*I am constantly working on a new version. Please let me know if there are images of any particular persons, costumes, etc. that you want to be included in the dataset for training.
This lora is trained based on the images of Japanese as well as Chinese young women. The weight should not be high; e.g., 0.4 is often high enough.