Uncensored concept (female)

Hey, thank you for everyone's likes, I just started a buymeacoffee account, since I had some serious hardware damages... It seems training does take a toll on the hardware. Please support~ thank you again o(////▽////)q https://www.buymeacoffee.com/reweik

Uncensored is not a new concept, but it is kind meh in most base models, especially for the lady parts.

Main trigger:

pussy (for ????? with hood, not major focus), spread pussy (lightly opened up ?????, hands may be involved, the les most likely will spread as well), cervix (open up more than spread ?????, so you can see the interior, and the the cute ??????), cleft of venus (I didn't focus on this tag, mostly just give you no hood ?????, most ???? models can already do that with no problem, it's a curved line representing ?????). If your base model understands these some other way, use underscore instead of space between words. Using them together may have unexpected result.

Roll the dice trigger:

uncensored

Focus trigger:

close-up (it will only work with spread ????? and cervix)

Other:

You can also get creative and use words like uretha, ????????, ????, and etc.. They're all capped in the caption files, whichever helps you to get the perfect thing you want.

Please note that the from back view can happen as well, since there's not tons of data, it may take a few tries to get it perfect (extra ???? and such), but the lora can clearly tell the direction of the pussy.

The point of this lora is to give more details to ?????, so it doesn't come out as in some unspeakable shape. It has some flaws, like it may give you extra uretha... or ????????... depend on your setting for highres. Or you can use upscaling, if it's not too much trouble.

I've used a few ????? loras on civ, they all feel a bit overfit or changing the style of base model way too much, including the previous pussy-gaping lora made by myself.

So I gathered about 6k uncensored imgs (anime) on internet, went through the hellish data selecting and editing (I can still see those pictures when I close my eyes), trained this lora, baring the idea that with enough material, the lora has enough knowledge so it's not biased on the style.

With a few tests, it significantly changes the ????? shape and is more responsive to the ???? triggers. Also, it can be worked on models with no "?????" capability to a degree, can work on realistic models as well (will change style obviously).