<h1 id="heading-4063">Intro | 简介</h1>This is another LoRA for Futanari or Femboy or Otokonoko. 又一个扶她LoRA，因为大鸡鸡女孩永远都不嫌多。<pre><code>What is the difference between this LoRA and two hundred other futa LoRA out there? Well I'm glad you asked. This LoRA does not have trigger words like "futanari", "otoko no ko", "1boy", "penis", "testicles" etc etc. It just works with "1girl", nothing else. 
这个LoRA与其他扶她LoRA不同点在于我去掉了各种复杂的触发词，你只需要输入1girl，并搭配相应的概念即可生成。所以请勿额外添加如"futanari", "otoko no ko", "1boy", "penis", "testicles"，它们是多余的，且可能造成未知的结果。</code></pre><h1 id="heading-4064">Usage | 食用方法</h1>Currently this LoRA contains four concept: erected penis, flaccid penis, chastity cage and penis in panties. Each has their own trigger words. 当前版本LoRA共包含四种概念：鸡鸡勃起（大小随机）、疲软（小小的也很可爱）、贞操锁以及被内裤覆盖。请使用以下触发词来生成对应概念。🔹勃起 | For erected penis start with:1girl, erection, | This concept contains the most images in the dataset, so it should be pretty stable.🔹疲软 | For flaccid penis start with:1girl, flaccid, | The flaccid concept I trained took inspiration from <a target="_blank" rel="ugc" href="https://civarchive.com/models/149989/flaccid-small-penis-flaccid-micropenis">this LoRA</a>. So it is very tiny. Personal taste, don't judge😉. Don't add <s>"</s><s>bottomless</s><s>"</s>or horrible things might show up when you img2img upscale.🔹贞操锁 | For chastity cage start with:1girl, chastity cage, | This is pretty stable.🔹内裤 | For penis in panties start with:1girl, penis in panties, bulge, | This one is a bit tricky to work. Adding "erection" will make penis pop out but testicles remains inside the panties.If it keeps generating panties, add "bottomless".<h1 id="heading-4065">Details &amp; Further Update</h1>This LoRA was trained on my <a target="_blank" rel="ugc" href="https://civarchive.com/models/252186?modelVersionId=284542">eroDeity | 艳神之仪 - r1_Prende</a> checkpoint. <s>So it's kind of only works well on it</s> also works on MeinaHentai. <s>Perhaps I will train a new version on NAI later</s> in the process. All images in my dataset were generated and photoshoped by myself. Dataset contains around 200 images. I tried to cover as many pose, outfit, time of day, location etc as possible. However I don't have much time recently so I decided to just train the first version and see how it goes.Currently this LoRA probably does not have the ability to generate sex pose. There is plan to include that in my future updates.

Files with SHA256: ebee2eef7da92a641834ea9bbca2e3182c60b4123c29a97084f3088d5d9e683b