Member
- Messages
- 89
Lately, when seeking suggestions for text prompts in our AI art generator, I've noticed GPT-4 treating the art generator's capabilities as on par with its text parsing skills. I'm in the process of establishing prompt limits to help the AI understand its constraints while exploring the exciting possibilities of Stable Diffusion.
One recent addition specified that the AI can only focus on the details of one subject at a time. For instance, if you request a couch and a chair in your image, only one can be detailed further. This limitation has shown promising results. Do you have any other valuable limitations to enhance prompt creations for Stable Diffusion?
One recent addition specified that the AI can only focus on the details of one subject at a time. For instance, if you request a couch and a chair in your image, only one can be detailed further. This limitation has shown promising results. Do you have any other valuable limitations to enhance prompt creations for Stable Diffusion?