A model for nylons and fishnets in their various styles and colors. Trained on a large, high-resolution dataset with some feet and legs solo content, some hardcore sex content, some amateur-style content. Still can't consistently generate normal-looking feet, especially closeups, for most positions.
Description
~650 images, 768-binned training, booru tags + some manual additions
FAQ
Comments (4)
Hello, was just wondering if there are any special weights or prompt triggers for this?
Mostly just what you see in the example image prompts. Tag frequency is embedded in the metadata. While this was a relatively focused model, you should be able to use any tag that was present in a decent number of the dataset images.
@honk_honk Cool, thank you. Gonna give this a try. :)
Is there anyway you could please create a veruca salt lora? 🙏