Por Qu\'e N\~ao Utiliser Alla Spr{\aa}k? Mixed Training with Gradient Optimization in Few-Shot Cross-Lingual Transfer

TL;DR

This paper introduces a novel mixed training approach with gradient optimization for few-shot cross-lingual transfer, outperforming traditional target-adapting methods across multiple NLP tasks and languages.

Contribution

It proposes a one-step mixed training method using stochastic gradient surgery, enabling a single model to handle multiple target languages simultaneously without large development sets.

Findings

Achieves state-of-the-art results on 4 NLP tasks across 48 languages.

Significantly outperforms target-adapting methods, especially for distant languages.

Demonstrates effectiveness without development sets and reduces overfitting.

Abstract

The current state-of-the-art for few-shot cross-lingual transfer learning first trains on abundant labeled data in the source language and then fine-tunes with a few examples on the target language, termed target-adapting. Though this has been demonstrated to work on a variety of tasks, in this paper we show some deficiencies of this approach and propose a one-step mixed training method that trains on both source and target data with \textit{stochastic gradient surgery}, a novel gradient-level optimization. Unlike the previous studies that focus on one language at a time when target-adapting, we use one model to handle all target languages simultaneously to avoid excessively language-specific models. Moreover, we discuss the unreality of utilizing large target development sets for model selection in previous literature. We further show that our method is both development-free for target languages, and is also able to escape from overfitting issues. We conduct a large-scale experiment on 4 diverse NLP tasks across up to 48 languages. Our proposed method achieves state-of-the-art performance on all tasks and outperforms target-adapting by a large margin, especially for languages that are linguistically distant from the source language, e.g., 7.36% F1 absolute gain on average for the NER task, up to 17.60% on Punjabi.