Update (3/2/2023)<ul><li>The Reverse Suspended Congress (Anal) LoRA has been released!</li><li>Ten new images were added to the dataset.</li><li>I suggest using this at 0.75-1.0 weight for best results.</li><li>USE <a rel="ugc" href="https://www.youtube.com/watch?v=UKHpGHepq4A">CLIP SKIP 2</a>.</li><li>Include 1boy, penis, reverse suspended congress, anal with the LoRA to activate it.</li><li>Embed and LoRA used in prompts: <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/Nerfgun3/bad_prompt">Bad Prompt </a>- <a target="_blank" rel="ugc" href="https://civarchive.com/models/6285/afrobull-style-lora">Afrobull Style</a> -</li><li>It has some of the same issues as the embed but good results are more consistent in the LoRA.</li><li>If you increase width to 768, you'll get a much wider shot and in my opinion - It makes a lot better.</li><li>You can use the LoRA in conjunction with the embed to create a slightly more intense effect.</li><li>I'll probably actually make the vaginal version sometime in the near future.</li><li>Everything I'm making will always be free but if you want to show a little bit more support, you can buy me a coffee - <a target="_blank" rel="ugc" href="https://ko-fi.com/ekune">https://ko-fi.com/ekune</a></li></ul>================================================This is a textual inversion embedding to generate Reverse Suspended Congress. Use RSC to activate the embedding.This was trained on Novel AI with a dataset of 130 Frontal pictures. Example Images were made on <a target="_blank" rel="ugc" href="https://huggingface.co/WarriorMama777/OrangeMixs/tree/main/Models/AbyssOrangeMix2">AbyssOrangeMix (Hard)</a>.This is my second sexual position embedding and I really like the results of this one. It's mainly anal. It was trained on mainly anal hence why it's the anal version.There were some obvious flaws but It mainly gave out great results. I don't know if this is because I poorly trained it, or my dataset was garbage or if it's unavoidable right now with AI but I shall continue to Improve.A character will most likely be next.Flaws I've seen:<ul><li>Sometimes won't generate the entire legs.</li><li>The arms holding up the girl will become legs.</li><li>The girls arms will become legs.</li><li>Anything that looks like an arm, will be a leg.</li><li>Everything comes out well, oh. The hands are feet.</li></ul>Other Embeddings Used: <a target="_blank" rel="ugc" href="https://huggingface.co/datasets/Nerfgun3/bad_prompt">Bad Prompt</a> - <a target="_blank" rel="ugc" href="https://civarchive.com/models/4536/kuchiki-rukia-ti">Rukia</a>Check out my other embeddings for characters and position here: <a target="_blank" rel="ugc" href="https://civarchive.com/user/ekune">https://civarchive.com/user/ekune</a>

Files with SHA256: f055e5ca191abdb8b7f3cfc6df6ea988d32be24c20da4736c5986fc52e432218