(No prompt) = 0.6-0.8 > For SFW images improves lighting, detail, adds realism and 3d, favors sexier poses, and adds some fun nsfw tendency's to the AI Generation, may need to add some negatives like "topless" with some models to keep generations sfw
Pussy (Small_Pussy, Anime Pussy) = 0.5-0.9 > use for nude standing or floor spreads. Small_pussy will give large labia and less spread
Cameltoe = 0.3-0.7 > adds detailed Clef of Venus under cloths works well with "thigh gap"
Description
FAQ
Comments (13)
it's named "1678868721671732874.safetensors" for real bruh?
Is that an issue? I never rename the safetensor i figures its like the seed
@aiFunNSFW its just annoying to find that way, you can name it correctly to whatever without any issue, so its best to name it appropriately.
@aiFunNSFW Yes, the name is very strange but I don't understand... is it for Pony or for XL or for 1.5 because it is not clearly stated. The notification says Pony but on the page it says 1.5. It is very confusing.
I'm not 100% certain about ComfyUI, but A1111 and Forge do not care about file names when it comes to seeds, it only looks at the file hashes, which never changes after the file is created. With the exception of LoRAs created by a couple of the more obscure training tools, you can name most LoRAs pretty much anything you want. It doesn't affect seeds or usablility at all.
Non-sensical file names become a problem for many (myself included), because if I download 10+ LoRAs in a single sitting and then looking at my downloads folder, I will have absolutely no idea what "1678868721671732874" is and will usually just delete it instead of wasting a lot of time trying to figure out what I downloaded.
If it's something I suspect I might really want to keep, I can use Civitai Helper in A1111 or Forge to pull the model's info and then rename the SafeTensor file accordingly to whatever I want.
Lastly, is this Pony or SD1.5? Your example images appear to be Pony, but you categorized this as an SD1.5 LoRA. If I'm mistaken, I apologize in advance. If not, you may want to correct it to avoid confusion and to avoid getting the site's moderators involved.
Thanks for all the responses the pictures were all generated on PixAi (FoolKat and FiaMix++ model). I also trained this Lora there (similar to SD 1.5). I recently trained one using Pony I will load it soon once I make new art with it. I thought if i choose to build it on Pony that it might work, but after some testing I changed it to SD 1.5. I'm just starting to branch out of PixAi so bear with me for a while. will port all my LORA over eventually
I fixed the name
@aiFunNSFW Thanks, I'll wait for the Pony version, as due to space issues I had to decide whether to keep 1.5, XL or Pony,
I like the concept, but I couldn't get it to work that well with the models & LoRA's I use, especially when prompting for an "anime pussy."
That's odd. What strength were you using. Usually you don't need to even prompt it. But some model and lora will always have issues. The examples were all made using FoolKat and Fia++, both do good with H, also you can try not using it and then bringing it in for image to image enhancements. If you want something like the cover photo you'll need 'Good Pussy' Lora, they complement well, I use both at 0.5 strength. Basically any other Lora affecting that area will mess with it
@aiFunNSFW Based on what I read on the description, I set the LoRA at 0.7. Could be the models I was using.
@betweenspectrums i haven't found many lora it wont play nice with. basically just some other nudity lora, usually poses are no issue