How can I properly parallelise the following for loops? (Python)

Question

A little bit of context first. I have around 10.000 pdf files. I want to convert each page of these files to a JPEG file and then OCR them using the Google Vision API. I know it