PLUG: Leveraging Pivot Language in Cross-Lingual Instruction Tuning
This addresses the problem of imbalanced language abilities in LLMs for users of lower-resource languages, though it is incremental as it builds on existing instruction tuning methods.
The paper tackles the challenge of applying instruction tuning to lower-resource languages by proposing PLUG, which uses a high-resource pivot language to enhance performance, resulting in a 29% average improvement in instruction-following abilities across four languages.
Instruction tuning has remarkably advanced large language models (LLMs) in understanding and responding to diverse human instructions. Despite the success in high-resource languages, its application in lower-resource ones faces challenges due to the imbalanced foundational abilities of LLMs across different languages, stemming from the uneven language distribution in their pre-training data. To tackle this issue, we propose pivot language guided generation (PLUG), an approach that utilizes a high-resource language, primarily English, as the pivot to enhance instruction tuning in lower-resource languages. It trains the model to first process instructions in the pivot language, and then produce responses in the target language. To evaluate our approach, we introduce a benchmark, X-AlpacaEval, of instructions in 4 languages (Chinese, Korean, Italian, and Spanish), each annotated by professional translators. Our approach demonstrates a significant improvement in the instruction-following abilities of LLMs by 29% on average, compared to directly responding in the target language alone. Further experiments validate the versatility of our approach by employing alternative pivot languages beyond English to assist languages where LLMs exhibit lower proficiency. Our code and data are available at https://github.com/ytyz1307zzh/PLUG.