I wasn't happy with how Illustrious v1 only made female subjects, so I generated 200 new images that included fifty male solo images, fifty pure landscapes, and fifty still lifes. I think this really helped even out the sets, as well as high x4000 step count per epoch.Anima v1 (Epoch 7) and v2 (Epoch 10) are from the same training data with both creating images in the style really well but with subtle differences that led me to keep and release the "v1" as it's own separate LoRA. I suggest trying both and seeing which one you like better. <h3 id="illustrious-v1">Illustrious v1</h3>Wanted to make a single LoRA for my soft and smooth style. Similar to the workflow that I used to create these images, this will pretty much only generate female main subjects. This is my first version, so I may end up scrapping it if things don't generate how I would expect them to.

B0rghese Soft and Smooth Style

B0rghese_Soft_and_Smooth_Anima_v2.safetensors

Files with SHA256: 155daa4381ad4ea4af46ec4511e81032b92bd11a0015678cc5b41a8a1dece572