This dataset was generated by reformatting coref-data/gap_raw into the indiscrim coreference format. See that repo for dataset details.
See ianporada/coref-data for additional conversion details and the conversion script.
Please create an issue in the repo above or in this dataset repo for any questions.
