Thanks for your great contribution to releasing a high-quality dataset for document understanding. I have tried to download part of ccpdf as you mentioned in README, but I didn't find OCR information of pdfs (whether by DjVu-based or ocr tools ), which makes it inconvenient to leverage this dataset. Could you release the ocr data publicly?
Thanks for your great contribution to releasing a high-quality dataset for document understanding. I have tried to download part of ccpdf as you mentioned in README, but I didn't find OCR information of pdfs (whether by DjVu-based or ocr tools ), which makes it inconvenient to leverage this dataset. Could you release the ocr data publicly?