Is with a great pleasure that i present this Lora.This lora is supposed to provide, a implement "angle" view from a spanking("Third person" from the implement to the butt) I recommend use the keywords<code>spanking implement focus (wooden spoon)</code><code>spanking implement focus (paddle)</code><code>spanking implement focus (small paddle)</code><code>spanking implement focus (hairbrush)</code><code>spanking implement focus (belt)</code><code>spanking implement focus (hand)</code>But getting a good seed on targeted implement on this is very hard ! Sometimes over 1/16+Belt HARDPaddle EASYHand HARDSpoon EASYHairbrush on correct side IMPOSSIBLE I recommend strength between 0.6 and 0.9<h3 id="heading-223">To upscale</h3>i recommend use img2img and reduce this lora strength to `0.6` or less<h3 id="heading-350">Training</h3>This were made using only 43 images, that's why is hard getting the correct implement.For making this actually good, i think i would need 20+ images for each implement<h3 id="heading-29"></h3>

Files with SHA256: a15f08c987756a4a62afb5ba88739e8b66021a4cf20fe9195cc7f62d9ffd5baf