This LORA attempts to create images in the style of Chinbotsu's (NSFW) doujinshi.
Another of my all-time favorite h-manga artist.
muscular futas

Recommended weight: 0.8 ~1.0
↓ !! CLICK "Show More" FOR RECOMMENDED USAGE !!
(TAGS UPDATED FOR RECENT VERSION)
Suggested tags:
1girl, solo, mature female,
close-up, | upper body,
muscular, muscular female, | abs, navel, obliques,
thick thighs,
large breasts | huge breasts,
nipples, puffy nipples,
lipstick, makeup,Optional tags:
futanari, penis, erection, veiny penis, veins, testicles, | huge penis,
(ahegao), tongue, tongue out, saliva, open mouth,
(close-up:1.3), (portrait:1.3), -> with ahegao tags above
dark skin, | tan, tanlines,
mole, | mole under mouth,
cum, | cum in pussy, | cum on hair, facial, | cum on body,
from below, | from above,Negative prompt ( IMPORTANT or you will get comic / black and white images ) :
(EasyNegative:0.8), -> or any negative embedding of your choices
(worst quality, low quality:1.2), (watermark, white border, artist name, signature, blurry, simple background, text, speech bubble, comic, monochrome, greyscale),Notes: You might need to add " ( ) " emphasis weight to your background/location tags in your prompt or you will get plain backgrounds.
The futanari tags don't always work and have issues with some body poses like "squatting | lying, on back" , "abs, navel" or "huge penis". Removing them might resolve that problem. Adjusting their weight could also help.
Their "shapes" are pretty limited due to the small dataset pool. Feel free to use any futa LORA.

Click on the 🔔 or Follow me if you want to receive a notification when I update or release new models.
If you somehow stumbled upon this LORA, check out my profile.
I have other models that might also pique your interest :)

I would love to see what you've generated with my LORAs.
Don't be shy to post below in the gallery or leave a Like.
Every little contributions motivate me to create more content.
Enjoy! :>
If the original artist does not approve of this model and wish it be removed, I will take it down.
Description
This version should be greatly improved overall.
Redone the tagging
Cleaned up dataset
Better training parameters