Impact of Cross-Lingual Intermediate-Task Training on Zero-Shot XTREME-R Performance

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: What is the effect of combining intermediate-task training with cross-lingual pretraining objectives (e.g., mT5, XLM-R) on zero-shot performance in XTREME-R compared to standard intermediate-task training on English-only data?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.9/10.