Gradio Demo for Donut, an instance of VisionEncoderDecoderModel fine-tuned on SROI (document parsing & information extraction).
To use it, simply upload your image and click 'submit', or click one of the examples to load them.
Output: extracts [date, company, total] from the document.