Simple docker container that exposes a gradio interface for automatic invoice processing, using docQuery.
Simply pull the repo and run docker compose up -d
Alternatively pip install -r requirements.txt
and python src/main.py
(make sure you have tesseract-ocr and poppler-utils installed)
A Gradio interface will be running on localhost:7860.
On the first run it should take about 30s to pull the model. After that inference takes about ~1s (on a 3090 GPU).