Using BUFFET, we perform thoroughĮvaluations of state-of-the-art multilingual large language models withĭifferent transfer methods, namely in-context learning and fine-tuning. BUFFET is designed to establish a rigorousĪnd equitable evaluation framework for few-shot cross-lingual transfer across aīroad range of tasks and languages. Introduce a new benchmark, called BUFFET, which unifies 15 diverse tasks acrossĥ4 languages in a sequence-to-sequence format and provides a fixed set ofįew-shot examples and instructions. To facilitate research on few-shot cross-lingual transfer, we Language processing, most models are developed and evaluated primarily inĮnglish. Download a PDF of the paper titled BUFFET: Benchmarking Large Language Models for Few-shot Cross-lingual Transfer, by Akari Asai and 8 other authors Download PDF Abstract: Despite remarkable advancements in few-shot generalization in natural
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |