Lius: Translation Model Based Instructional Lingustic Using Continual Instruction Tuning In Kupang Malay
For researchers working on low-resource language translation, this approach reduces reliance on large-scale parallel data, though it is incremental as it adapts existing instruction tuning to a specific language.
The paper tackles performance degradation of LLMs in low-resource language translation by proposing Lius, a model fine-tuned on Kupang Malay using Continual Instruction Tuning. Lius outperforms standard instruction-tuned models by 4-6 points and NMT/multilingual LLMs by 10-13 points on several metrics.
Large Language Models (LLMs) offer new potential for translation tasks but often experience performance degradation when handling low-resource languages. To address this limitation, we propose an approach for fine-tuning LLMs on a low-resource language, Kupang Malay. Our approach involves designing a set of instructions by leveraging explicit lexical and semantic features from a bilingual dictionary, and introducing Continual Instruction Tuning (CIT), a training paradigm that enables iterative instruction-based training. Experimental results demonstrate that our model, named Lius, yields notable improvements over standard instruction-tuned models by outperforming 4-6 points, and surpassing both Neural Machine Translation (NMT) and Multilingual LLM models by 10-13 points on several evaluation metrics. These findings highlight the potential of our approach to mitigate the reliance on large-scale parallel data in low-resource language translation.