SWEETAI | OC Claire Addams | V1

Some fun concept i had in mind, the first time i saw a goth girl was when i was a kid in the 90's (1991 version)

and you can bet your ass i goth the goth (lol) fetish right there

and why not mix the most beautiful girl in the world (Claire Redfield, RE2R) with Wednesday Addams (OG One, outfit and hair only, albeit more...skimpy), well? thats exactly what i did!

also remember, my work will be ALWAYS FREE and continuously improving, so all i ask is for a donation, if i don't get enough support i will stop providing free models eventually

oh and yeah i learned captioning (i had the click, finally i can sleep now!), i mastered SFM, i mastered Blender, and i will fucking master AI, learning alone with frustration increases one will power instead of relying on others

all is missing is NMKD fixing inpainting on 1.12 (it worked on NON-INPAINTING models on 1.9.1!!!) and adding ControlNet (gotta make these horny poses!) then i'm set, i don't give a fuck about

sdxl because it looks plastic and i read/heard that the training speed is abyssmal even on a 3090, that's absolutely not for me, and it is taking a long time to mature

note: there is some fuckery going on with her hair as the hair i got from Daz (a 2million vert hair, what.the.fuck) had a major split in the middle, so you might get some heavy baldness on that split in the middle.

note2: i will always include example prompts so you don't have to smash your head against the wall trying to guess then

note3: if you aren't getting enough braids make sure to include it on your prompts to give the AI a hand.

note4: people say you don't have control over AI and AI does all the job, let me change that, this is me doing already doing 60% of the lifting (i can do every single of these renders in blender, but slower ofc, AI is just a tool to speed an artist workflow not to replace the artist), we are like brothers now, working together for a world free of paywalls.

=====

Trigger Words are claire for face, and addams for outfit

Trained on 209 images (16 which are rendered at 4K 512 samples for close ups)

Trained on resolution 768x768

Tested on 1024x1024 resolution (high-res fix enabled)

Trained on BASE 1.5

CFG 7

Steps 40

Tested on CyberRealistic 3.2 only

Trained using NMKD GUI at a Learning Rate of 0.00070 and 740 Training Steps ( thank you https://talkai.info/)

Weight used on previews: 1

====

DISCLAIMER

I did not include ANY image of Jordan McEwen in this dataset, they are ALL exclusively rendered using Blender at a resolution of 1920x1920 using the in-game character extracted directly from the game (Resident Evil 2 Remake)

#========================#

TOOLS USED TO CREATE MY MODELS

#========================#

Blender 3.5 (Custom Dataset Creation for Characters)

DaVinci Resolve (Improve Dataset Quality using Various Techniques)

NMKD Gui (LyCORIS Model Training and Debugging, Image Previews)

#========================#

REQUESTS

#========================#

Don't forget to leave a request for a LyCORIS Model, either in the comment section or sending me a private message on my Patreon page.

#========================#

DONATION GOAL

#========================#

Help me reach a goal of $3,000/mo on Patreon https://www.patreon.com/SweetAI when the goal is meet every model released afterwards will include ALL images present in their Dataset, this way you can improve or create your own LoRA/LyCORIS based on then then.

#========================#

USAGE AND CREDITS

#========================#

Feel free to use ANY of my LyCORIS/LoRA models as you please, as long you credit me and link to my Patreon https://www.patreon.com/SweetAI

#========================#

LINKS

#========================#

https://www.patreon.com/SweetAI

https://civitai.com/user/SweetHentaiAI

https://www.deviantart.com/sweethentaiai

https://www.pixiv.net/en/users/95883777

#========================#

PREVIEWS

1-3

claire, raw photo, dof, bokeh, lifelike, a gorgeous addams with huge cleavage and piercing blue eyes on a foggy castle holding a red teddy bear

4-5

claire, raw photo, dof, bokeh, lifelike, a gorgeous addams with huge cleavage and piercing blue eyes on a foggy castle staring at the viewer with her toned abs

6-7

claire, cinematic film, dof, bokeh, lifelike, a gorgeous addams with huge cleavage wearing a black bride dress holding a bouquet of black dead roses, in front a foggy cemetery, black veil, black roses, braided hair

8

claire, cinematic film, dof, bokeh, lifelike, a gorgeous addams with huge cleavage showing her delicated huge nipples, on a boardwalk in the 40s,

NEGATIVES

(deformed iris, deformed pupils, semi-realistic, illustration, painting, cartoon, cg, cgi, 3d, render, doll, blurry, anime, jpeg compression, drawing, sketch, chromatic aberration, perfect)+, bad art, disfigured, deformed, extra limbs, open mouth, teeth