My goal is to sentencize documents which have already been tokenized and matched with gold annotations.
I have a dataset of annotated documents in the classic spacy f
