Abstract
Scarcity of parallel data causes formality style transfer models to have scarce success in preserving content. We show that fine-tuning pre-trained language (GPT-2) and sequence-to-sequence (BART) models boosts content preservation, and that this is possible even with limited amounts of parallel data. Augmenting these models with rewards that target style and content –the two core aspects of the task– we achieve a new state-of-the-art.
Original language | English |
---|---|
Title of host publication | Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing |
Editors | Chengqing Zong, Fei Xia, Wenjie Li, Roberto Navigli |
Place of Publication | Bangkok, Thailand |
Publisher | Association for Computational Linguistics, ACL Anthology |
Pages | 484-494 |
Number of pages | 11 |
Volume | 2 |
DOIs | |
Publication status | Published - 2021 |