This is a simple implementation of trocr (transformer based ocr) allowing you to retrieve handwritten and newpaper text
to read more about trocr:
https://arxiv.org/abs/2109.10282
pip install -r requirements.txtto predict handwritten text
python main.py handwrittento predict handwritten text
python main.py handwrittento predict newspaper text
python main.py newspaper