A lora of the <a target="_blank" rel="ugc" href="https://www.victoriassecret.com/us/vs/sleepwear-and-lingerie-catalog/1126294400/-/a/generic-11262944-choice-54A2/dare-to-dream-lace-strappy-triangle-teddy-black">Dare To Dream Lingerie Lace Teddy</a>. This was trained with only 4 images. The resulting model is slightly overbaked and without a strong prompt the training data will emerge, but I've had some decent results. I might make an updated version.With this lora you will need to have a strong prompt. You prompt will need to include the description of the subject, pose, and environment. Otherwise, your models will look a little thick. Additionally, i recommend a strength between .5-.8. But you can also get some ok results at 1.

Files with SHA256: 160fe7e6f5cf2e4902b53659691c4cd3ea8205e1c4d16904d2d4d91619772baa