Multilingual LLM instruction tuning

11 repos

Projects focused on adapting large language models to non-English languages through instruction-following datasets and fine-tuning. The cluster centers on the Alpaca model architecture extended across multiple languages (Spanish, French, Hindi, Chinese, Indonesian, Portuguese), enabling researchers and practitioners to build instruction-tuned models for low-resource and high-resource languages alike. This represents the practical engineering challenge of democratizing LLM capabilities beyond English-dominant training data.