Updated: Nov 9, 2025
concept
No trigger or phrase needed. Use with a positive weight of 0.5-1.5.
This lora is an experiment to fix the "same face" issue that some lora models seem to have. This model was trained on 1,000 images of 1,000 different people, both men and women. Please let me know your experiences, positive or negative. As I said, this is experimental and there is no guarantee that it will work. I will use your feedback to train a v2 or chain the training methods if necessary.