suggest weight: 0.9
recommend prompt: multiple penetration,multiple penises,ass focus,looking back
GLHF
Description
not stable
FAQ
Comments (3)
I'd love to use this model and recreate the images you've posted, but I have a few problems. The triggers listed don't work. The triggers aren't part of the trained data whatsoever. Why are your images censored? This isn't a japanese based site, so images don't require a censor, and is quite annoying. can you post the actual generated image with it's metadata so we can recreate them? Example images are meant to be a guide for users to know what the model can do and show how to do it.
the trigger word is exactly what I used to create these pics, but the post has other personal prompt like 'huge penis'、'simple background'etc, well if you cant recreate pics like post is also reasonable, as I've tried on different weight, this lora aren't stable enough, cant perform well on other checkpoint.
Ignore the censorship if you may, what I censored doesnt really concern the topic, the poses are more important.
@Tifa_ sigh you completely missed the whole point of my comment... I made the point that the triggers you listed aren't in the metadata of the lora itself. This can be easily checked with xypher7.github.io/lora-metadata-viewer/. Therefore, the triggers listed are not the actual triggers for this model, unless you have intentionally scrubbed some of your metadata (which opens up a whole new issue on its own). other users obviously agree with me based on the reactions my comment got. Feel free to prove me wrong though! upload an image including metadata that allows us to actually replicate your lora and I'll delete this whole thing. Until then however, I believe this whole lora is a fucking hoax. And regarding the censors... Editing the image resets the metadata, and removes any prompts that may have been used for the images. Just save yourself the effort and Don't censor them. This is a NSFW model, so there's no need to censor them.