I am trying to parallelize BERT\'s embedding because I have a very large dataset I want to use it on. However, when using the map functions from Python\'s
map
