Hi, there:
I am confused by the dalle-3 api when it creates simple entity, like carrot, banana, the generated result always contains drawing pencils, or a hand drawing the entity. even if I have added instruction “not to generate drawing-pencils”, it still doesn’t work.
However, the same prompt would be OK if I use the Dalle in the chat portal.
Any idea?
P: The following is the revised prompt returned by dalle-3 api:
Create a cartoon image that focuses on the subject: broccoli. Keep the background simple and avoid adding unnecessary information to ensure that the image is clear and easy to recognize.
The generated output is:
