<h1>Time for some socks &amp; soles!</h1><p>This LoRa is specially trained on socks with soles facing the camera. However, because the training context was very narrow, <strong>this model is not very flexible</strong>. If you want more flexible "girl wearing socks in X position" please check out my other LoRa here: <a target="_blank" rel="ugc" href="https://civarchive.com/models/25297/realistic-socks-sexy-legs-and-feet-nsfw">https://civarchive.com/models/25297/realistic-socks-sexy-legs-and-feet-nsfw</a></p><p></p><p>But if you want "girl showing her socks and soles" look no further!</p><p></p><h2>Usage:</h2><p>The settings are a bit finnicky. I'll update this LoRa in the future to be easier. For now, try to aim for a strength of 0.8, and specifically include the words "feet," "toes," "socks" and "soles." I usually get about 1 good image out of every 5 tries. Let me know if you find prompts or settings that help.</p><p></p><h3>Notes:</h3><ul><li><p>Use "Narrow" and "Thin" keywords - I guess sock soles are a bit "amorphous" in a sense, so the ai sometimes likes to squish them to be disproportianately short feet. I usually add descriptors "narrow" and "thin" to describe the feet and socks. This seems to help.</p></li><li><p>Sometimes (but rarely) the feet appear upside down. I believe this is because of a few images I had included in the training set. I'll remove them and try this LoRa again to see how it goes. (Updates in the future).</p></li></ul><p></p><h3>Suggested Sample Prompts:</h3><ul><li><p>"naked woman with narrow feet, thin feet, (toes), white socks, (soles)"</p></li><li><p>"woman showing her thin feet, soles, socks, narrow feet, toes, blue socks"</p></li></ul><p></p>

Files with SHA256: 98e86cae39f3bd0c54b31a2492f7cbdd371996e993ec42e34d1943c6c6e7a1e3