Can I read NusaMT-7B: Low-Resource Indonesian Translation on EtoBox?

NusaMT-7B: Low-Resource Indonesian Translation by aalghurro009 is a document available to read on EtoBox.

What is NusaMT-7B: Low-Resource Indonesian Translation about?

The document presents NusaMT-7B, a machine translation model designed for low-resource Indonesian languages, specifically Balinese and Minangkabau, utilizing a pre-trained LLaMA2-7B. It integrates techniques such as continued pre-training, supervised fine-tuning, and data cleaning to improve translation quality, achieving superior results in the FLORES-200 benchmark compared to state-of-the-art models. The findings highlight the potential of fine-tuned large language models to enhance translation for endang

Author

aalghurro009

Language

EN