This tool uses the Eagle-X5-7B model from NVIDIA to generate keyword-based captions for images in an input folder. Special thanks to NVIDIA for training this powerful model.
It's a fast and robust captioning model that produces comma-separated keyword outputs.