<h3 id="this-is-an-enhancer...of-sorts.-it-does-what-the-name-suggests:-adds-detail-and-realism-to-nsfw-output.-it-came-about-due-to-certain-sdxl-models-inability-or-unwillingness-to-portrait-proper-adult-content.">This is an enhancer...of sorts. It does what the name suggests: adds detail and realism to NSFW output. It came about due to certain SDXL models inability, or unwillingness, to portrait proper adult content. </h3><h3></h3><h3 id="please-note:-i-added-the-mosaic-to-the-demonstration-images-to-protect-other's-sensitivities.-enjoy-with-moderation.">Please note: I added the mosaic to the demonstration images to protect other's sensitivities. Enjoy with moderation.</h3><p></p>

NSFW

<p>I believe this is going to be the last version of this series, The dataset was enhanced via Topaz Photo AI, all traces of watermarks and logos removed, pictures are 1280 resolution, 1369 total in the dataset. Tagging was redone to eliminate any references to eye color, hair color, etc., location, places, objects and such. Only "Action". This is meant to be used in conjunction with another LoRA that will specify the "Subject". Best results by keeping the total LoRA strength at or below 1.80. such as:</p><p>&lt;lora:<strong>YOURCHARACTER</strong>:0.90&gt; &lt;lora:Xposed:0.90&gt;</p><p>Tried to include a variety of "actions". Try your hand, see what you can get.</p>

Files with SHA256: e8d5474b655b231dce56cc14112dc716284ca3514102ffef483b5921aee34245