Hybrid Batch Training for Zero-Shot Cross-Lingual Transfer in TyDi-QA and PAWS-X

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

Research goal: Does hybrid batch training (mixing English and non-English data) during intermediate-task fine-tuning improve zero-shot cross-lingual transfer performance on TyDi-QA and PAWS-X compared to English-only training?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 9.2/10.