Impact of Incremental Intermediate-Task Training on mT5-3B Zero-Shot Cross-Lingual Transfer 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 impact of incremental intermediate-task training (adding intermediate tasks sequentially) on the zero-shot cross-lingual transfer performance of mT5-3B on XTREME-R, measured in accuracy, and how does it compare to single-task intermediate training?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.0/10.