<h1 id="v2-is-available-now!"><u>V2 is available now!</u></h1><h2 id="trained-with-higher-quality-dataset-and-more-consistent-round-face.">Trained with higher quality dataset and more consistent round face.</h2><p><strong>Version 2 does have tendency to be too glossy on the skin (my fault), so you might have to play around with prompting or inpainting to get the glossiness out.</strong></p><p></p><p>Character Lora depicting generic female mannequin (can do males too). Can be used for SFW or NSFW purposes.</p><p>Adjust as needed to your preferences. It was trained without explicit reproductive anatomy, but you can change that if you want.</p><p>Made with <a target="_blank" rel="ugc" href="https://civarchive.com/models/827184/wai-illustrious-sdxl">WAI-illustrious-SDXL</a> checkpoint and <a target="_blank" rel="ugc" href="https://civarchive.com/models/1202699/facelerss-ilxl">Facelerss ILXL</a> Lora for body and face during training.</p><p></p><h3 id="v2-positive-prompt:"><u>V2 Positive Prompt:</u></h3><pre><code>femalemannequin, female, white body, bald, no face, no eyes, no nose, no mouth</code></pre><p>or if you want male:</p><pre><code>femalemannequin, male, white body, muscular, bald, no face, no eyes, no nose, no mouth</code></pre><p></p><h3 id="v1-positive-prompt:"><u>V1 Positive prompt:</u></h3><pre><code>femalemannequin, solo, completely white body, smooth, lines around limbs, breasts, faceless, no eyes,</code></pre><h3 id="v1-negative-prompt:"><u>V1 Negative prompt:</u></h3><pre><code>female flesh,</code></pre><p><strong><u>V1</u></strong> - This Lora has a strong tendency to include other mannequins in the scene. Include this in the negative prompt:</p><pre><code>another mannequin in scene,</code></pre><p>By doing that, you alleviate the frequency of having more mannequins in the background/foreground and if you only want one mannequin as the main subject.</p><p></p><p>May be more improvements in the future.</p>

Files with SHA256: c343b988b872cba0896a670ddfd8fb66469fd52d5a938d8080828edfa2d29489