The original data set is a selection of pictures I have made in the past year. It is based on my personal preferences and style. There are bonus BUFFs for weapons, armor, and some special effects. Joycaption is used for labeling. It is completely natural language, and the prompt words can be directly inferred from the original pictures. I am too lazy to write! For full-body images, it is recommended to enlarge them in high definition! I wish you all have fun!

If you are interested in my work, you can find me on Xiaohongshu. My Xiaohongshu ID is: 22757211789