A script to caption images using Florence-2-large. Script is presented as is and will probably not work out of the box for everyone. Requirements are probably NOT complete. Might help some get the right idea though. If somebody wants to finish it, feel free to.
It will generate a .txt file for jpegs, pngs and gifs. Every .txt file will have the same name as file.
How to use:
- Create venv
- Install requirements e.g. pip install -r requirements.txt (they might be incomplete)
- Change path to your image files in make_captions.py
- Run python make_captions.py
Cheers