[Feature Request]: multiple prompts input field

[MeinDeutschkurs]

Thank you very much for the effort put into this project. It is fantastic. As I am slowly starting to understand how everything works, creating text prompt-based images is incredibly fun for me. But I guess it's not just me. ❤️ Without DiffusionBee, I probably wouldn't have even been interested. Thank you for that!

During my research, I have repeatedly seen attempts to create animated GIFs from AI-generated individual images. Now I wonder if a kind of batch mode for creating interdependent individual images would be conceivable, where you specify frame by frame what should be seen.

With tools like ChatGPT, you could generate a variety of image descriptions that represent the individual frames. It would be so great if there was a kind of batch input field that could work through such a frame-by-frame prompt-for-prompt list. The individual prompts could be separated by ### from each other.

Unfortunately, or thanks to DiffusionBee, I don't have enough knowledge of the underlying technology to estimate how complex or cumbersome it is to integrate something like this.

One could also use the function to try out various prompts simultaneously to test them against each other.

Thanks for considering, Jo

[arya-coding]

Hi there,

This app stands out among many on GitHub, notably for its easily installation and smooth GPU performance on M3.

I'm looking for the same feature but for a different goal. I just want to process overnight plenty of alternative prompts. If anyone is curious, these two websites are great for prompts generation : https://www.aipromptgenerator.art/ https://sdpromptsgenerator.netlify.app/

I would suggest either a way to upload text file or an enhancement of the current textbox to support batch processing. Ideally, each line represents a separate prompt, and you can specify the number of images desired for each by appending a character and the number after the prompt, like so:

PROMPT1--10 PROMPT2--20 PROMPT3--15

EDIT :

Similar request here : https://github.com/divamgupta/diffusionbee-stable-diffusion-ui/issues/136 https://github.com/divamgupta/diffusionbee-stable-diffusion-ui/issues/25#issuecomment-1310868772 https://github.com/divamgupta/diffusionbee-stable-diffusion-ui/issues/17 https://github.com/divamgupta/diffusionbee-stable-diffusion-ui/issues/96

[MeinDeutschkurs]

I was able to submit my requests with python to its API. For your tasks this could result into a command line, something like this:

python3 myscript.py """Textprompt""" """negative prompt""" 10

You could even specify the model to be used. For your use case, you could pack every line into an .sh-file.

And with a prompt writing API, you could even pick up the prompts.

Honestly, I asked chatGPT to write me my user defined script, it was just in need of the documentation for the API of webui.