\item{cutoff_lower}{Cutoff value for minimum cosine similarity above which documents are considered duplicates (inclusive)}
\item{cutoff_upper}{Cutoff value for maximum cosine similarity, above which documents are not considered duplicates (for debugging and manual parameter tuning, inclusive)}
\item{words}{Document cutoff point in number of words. Documents are cut off at the last [.?!] before the cutoff (so document will be a little shorter than [words])}
dupe_objects.json and data frame containing each id and all its duplicates. remove_ids.txt and character vector with list of ids to be removed. Files are in current working directory