Meta supports building high-quality open-source Vietnamese datasets

This morning (March 14), Meta Group and the National Innovation Center (NIC) launched the 2025 Innovation Challenge Program, marking the third year of implementing the program with the goal of promoting the development of artificial intelligence (AI) in Vietnam.

Creating a high-quality open-source Vietnamese dataset

This year’s program focuses on the ViGen Project, which strives to create a high-quality open-source Vietnamese dataset to train, evaluate, and thereby improve the performance of large language models (LLMs), helping AI models better understand Vietnamese culture, context, and expressions. This project is expected to increase the presence of Vietnamese in the development of AI; at the same time, contribute to promoting the rapid and sustainable development of the digital economy.

The Innovation Challenge 2025 focuses on the ViGen Project with an effort to create a high-quality open-source Vietnamese dataset.

Mr. Vo Xuan Hoai, Deputy Director of NIC, said that AI is transforming the world. Therefore, developing large-scale, high-quality, open-source Vietnamese datasets for AI training and evaluation has become an urgent priority. The ViGen project is in line with Resolution 57 of the Politburo to promote breakthroughs in science, technology, innovation, and national digital transformation.

"The ViGen project requires joint efforts from policymakers, research teams, researchers, developers, experts, and users. Together, we will turn AI into a powerful tool for every Vietnamese person and make Vietnam an AI powerhouse in the world," said Mr. Hoai.

Mr. Sarim Aziz, Director of Public Policy at Meta shared that 2025 is a pivotal year for the development of AI in Vietnam. Meta is honored to cooperate with the National Innovation Center and AI for Vietnam Foundation with the ViGen Project.

"By supporting Vietnamese researchers, developers and businesses to access open-source AI models such as Llama and resources suitable for the local context, we hope to help them maximize their potential and promote highly applicable AI initiatives that are suitable for Vietnam's needs," said Director Sarim Aziz.

Professor Yann LeCun, Vice President, Chief Artificial Intelligence Architect of Meta Group, who is known as one of the fathers of AI, assessed that the ViGen project and the Vietnam Innovation Challenge not only aim to promote technology, but also aim to build a comprehensive AI future, honoring and integrating Vietnam's unique cultural and linguistic heritage.

Meanwhile, Mr. Chris Lewin, Director of AI & Data Capabilities, Asia Pacific, Deloitte said that as Asia Pacific undergoes a strong digital transformation, open source AI is becoming an important driving force in shaping the future of the public sector. "We are pleased to partner with Meta to equip agencies in this field with highly practical recommendations on applying next-generation AI (Gen AI), based on the principles of transparency and trust in AI," said Mr. Chris Lewin.

Unlock your potential with LIama

Alongside the ViGen Project, Meta and Deloitte also released a public sector handbook titled "Innovating in the Asia-Pacific Public Sector with Open Source AI: Unlocking Disruptive Potential with LIama".

The ViGen project is expected to increase the presence of Vietnamese in the development of Al.

This handbook will provide a roadmap for government agencies to effectively integrate open source AI technology. The combination of these initiatives aims to build a synchronized AI strategy in Vietnam, bringing benefits to both the public and private sectors.

According to Mr. Tran Viet Hung, Founder & CEO, Al for Vietnam Organization, the ViGen project will contribute to the community large and high-quality datasets in Vietnamese to improve the current status of Vietnamese as a language with a very modest presence in Al. The ViGen project also demonstrates the power and value of open source models such as Llama, allowing the development of innovative solutions with global impact that take into account the context of the Vietnamese language.

The ViGen project originated from a tripartite collaboration between Meta Group, NIC and Al for Vietnam; in which, NIC plays the role of the project owner, coordinator, ensuring the project is consistent with Vietnam's national goals. Al for Vietnam is the project implementation partner with technical and financial support from Meta Group. Strategic partners include NVIDIA, Viettel, and the Vietnam Academy of Science and Technology.

The mission of VIC 2025 - ViGen project is to make AI models support Vietnamese naturally and comprehensively from the core to unlock the potential of AI applications in Vietnam.

ViGen will build large-scale, high-quality open-source Vietnamese datasets to train and evaluate the capabilities of AI models. The ViGen project also contributes to ensuring that AI development in Vietnam is consistent with Vietnamese cultural values ​​and ethical standards, aiming to build a locally relevant and responsible open-source AI ecosystem.

To support the project, Meta will contribute open-source datasets from its AI and Data for Public Interest program, including insights on mobility and social connectivity, as well as training data from AI-powered population maps.

Source: https://www.baogiaothong.vn/meta-ho-tro-xay-dung-bo-du-lieu-tieng-viet-ma-nguon-mo-chat-luong-cao-192250314115403619.htm