Overcoming Catastrophic Forgetting in Zero-Shot Cross-Lingual Generation
Vu, Tu, Barua, Aditya, Lester, Brian, Cer, Daniel, Iyyer, Mohit, Constant, Noah
–arXiv.org Artificial Intelligence
In this paper, we explore the challenging problem of performing a generative task in a target language when labeled data is only available in English, using summarization as a case study. We assume a strict setting with no access to parallel data or machine translation and find that common transfer learning approaches struggle in this setting, as a generative multilingual model fine-tuned purely on English catastrophically forgets how to generate non-English. Given the recent rise of parameter-efficient adaptation techniques, we conduct the first investigation into how one such method, prompt tuning (Lester et al., 2021), can overcome catastrophic forgetting to enable zero-shot cross-lingual generation. Our experiments show that parameter-efficient prompt tuning provides gains over standard fine-tuning when transferring between less-related languages, e.g., from English to Thai. However, a significant gap still remains between these methods and fully-supervised baselines. To improve cross-lingual transfer further, we explore several approaches, including: (1) mixing in unlabeled multilingual data, and (2) explicitly factoring prompts into recombinable language and task components. Our approaches can provide further quality gains, suggesting that robust zero-shot cross-lingual generation is within reach.
arXiv.org Artificial Intelligence
Oct-23-2022
- Country:
- Africa > Nigeria (0.04)
- North America > United States
- Massachusetts > Hampshire County > Amherst (0.04)
- Europe > Romania
- Asia
- Pakistan (0.04)
- Nepal (0.04)
- Middle East > Israel (0.04)
- India (0.04)
- China > Guangxi Province
- Nanning (0.04)
- Genre:
- Research Report > New Finding (0.46)
- Technology: