Lora trained in the G1_(jjw00123)'s art style. Use about 300 images for training, and manually proofread the prompt words. Only contains human male, and the style is biased towards bishounen. Of course, female or furry content can also be generated, but the effect is not guaranteed.