Facebook's super AI, Llama (Large Language Model Meta AI), will be standardized into Vietnamese through the ViGen project, an initiative implemented by Meta in collaboration with the National Innovation Center (NIC).
ViGen aims to build a high-quality open-source Vietnamese language database, helping AI models better understand culture, context and expressions in Vietnamese.
On the morning of March 14, Meta Group and NIC officially launched the Innovation Challenge 2025, focusing on the ViGen project.
This project aims to ensure AI models supporting Vietnamese naturally and comprehensively right from the core, creating a foundation for the development of AI applications in Vietnam.
ViGen will build a large-scale open-source dat set, contributing to enhancing the presence of Vietnamese in AI, while promoting the digital economy.
Meta is committed to contributing open source databases, including mobility data, social connectivity, and AI-powered demographic maps.
At the same time, Meta and Deloitte also announced the handbook "Innovation in the public sector in Asia - Pacific with open source AI: Exploiting breakthrough potential with Llama".
This handbook helps government agencies effectively integrate open source AI technology, creating a synchronous foundation for AI development strategies in Vietnam.
According to Mr. Sarim Aziz - Director of Public Policy at Meta, supporting researchers and businesses to access open source AI such as Llama will help them promote their potential and promote highly applicable AI initiatives.
Meanwhile, Mr. Tran Viet Hung - Founder & CEO of AI for Vietnam - emphasized that ViGen will provide high-quality Vietnamese data sets to improve the current situation of Vietnamese, which is considered less present in the AI field.
The ViGen project not only contributes to standardizing Vietnamese in AI but also lays the foundation for an open source AI ecosystem, in line with the context and ethical standards in Vietnam.