Can General-Purpose Large Language Models Generalize to English-Thai Machine Translation ?
This highlights the need for specialized models to maintain performance in low-resource settings, which is incremental as it builds on known limitations of LLMs.
The study tested general-purpose large language models (LLMs) and specialized translation models on English-Thai machine translation and code-switching, finding that LLMs fail under strict computational constraints like 4-bit quantization, while specialized models outperform them with comparable or lower requirements.
Large language models (LLMs) perform well on common tasks but struggle with generalization in low-resource and low-computation settings. We examine this limitation by testing various LLMs and specialized translation models on English-Thai machine translation and code-switching datasets. Our findings reveal that under more strict computational constraints, such as 4-bit quantization, LLMs fail to translate effectively. In contrast, specialized models, with comparable or lower computational requirements, consistently outperform LLMs. This underscores the importance of specialized models for maintaining performance under resource constraints.