This LoRA is trained with stock pictures of the bra brand Nessa.The current LoRA knowns these following models:<ul><li>clarisse</li><li>cleo</li><li>klara</li><li>kora</li><li>leticia</li><li>mamma</li><li>moly black</li><li>moly nude</li><li>naomi</li><li>rubin</li><li>selena black</li><li>selena red</li><li>selena white</li><li>tatuaz</li><li>tess</li></ul>strength could be around 0.4~0.8as all image in the dataset have head out of frame and very simple background, the LoRA tends to cut head and simplify background. strength higher than 0.8 is not recommended.

Nessa bra

Less faithful to the training dataset than the SDXL version, but still produces very good images.Trigger words are the same than sdxl version.

Files with SHA256: 6fe24bd33f422bfa76706beca54e94ef312e2129f103fbd65c6649cfb3b0c8a8