Training Language Models for Long-Span Cross-Sentence Evaluation.

Kazuki Irie, Albert Zeyer, Ralf Schlter, Hermann Ney

Browse the full ASRU paper archive.