Computational Efficiency and Performance Trade-offs in Intermediate-Task Fine-Tuning for Zero-Shot Cross-Lingual Transfer

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: How does the computational efficiency (training time, inference speed) of intermediate-task fine-tuning scale with model size (1B to 50B+ parameters) when applied to zero-shot cross-lingual transfer, and what is the trade-off between efficiency and performance gains on the XTREME benchmark?

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