Language and Task Arithmetic with Parameter-Efficient Layers for Zero-Shot Summarization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chronopoulou, Alexandra, Pfeiffer, Jonas, Maynez, Joshua, Wang, Xinyi, Ruder, Sebastian, Agrawal, Priyanka
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929536571015168
author Chronopoulou, Alexandra
Pfeiffer, Jonas
Maynez, Joshua
Wang, Xinyi
Ruder, Sebastian
Agrawal, Priyanka
author_facet Chronopoulou, Alexandra
Pfeiffer, Jonas
Maynez, Joshua
Wang, Xinyi
Ruder, Sebastian
Agrawal, Priyanka
contents Parameter-efficient fine-tuning (PEFT) using labeled task data can significantly improve the performance of large language models (LLMs) on the downstream task. However, there are 7000 languages in the world and many of these languages lack labeled data for real-world language generation tasks. In this paper, we propose to improve zero-shot cross-lingual transfer by composing language or task specialized parameters. Our method composes language and task PEFT modules via element-wise arithmetic operations to leverage unlabeled data and English labeled data. We extend our approach to cases where labeled data from more languages is available and propose to arithmetically compose PEFT modules trained on languages related to the target. Empirical results on summarization demonstrate that our method is an effective strategy that obtains consistent gains using minimal training of PEFT modules.
format Preprint
id arxiv_https___arxiv_org_abs_2311_09344
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Language and Task Arithmetic with Parameter-Efficient Layers for Zero-Shot Summarization
Chronopoulou, Alexandra
Pfeiffer, Jonas
Maynez, Joshua
Wang, Xinyi
Ruder, Sebastian
Agrawal, Priyanka
Computation and Language
Parameter-efficient fine-tuning (PEFT) using labeled task data can significantly improve the performance of large language models (LLMs) on the downstream task. However, there are 7000 languages in the world and many of these languages lack labeled data for real-world language generation tasks. In this paper, we propose to improve zero-shot cross-lingual transfer by composing language or task specialized parameters. Our method composes language and task PEFT modules via element-wise arithmetic operations to leverage unlabeled data and English labeled data. We extend our approach to cases where labeled data from more languages is available and propose to arithmetically compose PEFT modules trained on languages related to the target. Empirical results on summarization demonstrate that our method is an effective strategy that obtains consistent gains using minimal training of PEFT modules.
title Language and Task Arithmetic with Parameter-Efficient Layers for Zero-Shot Summarization
topic Computation and Language
url https://arxiv.org/abs/2311.09344