Denoising based Sequence-to-Sequence Pre-training for Text Generation

Liang Wang, Wei Zhao, Ruoyu Jia, Sujian Li, Jingming Liu
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
This paper presents a new sequence-tosequence (seq2seq) pre-training method PoDA (Pre-training of Denoising Autoencoders), which learns representations suitable for text generation tasks. Unlike encoder-only (e.g., BERT) or decoder-only (e.g., OpenAI GPT) pre-training approaches, PoDA jointly pretrains both the encoder and decoder by denoising the noise-corrupted text, and it also has the advantage of keeping the network architecture unchanged in the subsequent fine-tuning stage. Meanwhile, we
more » ... esign a hybrid model of Transformer and pointer-generator networks as the backbone architecture for PoDA. We conduct experiments on two text generation tasks: abstractive summarization, and grammatical error correction. Results on four datasets show that PoDA can improve model performance over strong baselines without using any task-specific techniques and significantly speed up convergence. 1
doi:10.18653/v1/d19-1412 dblp:conf/emnlp/WangZJLL19 fatcat:b6w57svbkjhrziz5it6wo472ua