Impact of Multilingual Intermediate-Task Training on Zero-Shot Cross-Lingual Transfer Performance in XNLI
Abstract
Description
<p>Intermediate-task training---fine-tuning a pretrained model on an intermediate task before fine-tuning again on the target task---often improves model performance substantially on language understanding tasks in monolingual English settings. We investigate whether English intermediate-task training is still helpful on non-English target tasks. Using nine intermediate language-understanding tasks, we evaluate intermediate-task transfer in a zero-shot cross-lingual setting on the XTREME benchmark. We see large improvements from intermediate training on the BUCC and Tatoeba sentence retrieval tas</p><p><strong>Research goal:</strong> Does intermediate-task training with multilingual intermediate tasks (instead of English-only) improve zero-shot cross-lingual transfer performance on XNLI compared to English intermediate-task training?</p><p><em>Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 9.3/10.</em></p>
This report was generated autonomously by Assignee Research, an owner-gated autonomous research lab. The content synthesizes findings from peer-reviewed papers. Tribunal score: 9.3/10.
This report was generated autonomously by Assignee Research, an owner-gated autonomous research lab. The content synthesizes findings from peer-reviewed papers. Tribunal score: 9.3/10.
Keywords
intermediate-task , training , multilingual , intermediate , tasks , instead , English-only , improve
Citation
Collections
Endorsement
Review
Supplemented By
Referenced By
Rights and licensing
info:eu-repo/semantics/openAccess
Creative Commons Attribution 4.0 International
https://creativecommons.org/licenses/by/4.0/legalcode
Creative Commons Attribution 4.0 International
https://creativecommons.org/licenses/by/4.0/legalcode