The Quicksign OCRized Text Dataset is a collection of more than 400,000 labeled text files that have been extracted from real documents using optical character recognition (OCR). Each file is ...