This is a LORA model based on the training of Ju Jingyi with the main trigger word Jingyi, chinese Idol.
I used 100 training sessions per image, based on 33 of published photo. Results may have some facial distortion, but are still relatively effective.