I am trying to insert a series of JPEGs into a Word document using python-docx but it seems that some of them may have non-UTF-8 metadata included, which is causing docx to issu