Impact of Intermediate-Task Training Duration on Zero-Shot Cross-Lingual Accuracy and Inference Efficiency in mT5-3B

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 intermediate-task training duration (e.g., 1 vs. 10 epochs) on the trade-off between zero-shot cross-lingual accuracy and inference efficiency of mT5-3B across different XTREME-R tasks?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.5/10.