Abstract
Task-adaptive pre-training (TAPT) and Self-training (ST) have emerged as the major semi-supervised approaches to improve natural language understanding (NLU) tasks with massive amount of unlabeled data. However, it’s unclear whether they learn similar representations or they can be effectively combined. In this paper, we show that TAPT and ST can be complementary with simple TFS protocol by following TAPT -> Finetuning -> Self-training (TFS) process. Experimental results show that TFS protocol can effectively utilize unlabeled data to achieve strong combined gains consistently across six datasets covering sentiment classification, paraphrase identification, natural language inference, named entity recognition and dialogue slot classification. We investigate various semi-supervised settings and consistently show that gains from TAPT and ST can be strongly additive by following TFS procedure. We hope that TFS could serve as an important semi-supervised baseline for future NLP studies.- Anthology ID:
- 2021.findings-emnlp.86
- Volume:
- Findings of the Association for Computational Linguistics: EMNLP 2021
- Month:
- November
- Year:
- 2021
- Address:
- Punta Cana, Dominican Republic
- Editors:
- Marie-Francine Moens, Xuanjing Huang, Lucia Specia, Scott Wen-tau Yih
- Venue:
- Findings
- SIG:
- SIGDAT
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 1006–1015
- Language:
- URL:
- https://aclanthology.org/2021.findings-emnlp.86
- DOI:
- 10.18653/v1/2021.findings-emnlp.86
- Cite (ACL):
- Shiyang Li, Semih Yavuz, Wenhu Chen, and Xifeng Yan. 2021. Task-adaptive Pre-training and Self-training are Complementary for Natural Language Understanding. In Findings of the Association for Computational Linguistics: EMNLP 2021, pages 1006–1015, Punta Cana, Dominican Republic. Association for Computational Linguistics.
- Cite (Informal):
- Task-adaptive Pre-training and Self-training are Complementary for Natural Language Understanding (Li et al., Findings 2021)
- PDF:
- https://preview.aclanthology.org/proper-vol2-ingestion/2021.findings-emnlp.86.pdf
- Data
- CoNLL 2003, GLUE, MultiNLI, QNLI, SST, SST-2