORA14DM120anything-v4.5-vae-swapped - LORA1024CV1024anything-v4.5-vae-swapped

• LORA14DM320robo-diffusion-v1 - LORA14DM120robo-diffusion-v1 - LORA1024CV1024robo-diffusion-v1

• LORA14DM320Silicon29 - LORA14DM120Silicon29 - LORA1024CV1024Silicon29

Updated 1.1 version:

Trigger word "ElysiaHoH"

Replace "ElysiaHoH" with other cloth tags to get other custom. Tested:
"chinese dress"

"bikini"

"jacket, jeans"

"nsfw, naked"

触发词: ElysiaHoH。

使用触发词可以获得人之律者服装。把触发词替换成其他衣服可以换装，已测试过的有：

"nsfw, naked"

"chinese dress"

"bikini"

"jacket, jeans"

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。

lora强度推荐0.6-0.7。chilloutmix出图，词条看示意图，负面提示词我用了EasyNegative.pt：EasyNegative | Stable Diffusion TextualInversion | Civitai

素材来自网络，只有面部。

请及时下载保存，真人模型随时可能被删除。

喜欢请给五星。

公共频道：t.me/dawn66666666

Here is the Lora for lick my feet !

Here are the trigger words suggested, foot, feet, foot_focus

Hope you enjoy my work!

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

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

Basic model:AnyLora (!!!Vae Not Nedeed!!!)

Images were taken here:https://vk.com/m3rii44

don't forget to send in your results

P.S. I opened a discord server where you can offer your ideas for models, link:https://discord.gg/HDfvejBPMJ

Discount for commission (10 slots 5 days delivery) and some exclusive LoRA, Visit my BMC page

(EXPERIMENTAL .Ver)

This Lora contains NSFW , please use it wisely!!
WARNING! "NSFW prompt" are easily triggered in this version |

It's perfect for the "Chillout-mix" checkpoint, besides that I haven't tested it yet

I recommend using weights 1 or slightly below (0.750 - 0.950), 30-35 Sampling steps, uncheck restore faces for better results!

V2.0 Update

I learned a lot in this month, and my graphic card also suffered a lot :P, till now I can publish a better version for this model

For this version, recommend weight is 0.45~0.7, only one trigger word nahida \(genshin impact\) is needed, but has better restoration of characteristic

My suggestion is use lower weight for model, and higher weight for trigger word, so you can easily change her suit

All known problems in previous version are solved

The next version will be a LyCORIS model, please stay tuned!

Any suggestions are welcome!

V1.0 Release

Genshin Impact Nahida LoRA, training with base model final-prune[925997e9].

Recommend weight 0.45~0.8. around 0.7 is a sweet spot for customize and keep characteristic.

Models used for examples:

• Cetus-Mix

• MeinaMix

• AOAOKO [PVC Style Model] with VAE Counterfeit-V2.5

Use tags braid, sidelocks, side ponytail, multicolored hair, white hair, green hair, gradient hair to fix hair style.

Use tag bracelet to generate a in-game style bracelet (working better in higher weight).

Use tag white bloomer to add in-game style bloomer for charater.

Known problem:

• Lower weight may cause a little incorrect on the green part of the hair color.

• Sometimes a unexpected crown would appear on Nahida's head (mostly in higher weight).

Any suggestions are welcome!

Fiora from Xenoblade Chronicles. It might be a little bit overcooked, so lowering the weight to 0.8 seems to work better. Includes her Homs version, and her Mechon version. The Mechon version was a much smaller sample size, so its not as good as the Homs.

If you want to generate her normal outfit:

fiorahoms, tube top, detached sleeves, skirt

put in the negative: cyborg, short hair (unless you want her short haired version from post game)

To generate her Mechon version:

fioramechon, cyborg,

put in the negative: bare shoulders,

Trained on NAI anime, 55 homs images, 30 mechon

M. V. Lomonosov Moscow State University (MSU; Russian: Московский государственный университет имени М. В. Ломоносова, romanized: Moskovskiy gosudarstvennyy universitet imeni M. V. Lomonosova) is a public research university in Moscow, Russia, and the most prestigious university in the country.

The university includes 15 research institutes, 43 faculties, more than 300 departments, and six branches (including five foreign ones in the Commonwealth of Independent States countries). Alumni of the university include past leaders of the Soviet Union and other governments. As of 2019, 13 Nobel laureates, six Fields Medal winners, and one Turing Award winner had been affiliated with the university.

The university was ranked 18th by The Three University Missions Ranking in 2022, and 76th by the QS World University Rankings in 2022, #293 in the world by the global Times Higher World University Rankings, and #326 by U.S. News & World Report in 2022. It was the highest-ranking Russian educational institution by QS in 2020, and according to the Nature Index in 2019 the highest ranked Russian university for research output. Moscow State University is generally accepted as the leading higher educational institution in the former Soviet Union.[8]

This is my first time training a model. It may not be very successful, but there is no way, just this ability.
It is best to use the CHILLOUTMIX model. When using other models, the effect may not be very good.

If only use it to generate a face, the effect will be the best. To generate a full upper body, a resolution of 768*768 or more is required. The effect of generating the whole body is very poor. 

Alisabelle is a prefessional model from Met-Art. Trained on RealisticVisionV2. Best results found from weights 0.7 to 1.0.

快回来，村里发魅魔了

推荐权重：0.3-0.6（越大越不真实）

Hi-res时适当提高，重绘幅度建议小一点

效果问题：realistic类咒语会明显削弱翅膀和尾巴，小幅削弱淫纹，可以自行补充wings、tattoo等关键词或加描述

水印问题：恶魔外形感觉和同人网站水印有强关联，出现概率贼高，但是去水印负面加的太高会影响图片质量甚至削到淫纹头上，建议还是允许出现水印然后修图或裁切

关键词及世界观设定解读（絮絮叨叨）

1girl：隐藏形态，但角经常会幻化为兽耳或发髻，而且腰比普通瘦的女性会细20%，疑似肋骨和内脏比人类更少或更小

Village_Succubus：恶魔特征的角、翼、尾，眼睛有时会变色，四肢有时自带长筒手套、长筒袜的纹路，或者鳞片、咒文类的纹路

obscenity_tattoo：魅魔魔力的显现，特点为爱心形状的淫纹，魅魔魔力越强，爱心越多越大越精致，有的魅魔不喜欢显露淫纹，有的魅魔反之只显露淫纹不张开翅膀

darkness_tattoo：遇到抖M忍不住黑化的魅魔形态，黑皮白毛+夜光纹身，战斗力和“战斗力”都更强

（搞不定翅膀和尾巴，明明训练都是双翼、负面也试着排除single_wing，就是乱七八糟，偶尔自动变成head_wings但还算能看，侧身和背后的位置识别也是离谱，估摸着要专门训练或者我有空去学习一下controlnet试试……

最难的尾巴正面我都没让他稳定下来，偶尔还会和淫纹混淆，只能暂时放弃了，我还是很想要爱心尾巴的。

淫纹还算成功，真人模型用各种tattoo跑出来的感觉太丑了，长个鬼脸兽头看着不碍眼吗，所以也专门练了一下爱心tattoo）

生成Rwby 主角几人团的全身,头像等

Discount for commission (10 slots 5 days delivery) and some exclusive LoRA, Visit my BMC page

(EXPERIMENTAL .Ver)

This Lora contains NSFW , please use it wisely!!
WARNING! "NSFW prompt" are easily triggered in this version |

It's perfect for the "Chillout-mix" checkpoint, besides that I haven't tested it yet

I recommend using weights 1 or slightly below (0.750 - 0.950), 30-35 Sampling steps, uncheck restore faces for better results!

Warning: Lora with too many degrees of freedom: First of all this is a mixed version and not only of characters, among everything done by the artist, so the results of his creations can be quite random and must generate several attempts and necessarily use the tags.
Since the version allows many degrees of liberation and the author creates surreal things, then the results are going to look broken unless they are wasted on using the tags correctly. Play with the tags

macoco,<lora:pon-10:1> --->Activacion

Top 50 tags:

1girl (689)

solo (599)

black hair (480)

ass (414)

long hair (412)

breasts (404)

nude (372)

sweat (331)

blush (308)

large breasts (302)

pussy (292)

anus (281)

censored (247)

tentacles (204)

from behind (202)

brown hair (196)

short hair (194)

simple background (191)

anal (185)

penis (182)

hetero (164)

open mouth (163)

1boy (156)

skirt (145)

spread legs (140)

thighhighs (135)

cum (134)

nipples (127)

slime (substance) (125)

earrings (125)

jewelry (123)

what (121)

uncensored (117)

black eyes (116)

gaping (115)

brown eyes (112)

object insertion (112)

heart (108)

shirt (107)

vaginal (106)

dark skin (104)

monster (103)

looking back (101)

monochrome (101)

looking at viewer (96)

white background (95)

sex toy (93)

panties (92)

barefoot (90)

underwear (89)

猎人×猎人角色奇犽·揍敌客lora

使用tag

Ranamon,monster girl,helmet,(hairless),green skin,red eyes,gloves,blue leotard,fins,dutch angle,clothing cutout,side cutout,anklet,digimon

基本算是一个实验作品了，主要也是帮助上一个评论区的留言练了这个模型，以后练这类图片资源较少的角色应该慎重考虑一下了，困难度并不是我这个水平能处理的。而且炸了几次之后这一炉也只是相对好一点，模型的问题还比较大，主要问题还是出在了手部上，AI识别不了那种特异形状的手，这个我也没有解决办法，等有能力的作者再弄这个角色吧

It's basically an experimental work. I helped the comments in the previous comment section practice this model. In the future, I should carefully consider characters with fewer image resources. The difficulty is not something I can handle at my level, and after a few explosions, this batch is only a bit better. The model still has serious problems, The main problem is still with the hands. AI cannot recognize those hands with special shapes, and I don't have a solution to this, Wait for a capable author to play this role again

权重（weigh）0.8~1，相关性（cfg）在6左右，一般5就行

不懂为什么，弄完之后感觉自己好像练得是洛克人（）

I don't understand why. After finishing it, I feel like I'm practicing as a Rockman

这是一个用于生成胶片风格AI写真照片的LoRA。与ChilloutMix搭配使用，可以生成逼真的胶片风格照片。因为用于训练的图像均为横向构图，因此生成横图或者方图的效果比较好。

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，因为我用这个模型生成的图像不算多，大家可以摸索摸索多试试。

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。

by 豹仙人Leopa-sama

Wang Xiaoyu is a former member of the badminton national youth team. She has been selected for the national youth team twice. After retiring, she started to operate a short video account and became the "No. 1 large" short video platform for badminton⁴. Unfortunately, I could not find more information about her badminton career.

Statement

1. This Lora model is prohibited from generating NSFW images.

2. This Lora model is prohibited from commercial use.

Usage (用法):

Cameltoe pussy, <any pose, any style, any person you like>

Recomment weights(推荐权重): 0.7~0.9

This is pretty strong Lora, try multiple setting to see which one you like most.

Example: 示例

cameltoe pussy, naked, 20 years old young white woman, pink bra

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: 263121448, Face restoration: CodeFormer, Size: 512x768, Model hash: fc2511737a, Model: chilloutmix_NiPrunedFp32Fix

Recommend model(推荐模型):

https://civitai.com/models/6424/chilloutmix

Impacts of the weights

Please use this LoRA to give birth to good-looking Aries warriors, and please share your works for everyone to appreciate, thank you

v1.0 2023.04.17

Recommend options

• LoRA weight 0.6~0.8

• Trigger words

  • base

    • hikami smire

  • Common recommended keywords

    • hair ribbon,purple hair,blunt bangs,

Can change outfits

• starlight academy uniform

　starlight academy uniform,white skirt,blue jacket,pleated skirt,red ribbon

• code 1 (Difficult to output)

  bow, checkered, checkered background, checkered skirt, gloves, hair ribbon, hammer and sickle, idol, long hair, meer campbell, open mouth, pentagram, plaid, plaid background, plaid bow, plaid dress, plaid skirt, plaid vest, ribbon,short sleeves, skirt, star-shaped pupils, star choker, star earrings, code 1

• code 2

  crown,jacket,black jacket,knee boots,black purple skirt,white legwear,choker, black footwear, boots, choker, dress, flower, jewelry, thigh strap, code 2

• code 3

  code 3, apple, bangs, berry, blueberry, bow, camellia, candy apple, cherry, cowboy shot, dress, earrings, food, food-themed hair ornament, food on head, fruit, fruit on head, grapes, holding, holding food, holding fruit, jewelry, long hair, looking at viewer, mandarin orange, orange \\\(fruit\\\), orange slice, peach, red apple, red flower, red rose, short sleeves,strawberry, strawberry hair ornament,watermelon, whipped cream, wrist cuffs,

• The output is not perfect, but it is quite similar.

Use animefull based model

• Settings

  • Stable Diffusion WebUI by Automatic1111

  • DPM++ Series(SDE Karras, 2M Karras, etc.)

  • about 20 steps, CFG scale 3.5~6.5

  • kl-f8-anime2.vae

  • CLIP skip = 2

LoRA training info

• trained on sd-scripts by kohya_ss. Thank you a lot!

• 107 images of hikami smire

• resolution : 512x512rank=alpha=16

• batch size = 4, 10 epochs trained

This LoRA may be compatible with animefull-based models (ex. AbyssOrangeMix Series...)

This LoRA can generate nsfw images of hikami smire

氷上スミレ（アイカツ！）

• 設定などは上記を参考

There are 3 versions for different outfits! I separate them to make more consistent generation outputs.

To replace the preset outfit with others (e.g. school uniform), please use the black kimono ver.

Highres fix is strongly recommended for all 3 versions and I'm using ESRGAN_4x with denoise strength=0.5 for these samples.

white coat purple kimono ver.

LORA for character with anime original white coat and purple kimono. It's less consistent (on coloring part) comparing to previous 2 outfit.

Trigger words: 1girl lacus clyne pink hair blue eyes

Outfit: ponytail open clothes vest uniform long sleeves purple sleeves purple kimono japanese clothes

white and purple skirt ver.

LORA for character with anime original white and purple skirt. I found it more consistent when the LORA is focusing on one single outfit, but it still can not reproduce the outfit 100% correct, especially on the lower part of the dress😢.

Trigger words: 1girl lacus clyne pink hair blue eyes

Outfit: straight hair white sleeves detached sleeves purple skirt purple frills

black kimono ver.

A requested character LORA for Lacus Clyne from Mobile Suit Gundam SEED! Suggested weights: 1

Character trigger words: 1girl lacus clyne blue eyes pink hair

Hair style: straight hair or ponytail

Outfit triggers (less consistent in realistic models): japanese clothes black kimono detached sleeves

This is a character LoRA for Conan.

v1：The base model used is the original stable-diffusion-1.5, training with 2500 steps on nearly 30 images.

v2：The base model used is https://civitai.com/models/4855/anythingelse-v4, training with 2500 steps on nearly 30 images.

The model is in the original Ghibli style and can be used with other Loras in the generative animation style which can create fantastic scenes and buildings。

Download counterfeit-v2.5 model before use

v2.0 The light and shadow effect of the structure is optimized to be more friendly to the architectural expression. Please adjust Tag to avoid unreasonable building structure

v3.0 I improved the light, shadows and scenes, added a lot of detail, and tried to get away from the original style, which is more animation oriented. If you don't like this style, please continue to use v2.0, of course, if you want to try new style and details, you can try it

v4.0 Added more architectural style and more friendly for close-up shots. More details, of course, if you like the style of the previous three versions, please download free

Q群 692574838

If you have any questions, please contact me by email

This model is not for commercial use, only for exchange learning

一直找不到很好的关于不知火舞的lora模型，所以自己就找了很多图片训练了一个。

本来是打算自己用的，想着跟别人拿了这么多，也该贡献一下。

触发词：maiShiranui，maiShiranui_face,maiShiranui_cloth

权重0.5左右比较好，二次元，2.5D和真实模型我都试用过，有些效果很好，有些一般。

I haven't been able to find a good lora model about Mai Shiranui, so I found a lot of pictures and trained one.

I originally planned to use it myself, thinking that I have taken so much from others, so I should contribute.

Trigger words: maiShiranui, maiShiranui_face, maiShiranui_cloth

A weight of about 0.5 is better. I have tried the two-dimensional, 2.5D and real models, and some of them are very effective, while others are not so good.

Demo images use Danger Zoo ~ Johnsons Negative TI Collection ~

"modelname style" is the main trigger word
additional trigger words are listed to keep you close to the trained style/concepts.

This LORA is designed for used with other djz LORA's and/or on top of Mixed pairs of djz CKPT's or a Triple merge mix including other CKPT models. This is a specific method for merged CKPT models while running a stack of these Style LORA models on top. You may still use this alone and it is recommended to include the additional trigger words in your prompts.

DJZ Season 2.5

V2.0新增皮肤（V2.0 add skin update），校服和婚纱（slow_ahead!、blissful_purity）

唤醒词（Trigger Words）

原版（origen）:javelin\(azur lane\),\(origen\),1girl,

校服（slow_ahead!）:javelin\(azur lane\), \(slow_ahead!\), school uniform,

婚纱（blissful_purity）:javelin\(azur lane\), \(blissful_purity\), wedding dress,

可选提示词（Optional prompt words）

原版（origen）:solo, bracelet,white gloves, purple_hair, mini_crown,white_camisole, hair_ornament, black_ribbon, hair_ribbon,blue hood,

校服（slow_ahead!）:1girl, solo, purple_hair,mini_crown, ribbon, hair_ornament,blue shirt, jewelry,

婚纱（blissful_purity）: 1girl, solo,purple_hair,gloves, jewelry,

V1.0

可选提示词：bracelet,white gloves, purple_hair, mini_crown,white_camisole, hair_ornament, black_ribbon, hair_ribbon,blue hood,

another art style lora. this time i used more base arts and training steps than the previous ones.

still need more testing. might need to tune down weight when use with other loras.

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

support the artist:

https://twitter.com/sakula_yuki

https://www.pixiv.net/en/users/1661253

Trained in Anything V4.5 but should work fine in any anime model.

Lora weight: 0.6 ~ 0.85

Upscaler/settings I use

Highres: 4x animeSharp

Hires steps: 20
Denoising strength .55 ~ .75

Megami No Cafe Terrace anime full LoRa:

trigged words:

riho tsukishima

shiragiku ono

ouka makuzawa

ami tsuruga

akane hououji

hayato kasukabe

background

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

if you get bad results - try https://civitai.com/models/4669/corneos-7th-heaven-mix this checkpoint and write in prompt position(like "front view, face"), because dataset not only front-face.

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

https://plisio.net/donate/pm9qVYci

Works best at 0.9-0.7

BlindBox Artstyle

更新了V3版本，调整了眼睛，俗话说眼睛是心灵的窗口。但是目前还是实验性质的，从稳定性上还是推荐v1_mix版本。

如果你喜欢我的作品，可以点个Like，或者请我 喝杯咖啡。

爱来自瓷器。

[机翻英语]

Updated v3 version, adjusted the eyes, as the saying goes, the eyes are the windows to the soul. But it is still experimental, from the stability or recommended v1_mix version.

If you like my work, you can click "Like" or buy me a coffee.

Love from China.

blindbox_v1_mix 👍

• Add the prompt full body,chibi,

• weigh : 1

• checkpoint : RevAnimated or other you like

mix PVC material, is good.

🖖I would appreciate it if you could share with me the results of your use.

blindbox v1

• Add the prompt full body,chibi,

• Recommeneded weigh : 1

• checkpoint : RevAnimated

Other checkpoint may produce fantastic results, try it yourself.

Can be used with other LoRA.

blindbox v2

• Reduces the probability of a big head (maybe)

• Lowered network_dim value, smaller file size

• More Chibi, Less blindbox

v1 version has good blindbox artstyle

blindbox v3

• 调整眼部细节

ーーーーーーーーーーーーーーーーーーーーーーーーーーーーーーーー

盲盒风格的LoRA，一开始是这么想的。

v1版本 还是比较盲盒的。

增加了PVC的质感，更新了v1_mix，效果很不错，建议使用👍

增加关键词chibi, full body,

减少负面描述会好一点，也可自行尝试。

经过测试，如果要和别的LoRA一起使用，还是把那些负面关键词加上吧😅

推荐使用RevAnimated模型。

跟一些角色LoRA也有比较好的相性。

v2版本减少了大头的概率，更偏向二头身。

适合一些动漫画风了，也许。

有问题可以去B站问，如果我能解决的话😄

姑且放上一个没人气的视频

https://www.bilibili.com/video/BV1Ah411G7Zs/

model : ironcatfatetoon_v10

Sampling method : Euler a

suggest lora set 0.8-1.0 and add "Kozuki Hiyori" in prompt

Have fun!

This model is prepared for 《名侦探柯南：黑铁的鱼影》（日语：名探偵コナン黒鉄の 魚影 （ サブマリン ） ）Détective Conan : Le Sous-marin noir.

This LoRA supports to generate images of both Conan and Haibara. The recommended basemodel is https://civitai.com/models/4855/anythingelse-v4.

Tsurumaki Kokoro, Character of BangDream!

All I can say is she so cute!

Happy Lucky Smile!

tested models:

-Anything V4/V3

-AbyssOrangeMix2 - SFW

note: I mostly used AbyssOrangeMix2 - SFW for my previews

also, if you have the problems of duplicate pupils in one eyes, put "double pupils in one eye" in the negative. character with dots eyes are more difficult to train.

I've noticed that many creators tend to delay uploading their models, waiting for download counters and likes to be fixed. However, I don't wait (in fact, I upload even more now). I'm not chasing the numbers; I'm simply delivering my best work to you all. Enjoy!

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

Also, don't hesitate to send me suggestions. I tend to alternate between creating what I want and what I'm asked for. Rest assured, nothing goes unnoticed.

斗破苍穹云韵2.0

有使用问题可以在 b站视频 内留言。

带 prompt 图片可以在公众号【老梵AI】后台回复【云韵模型2.0】获取

本人所有图片和模型均免费公开，如果有人向你收费，可以举报他。

1. 主模型：chilloutmix_Ni

2. 依赖 Lora：koreanDollLikeness_v15

3. 其他参数可参考图片 demo

4. 建议权重：

   1. 云韵：0.5-0.6

   2. koreanDoll：0.5-0.6

      

Lelei la lalena from the manga/anime: Gate: Jieitai Kanochi nite, Kaku Tatakaeri.
Trained on NovelAI with resolution of 768x1024 and on 40 images screenshotted from the anime.

The staff is especially hard to get perfect. Use weight between 0.6-0.9 for best results.

That model was trained on 13 pictures using AOM3

good

一种艺术风格，看到封面图就会明白它的用途

训练所使用的模型是Anything v4.5，经过测试在其他动漫模型上也能取得较好的效果，但对于上色较平的模型(如color box)只能保持脸部特征，皮肤的着色风格会损失掉。

训练所用的图片是512*768，建议生成这个分辨率的图像，训练用的图像经过了一定的裁剪/背景的涂抹。

注意：

1.如果角色经常将手伸出画面外，画面构图像是从一张完整的图像中截取下来的，可以添加对于角色服装，身体部位的描述tag（如shoes，hand，leg），用这种方法来让角色完整起来。

2.lora对于兽耳有过拟合，可以添加负面tag: cat ears，animal ears

3.在不使用高清修复的情况下，这个lora对于脸部的绘制效果不好，我一般会添加Denoising为0.4的重绘

4.可以使用Lora Block Weight插件分层控制Lora权重，当与其他具体角色Lora结合使用时可以参考以下做法:

Nico Lora 使用MIDD + OUTALL,或者只使用OUTALL

具体的角色 Lora 使用MIDD，

如果角色Lora对于风格的过拟合够低（从二次元到写实都能生成），那可以产生不错的效果。

Hi all, this is a sister project of AnimeFus - Kyo Mix v0.3, which trained with anime characters from KyoAni into a single model. On the other hand, AnimeChar aims to remake the character with higher quality using Lora based on our dataset.

Please check "About this version" to know more details about the model version and included characters. The model base is Anything v4.0.

-- Please also support --
AnimeChar - Tamako Mix | Kitashirakawa Tamako (北白川 たまこ) & Asagiri Shiori (朝霧 史織)

AnimeFus - Kyo Mix v0.3

some original character, 60 images used

weights: ~0.8

models used for post: grapefruit

model: most of anime-like models

triggle words :

HoshinoAi, long hair, purple hair, purple eyes, hair ornament, star-shaped pupils,

This is a character LoRA for Haibara.

The base model used is the [anythingelse-v4](https://civitai.com/models/4855/anythingelse-v4), training with 2500 steps on nearly 30 images. I find that a good anime base model is necessary here.

Demo images use Danger Zoo ~ Johnsons Negative TI Collection ~

"modelname style" is the main trigger word
additional trigger words are listed to keep you close to the trained style/concepts.

This LORA is designed for used with other djz LORA's and/or on top of Mixed pairs of djz CKPT's or a Triple merge mix including other CKPT models. This is a specific method for merged CKPT models while running a stack of these Style LORA models on top. You may still use this alone and it is recommended to include the additional trigger words in your prompts.

DJZ Season 2.5

Demo images use Danger Zoo ~ Johnsons Negative TI Collection ~

"modelname style" is the main trigger word
additional trigger words are listed to keep you close to the trained style/concepts.

This LORA is designed for used with other djz LORA's and/or on top of Mixed pairs of djz CKPT's or a Triple merge mix including other CKPT models. This is a specific method for merged CKPT models while running a stack of these Style LORA models on top. You may still use this alone and it is recommended to include the additional trigger words in your prompts.

DJZ Season 2.5

Demo images use Danger Zoo ~ Johnsons Negative TI Collection ~

"modelname style" is the main trigger word
additional trigger words are listed to keep you close to the trained style/concepts.

This LORA is designed for used with other djz LORA's and/or on top of Mixed pairs of djz CKPT's or a Triple merge mix including other CKPT models. This is a specific method for merged CKPT models while running a stack of these Style LORA models on top. You may still use this alone and it is recommended to include the additional trigger words in your prompts.

DJZ Season 2.5

Demo images use Danger Zoo ~ Johnsons Negative TI Collection ~

"modelname style" is the main trigger word
additional trigger words are listed to keep you close to the trained style/concepts.

This LORA is designed for used with other djz LORA's and/or on top of Mixed pairs of djz CKPT's or a Triple merge mix including other CKPT models. This is a specific method for merged CKPT models while running a stack of these Style LORA models on top. You may still use this alone and it is recommended to include the additional trigger words in your prompts.

DJZ Season 2.5

Demo images use Danger Zoo ~ Johnsons Negative TI Collection ~

"modelname style" is the main trigger word
additional trigger words are listed to keep you close to the trained style/concepts.

This LORA is designed for used with other djz LORA's and/or on top of Mixed pairs of djz CKPT's or a Triple merge mix including other CKPT models. This is a specific method for merged CKPT models while running a stack of these Style LORA models on top. You may still use this alone and it is recommended to include the additional trigger words in your prompts.

DJZ Season 2.5

Lee Chae-young

Trigger word : chaeyoung

All member, version + gyuri

here : -> fromis9+gyuri <-

Use this wisely. I'm not responsible for anything you create with this LoRA.

Jang Gyu-ri

Trigger word : gyuri

All member, version + gyuri

here : -> fromis9+gyuri <-

Use this wisely. I'm not responsible for anything you create with this LoRA.

本人不喜炼人，酷爱炼机甲

权重暂时建议0.8-1

后续会增加训练素材

未来继续炼更多各种机甲

I don't like refining people, I love refining mechs

The weight is temporarily recommended to be 0.8-1

Training materials will be added later

In the future, we will continue to refine more various mechs

另外放一个小链接https://www.bilibili.com/video/BV1fm4y1177z/?vd_source=fd14ef3c7aa9e8cb3e3f08b04caed768

LoRA for the actress Lily James.

Use it with 1 weight.

Feedback and constructive criticism is welcome in the comments.

Shinju Inui 乾 心寿 | My Dress-Up Darling

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

Here is the Lora for condom in mouth!

Here are the trigger words suggested, condom in mouth, condom wrapper (in v2), used condom (in v3)

Hope you enjoy my work!

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

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

Form for requests.

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

Well it is done. That was honestly way more difficult then I imagined. The main problem with making this lora is quite honestly just the dataset.
I have no idea why, but I couldn't find two pictures in a row where she had the same hair style. Not to mention that 99% of the available images were in her sailor uniform.
I still hope you will be able to make cool pictures with it.

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

The main prompts are "SMMars"

You can try "very long hair" and "parted bangs" for the hair style (i don't know her hair style is really inconsistent)

Add "1990s \(style\)" to your prompt, if you want that style.

For her outfit try:

"SMMars, sailor senshi uniform, red sailor collar, red skirt, elbow gloves"

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.

We will comply with the 774inc. secondary creation guidelines.

It's my wife!!!!!

22222

TLDR (Because I don't care)

Use Honey Lemon <lora:honeyLemonFromBigHero5_v10:1> in positiv promt.

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

Originally, I wanted to make a LoRa of Honey Lemon in the actual Big Hero 5 movie, but she just didn't get enough screen time there, and the fact that she was wearing a helmet for half the movie didn't help at all.

So I used the animation series instead. I do not really like the art style, the clothes she is wearing here look like a sack of potatoes most of the time, and in general Honey Lemon does not really look like the character in the actual film.

But there was a lot of potential training data (and besides; it's completely 2D, so another challenge :D), so I was hooked. ^^

Training

I stuck to the new normal (at least as far as animated stuff goes) and looked for good scenes of Honey Lemon in the whole series, I barely managed to get 50 pics together as she's quite underplayed.

Seriously, the rich guy got several episodes that mostly told his story. Honey Lemon only got one episode where she was the centre of attention.

But okay... that's one of the problems we have to live with when doing (western) animated stuff, changing clothes is not allowed.

Usage

No weighing is required, so it's as simple as that:

Honey Lemon <lora:honeyLemonFromBigHero5_v10:1>

It heavily depends on the Checkpoint you are using which sampler should be used, so no recommendation there this time.

In addition, some checkpoints did not work too well (Completely ugly colours, even worse than in the case of a missing VAE) with v1... so I re-trained a v2 with a noise offset set to 0 (v1 had an offset of 0.1), which fixed the problem (if anyone could enlighten me as to why this is the case, please leave a comment).

Because in some cases the image generated by v2 did not look right with DreamShaper v5, I also provided v1. I added the "maid" image as example for both versions...

属于意外的惊喜，效果还不错，如果喜欢的话就拿走吧，大概率不会更新了，会尝试其他新的模型。推荐使用Rev模型，Rev YYDS! 二次元模型效果也很好,不同的大模型有不同的效果，希望大家能喜欢哈~

It is an unexpected surprise, the effect is not bad, if you like it, take it away, i cant promise i will update, and I will try other new models. It is recommended to use the Rev model, and the other models are also very effective.hope you like that

Here is the Lora for yuri tribadism !

Here are the trigger words suggested, tribadism, sex

Hope you enjoy my work!

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

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

当你看到这个lora的时候，就表明我在

Pov_Hand-Early Access | Stable Diffusion LORA | Civitai

这个lora的研究中取得了阶段性的进步

伸出两只手都能画好了，一只手还远吗？

When you see this lora, it means that I am in

Pov_Hand-Early Access | Stable Diffusion LORA | Civitai

I've made some progress with this lora

I can draw with two hands outstretched, is one hand far behind?

Lora of Yuzuriha from Jigokuraku, works around 0.7-1.0

地狱乐杠的模型，推荐权重0.7-1.0

-

Official Outfit-原皮肤=OfficialOutfit, Kunoichi Clothing

-

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

Try and learn to generate mockups for brand design.

Trained based on the ChilloutMix model.

This is the character Miyu Kozuki from the animation Da! Da! Da! (UFO baby).

The word "daiyonchugakkou seifuku" is her school uniform.

LoRA of Yukari in PCR, works between 0.7-0.9

公主连接黄骑的模型，推荐权重0.7-0.9

-

Supports 7 sets of official outfit

支持7套官方外观，

-

1-2starDefault-1-2星外观=white gloves, 1starOutfit, Yuraki 1star Dress

3-5starDefault-3-5星外观=3starOutfit, Paladin Outfit, gauntlet

6starDefault-6星外观=6starOutfit, Advanced Armor Dress, gauntlet

Summer-泳装外观=SummerOutfit, Yukari's Swimsuit

Yukari Camp-露营外观=CampOutfit, Yukari's Camping Clothing

Yukari Holiday-圣诞外观=ChristmasOutfit, Yukari ChrismasDress

Yukari in reality-现实由加莉=RealityYukari, human, black hair + "elf, pointy ears, ears" in negative prompt

-

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

You can get an idea about the specifics of this LoRA by looking at example images (the dataset source is mostly わす Youtube Channel).

The base model is Suzumehachi.

If you are the artist whose style is conveyed in part or fully by this LoRa, I suggest that you introduce this tool into your workflow. That way you can please your fans with even more art.

Character model

Trained on anything v4.5. Different characters will need different weights to work well. It's more versatile at lower weights, while at higher weights it will be more rigid but the style will be very close to the one of the show.

Hi guys this is my first Pose model it.Half of the time works perfectly,the other half,lets just say its not for the light hearted people,feel free to experiment if you get a good image use image2image for better results hope you like it

Use weight around 0.5-0.7

If you are able to support me please consider checking my Patreon

All my models are free and will remain that way

Bundang Thorn Field here..........

Gahyeon ~yuhuu...

Trigger word : gahyeon

V2

here : -> Gahyeon VIP <-

Use this wisely. I'm not responsible for anything you create with this LoRA.

Made with chilloutmix_NiPrunedFp32Fix with 90+ images. Tested with:

• 768,768 image and lower. Use Hirezfix for bigger and better results.

• Use DPM++ 2M Karras or DPM++ SDE Karras. Euler a should be fine also

• Use 20-30 sampling steps

• Strength: 0.8 - 1.0

• Restore faces: ON

Legal & Risk

• NOT FOR NSFW USE! NO FOR PORN USE!

• All content created are free, BUT CANNOT be used for "commercial use".

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

这是一个水彩风格的LoRA

权重推荐使用0.75-0.9，如与其他lora搭配请自行判断。

Suits from bsg show, right now battlestar officer suit and viper pilot suit are present.

Trigger words for the newest version are battlestarofficer and vipersuit. The Lora can also generate Vipers, the trigger phrase is colonialvpr fighter jet. Have in mind that those vipers are intended to use as parked in the hangar.

You can also add "in galacticaCIC room" when generating officers to get a nice background. This trigger word is however not very flexible at the moment.

By using hollowstrawberry's dataset maker. Testing if this work.

v2 is out. Suggested weighting at 0.6-0.9

Yuka Tachibana 橘結花 | Tachibana-san's Circumstances With A Man

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

蠢沫沫

Kishiyo's style. He is famous for designing Takao class in azur lane and type 95/97 in girls frontline.

No commercial use. All credits go to Kishiyo.

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

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

Please use ChilloutMix and based on SD1.5

• 768,768 image

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

• Use 18 sampling steps

• Strength: 0.7-0.8

• Hires. fix: on

• Hires step: 18

• Sampling step: 18-20

• Denoising: 0.7

漆原智志风格的复古二次元lora，建议大模型用二次元模型，如Nai，any，橘子等

免责声明：学习素材来自于互联网，不得用于任何商业用途，否则后果自负。

抱歉英语不好，以下机翻：

Sorry ！ my poor English，Here is the machine translation:

Satoshi Urushihara's style of retro ANIME lora, it is recommended to use ANIME models for large models, such as Nai, anything, orange, etc

Disclaimer: Learning materials are from the Internet and shall not be used for any commercial purposes, or the consequences will be taken at your own risk.

This model can be used to create black and white hand drawn art works

Sylvia van Hossen | Princess Lover

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

Activation Prompt: chibi

Recommended Negative Prompt: text

Test of Dreambooth check point converted to LoRA

Chloe クロエ | Kuroinu 黒獣

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

https://www.animecharactersdatabase.com/characters.php?id=21629

https://vndb.org/c32490

https://anidb.net/character/39553

I offer all my content at no cost to you. Should you wish to show your appreciation and support, kindly consider buying me coffee via the following links: https://www.buymeacoffee.com/konyconiN , https://ko-fi.com/konyconi7992 .
Your generosity is greatly appreciated!

Do not hesitate to give me suggestions.

CHECKPOINT:ChilloutMix

weigh:0.5~0.7,use hires fix 0.5~0.58

This lora is trained based on v1-5-pruned. It simulates the comic style of Ping_Pong,hope you like it! Any feedback would be appreciated!

This is LoRA, which embodies the 'Seiko-chan cut', the hairstyle of Japanese idol Seiko Matsuda in the 1980s.

Recommended weights : 0.6 ~ 0.7

Trigger Words : Seikochan cut

My first LoRA! thanks.

Korean Weather Girl Mix

Recommend to use 0.6~0.8 weights, but may look great on 0.3~1.

Works great on basil mix, but works good on ChilloutMix, etc.

https://linktr.ee/promqueen1