Autoregressive Structured Prediction with Language Models

Benchmark Model Rank Results
coreference-resolution-on-conll-2012ASP+T0-3B#4Avg F1: 82.3
coreference-resolution-on-ontonotesASP+T0-3B#3F1: 82.3
named-entity-recognition-ner-on-conll-2003ASP+T5-3B#5F1: 94.1
named-entity-recognition-ner-on-conll-2003ASP+flan-T5-large#11F1: 93.8
relation-extraction-on-ace-2005ASP+T5-3B#3RE Micro F1: 72.7RE+ Micro F1: 70.5NER Micro F1: 91.3
relation-extraction-on-conll04ASP+T0-3B#9RE+ Micro F1: 76.3NER Micro F1: 90.3