Here is version 2 of mthxact (that’s the trigger word). Seeking greater realism in the depiction of sexual interaction—specifically regarding textures and human body elasticity—I trained this LoRA using a diverse dataset ranging from simple finger insertion to enthusiastic triple penetration (oral, vaginal, and anal). I carefully selected images to cover a wide range of lighting conditions: studio flash, natural light, harsh, diffused, direct, white, golden, etc. For some reason, v1 didn't remain stable for long—barely a week before it became impossible to load (perhaps due to a Comfy update?). I switched training models, and I’m really pleased with the results so far. Over time, I’ll expand the dataset, particularly to improve stability for certain positions. For now, though, I’ll let you test it out and share your feedback—good or bad. Feel free to play around with the weight slider; I prefer it around 1.10, with 1.30 as the maximum. base <span style="background-color:rgb(26, 27, 30);color:rgb(193, 194, 197);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, Helvetica, Arial, sans-serif, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;;font-size:16px">model : <a target="_blank" rel="ugc" href="https://civarchive.com/models/2735032/pornmaster-krea2?modelVersionId=3171380">https://civarchive.com/models/2735032/pornmaster-krea2?modelVersionId=3171380</a>

Files with SHA256: 8959e55c93a6507d5f9e69ea89b51664dda8747f9ddfa13d0d9eb344d3537c2a