Unlocking Parameter-Efficient Fine-Tuning for Low-Resource Language Translation

TL;DR

This paper evaluates various parameter-efficient fine-tuning methods for low-resource language translation, demonstrating that several outperform baseline models and identifying the Houlsby+Inversion adapter as the most effective overall.

Contribution

It provides a comprehensive empirical comparison of 8 PEFT methods across multiple low-resource language domains and sizes, highlighting the best-performing architectures.

Findings

6 PEFT architectures outperform baseline models

Houlsby+Inversion adapter achieves the best overall performance

PEFT methods significantly improve translation accuracy in low-resource settings

Abstract

Parameter-efficient fine-tuning (PEFT) methods are increasingly vital in adapting large-scale pre-trained language models for diverse tasks, offering a balance between adaptability and computational efficiency. They are important in Low-Resource Language (LRL) Neural Machine Translation (NMT) to enhance translation accuracy with minimal resources. However, their practical effectiveness varies significantly across different languages. We conducted comprehensive empirical experiments with varying LRL domains and sizes to evaluate the performance of 8 PEFT methods with in total of 15 architectures using the SacreBLEU score. We showed that 6 PEFT architectures outperform the baseline for both in-domain and out-domain tests and the Houlsby+Inversion adapter has the best performance overall, proving the effectiveness of PEFT methods.