Computational Efficiency of English vs. Multilingual Intermediate-Task Training for Zero-Shot Cross-Lingual NER 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 computational efficiency (measured in FLOPs and inference latency) of models trained with English intermediate-task training versus multilingual intermediate-task training when deployed for zero-shot cross-lingual named entity recognition on XTREME-R, while maintaining a target macro-F1 score threshold?

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