By default, this LoRA generates NSFW images in 3D style. It's good at generating curvy women with wide hips and large breasts. It does big butts and images from behind very well. It can generate 2D images, but it needs be used with 2D checkpoint like <a target="_blank" rel="ugc" href="https://civarchive.com/models/24149?modelVersionId=692489">Mistoon_Anime - Pony Alpha, </a><a target="_blank" rel="ugc" href="https://civarchive.com/models/2220?modelVersionId=1118499">Babes - 1.21 Pony </a>or <a target="_blank" rel="ugc" href="https://civarchive.com/models/439889?modelVersionId=1047139">Prefect Pony XL - v4.0.</a> On checkpoint like <a target="_blank" rel="ugc" href="https://civarchive.com/models/257749?modelVersionId=290640">Pony Diffusion V6 XL </a>it generates images in 3D/2.5D style very similar to<a target="_blank" rel="ugc" href="https://civarchive.com/models/257749?modelVersionId=290640"> </a><a target="_blank" rel="ugc" href="https://civarchive.com/models/331116?modelVersionId=464939">Pony 🦄FaeTality - v1.1</a>. It does this kind of style really well with 1 strength without needing other LoRA's.This LoRA was trained using 920 nsfw images that were generated using different Pony models and LoRA's. Many of the training images were of Loba, but the LoRA seems to handle different women characters as well. (Although it tends to favor Loba, twin braids and images from behind, if hairstyle or angle are left unprompted). This is because of the images that were used in training. This can be prevented with tags like "headshot, portrait, from front, and full body portrait" and by specifying hairstyle and other features in the prompt.

Files with SHA256: f492f32965ca27d91e8710c801e6c379f1f1b29dc9f960d27c49e93b4213873c