Extending Input Contexts of Language Models through Training on Segmented Sequences.

Petros Karypis, Julian J. McAuley, George Karypis

Browse the full NAACL paper archive.