These is a custom dataset made using various programs such as LibreOffice and GIMP used for my Jupyter Notebook binders here on Kaggle and at Binder.
Currently, the dataset is composed of images used for optical character recognition (OCR) for example notebooks of different OCR engines and packages.