Arabic-OCR-final-project

the dataset and pretraining model and the model weight in this Link