Abstract
Curriculum learning has shown promising improvements in multiple domains by training machine learning models from easy samples to hard ones. Previous works which either design rules or train models for scoring the difficulty highly rely on task-specific expertise, and cannot generalize. Inspired by the “easy-to-hard” intuition, we propose to do in-sample curriculum learning for natural language generation tasks. Our learning strategy starts training the model to generate the last few words, i.e., do sequence completion, and gradually extends to generate the whole output sequence. Comprehensive experiments show that it generalizes well to different tasks and achieves significant improvements over strong baselines.- Anthology ID:
- 2023.acl-long.666
- Volume:
- Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
- Month:
- July
- Year:
- 2023
- Address:
- Toronto, Canada
- Editors:
- Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
- Venue:
- ACL
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 11937–11950
- Language:
- URL:
- https://aclanthology.org/2023.acl-long.666
- DOI:
- 10.18653/v1/2023.acl-long.666
- Cite (ACL):
- Qi Jia, Yizhu Liu, Haifeng Tang, and Kenny Zhu. 2023. In-sample Curriculum Learning by Sequence Completion for Natural Language Generation. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 11937–11950, Toronto, Canada. Association for Computational Linguistics.
- Cite (Informal):
- In-sample Curriculum Learning by Sequence Completion for Natural Language Generation (Jia et al., ACL 2023)
- PDF:
- https://preview.aclanthology.org/add_acl24_videos/2023.acl-long.666.pdf