DocRED: A Large-Scale Document-Level Relation Extraction Dataset

Benchmark Model Rank Results
relation-extraction-on-docredBiLSTM#47F1: 51.06Ign F1: 44.73
relation-extraction-on-docredDocRED-Context-Aware#48F1: 50.64Ign F1: 43.93
relation-extraction-on-docredBiLSTM#49F1: 50.12Ign F1: 43.60
relation-extraction-on-docredDocRED-CNN#50F1: 42.33Ign F1: 36.44