<h3 id="prompt-words:">Prompt words:</h3><pre><code>socks-ill,ankle socks,calf sock,Knitted socks,lace socks,cotton socks,Lace trim,bow,ribbon,Perforated,</code></pre>This is LoRa I trained to generate better textures for socks. It can now basically distinguish between knit and lace materials. However, the definition of sock length is still vague; it often fails to differentiate between ankle socks and calf socks. This is probably because my labeling of the dataset images was also ambiguous.Also, I don't know if it's due to overfitting or some other reason, when loading this LoRa, the sock texture overlaps with other clothing images, even though there are no other clothing images in my dataset. I hope someone can answer this question.The example image uses checkpoints from Crody:novaMatureXL_v40

Files with SHA256: dc284d470c7c16788265d5433046dda467bc8e54d35c047883d073d6463780e9