What is the impact of multimodal intermediate-task training (e.g., image-text alignment tasks like CLIP) on zero-shot
Description
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 multimodal intermediate-task training (e.g., image-text alignment tasks like CLIP) on zero-shot cross-lingual performance compared to text-only intermediate tasks, evaluated using M3IQA accuracy and downstream task throughput?
Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 7.5/10.
Notes
Files
paper.pdf
Files
(86.1 kB)
| Name | Size | Download all |
|---|---|---|
|
md5:add2f4b756d8cafc5ae68af04f48bbce
|
86.1 kB | Preview Download |