V4: 1000+ images, it should be better all around than v3. Different language training and better balanced than v3. Detail is significantly improved. Let me know what you think!
V3: 300+ images for training, better training setup, more variety and even higher realism. Strength 1 will be realistic, but feel free to change it around a bit.
V2: Much improved fidelity, 180+ images for training, higher resolution dataset with more variation in poses and tagging. Strength at 1 will be realistic.
V1: For the meat lovers, 80+ images for training. It seems to produce the results I want. Strength at 1 will be realistic, but can push it higher for thicker labia.
Description
Completely revamped training style, dataset, and weights. Should be more detailed, more flexible, and overall more capable than v3. No trigger words required, but vulva and labia should be used in the prompt to engage this data.