<h1 id="heading-1024">Women of A.I. - <span style="color:rgb(250, 82, 82)">LivvyLuv</span></h1><p>Experimenting with methods of creating fictitious models. LivvyLuv is my third LoRA not based on an actual person. The idea is to create models that could be celebrities if they were real. Models that you look at and think haven't I seen that person before.</p><p></p><p>L1vvyLuv was not based on an actual person. The LoRA was created on a picture set of AI generated images from my FULLY_REAL_XL checkpoint. I still have some work to do to improve the process. <span style="color:rgb(64, 192, 87)">I recommend a LoRA strength of .85 to .95 depending on checkpoint and prompt size.</span></p><p></p><h2 id="heading-166"><span style="color:rgb(121, 80, 242)">Try this with the FULLY_REAL_SD checkpoint. I will add it to the suggested resources. All sample images created on that checkpoint for this LoRA.</span></h2>

Files with SHA256: 3a6f074d6bf252463ee468ed8b54536403a8619cbf5cee6ceffb669e43de7743