We ( Ananth, @AlonTalmor, @sameer_ , @nlpmattg) , are pleased to announce the release of ORB, an Open Reading Benchmark. This is an evaluation server that tests a single model on a variety of reading comprehension datasets (SQuAD, DROP, Quoref, ...). leaderboard.allenai.org/orb