Scaling Intermediate-Task Benchmarks for mT5-XL 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: Does scaling the size of intermediate-task benchmarks (e.g., increasing the number of tasks or examples) improve the zero-shot cross-lingual transfer performance of mT5-XL on XTREME, measured by F1 score differences?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 9.0/10.