Scrambled text: training Language Models to correct OCR errors using synthetic data

SyDa
Re-assign community