To help developers better understand large-scale language models (LLMs) globally, the company funded and published professional translations of the Large-Scale Multi-Task Language Understanding (MMLU) benchmark in 14 languages through Hugging Face.
“Helping (developers) understand the unique cultural, economic, and social dynamics of their communities helps ensure that AI applications are tailored to local needs,” the company explained.
The MMLU benchmark is available in multiple languages including Arabic, Bengali, Chinese, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Spanish, Swahili, and Yoruba.
