This is an NSFW LoRA intended to produce realistic/3d futanari. This one was really difficult to train!Anyways, I hope you enjoy and your image results are satisfactory!Side note: I am aware that the LoRA may have some issues. I hope to correct them in the future, so please bear with me.<h2 id="heading-12377">Usage:</h2><ul><li>The keyword futacock is absolutely necessary. There are a couple ways I prefer to use the keyword in a prompt. Of the methods below, there's not THAT much difference so use whatever feels comfortable.<ul><li>As a noun (my preferred): Use the keyword futacock in replacement of a term like "woman" when describing your subject.<ul><li>Example prompts:<ul><li>"rainy city, film noir, black and white, futacock"</li><li>"attractive, futacock, short red hair, bangs, freckles, nude, huge breasts"</li></ul></li></ul></li><li>As an adjective/descriptor: Use the keyword with "woman" or other noun when describing your subject.<ul><li>Example prompts:<ul><li>"an attractive blonde viking futacock woman"</li><li>"an attractive blonde viking woman, futacock"</li></ul></li><li>In my limited testing I found this method gets the penis anatomy wrong a tiny bit more often</li></ul></li></ul></li></ul><ul><li>Some other default tags I include in my prompts are just the usual: beautiful, masterpiece, depth of field, attractive, etc.</li></ul><h2 id="heading-12378">Some quirks and issues you might encounter:</h2>TL;DR - Weird niche issues you might deal with<ul><li>Use the "muscular" tag with caution. It gives women a much more masculine face 9/10 times.<ul><li>I know some people like their muscle futa so I hope to incorporate better sample data in the future for this</li></ul></li></ul><ul><li>Your subject may sometimes have strangely smooth/plastic looking skin.<ul><li>To fix this, tinker with your tags surrounding your lighting and environment. It's usually a prompting issue.<ul><li>You can put things like "beautiful", "photography", "realistic" at the start of your prompt and when describing your subject, include "attractive" as a descriptor</li><li>Start with "glossy skin" in your negative prompt, but...<ul><li>If you get really desperate, try putting "barbie doll, plastic, fake" in your negative prompt. This might make your subjects a touch less attractive though.</li></ul></li></ul></li></ul></li></ul><ul><li>The tag "huge breasts" is great in my opinion, but could be very heavy-handed for others. Use as you see fit.</li></ul><ul><li>Most of these quirks stem from the nature of the training data, there's not really much I can do to avoid it at this point. Hope to fix in future updates :3</li></ul><h2 id="heading-10531">Limits &amp; Future Improvements:</h2>TL;DR - LoRA is not as flexible yet, hope to fix in the future. Built for 3D renders/photography/realism, you get it<ul><li>This LoRA was trained around realism and 3D renders, so I make no guarantee how well it interacts with other style tags (like 2D or anime) or LoRAs. I hope to improve this in the future though.</li><li>Pose variety can be limited.<ul><li>Most of the training data involves standing, sitting, or squatting so that's what you'll get in your generations.</li><li>Try experimenting with controlnet if you really want more variety.</li></ul></li><li>Your subject will almost always be nude by default.<ul><li>The majority of the training data didn't involve clothes, so the LoRA might struggle to do full outfits. Forcing clothing will probably affect the anatomy of the penis as well.</li><li>For safekeeping, feel free to include the tag "nude" even though it's not super necessary</li><li>It can do things like jewelry or thigh-highs to some success, though.</li></ul></li><li>My future goal is to have an powerful, flexible futanari SDXL LoRA that can handle all sorts of styles, poses, outfits, etc. but I needed to establish a baseline first :3</li></ul>

Files with SHA256: 6246771faf89f3803b6f86e660b49a6d1e599047842975e4e07547c55d27e97c