Trade-off Analysis of Training Time and Zero-shot Cross-lingual NER Performance in XLM-R-large on XTREME-R

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 trade-off between intermediate-task training time and zero-shot cross-lingual NER performance of XLM-R-large on XTREME-R, quantified by comparing F1 scores against training throughput (samples per second) across different task combinations?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.7/10.