Works well with specific details for Detail 1 and 2, and then a semantic detail for the third one. Like in the example image i used: A cartoon [SETTING] scene with crayon colored [Detail1] and [Detail2] and [Detail3], like that of a learning child. Setting="Sunset" Detail1="trees" Detail2="bushes" Detail3="creek flowing through forest" 来源：https://prompts.chat/prompts/cmjhpxfd60005jr04ro88q7nh
