Awarri unveils Nigeria’s first multilingual language model to boost AI

Awarri has unveiled Nigeria’s first Multilingual Language Model (LLM) to boost Artificial Intelligence technology in Africa.

The innovative approach aims to build the largest dataset of native Nigerian languages, laying the foundation for the development of AI models and applications tailored to Nigeria’s linguistic diversity.

Developed in collaboration with the National Information Technology Development Agency (NITDA), the National Centre for Artificial Intelligence and Robotics (NCAIR), and data.org, this initiative marks a significant milestone in Nigeria’s AI landscape.

The LLM, powered by Awarri’s Artificial Intelligence technology, is designed to be trained in five indigenous Nigerian languages.

Currently supporting five of Nigeria’s most spoken languages – Yoruba, Igbo, Hausa, Pidgin, and Ibibio – alongside accented English, the platform invites users to contribute by translating English sentences into their local language through audio recordings.

Awarri plans to expand its language coverage to include more Nigerian languages in the future.

Silas Adekunle, CEO of Awarri, emphasised the significance of this launch for Nigeria’s AI development.

“A first for Nigeria, the launch is a critical step in the development of AI in Africa. One of our goals with the LLM is to demonstrate the key role technology can play in persevering Nigeria’s many cultures,” Adekunle said.

Adekunle further highlighted the importance of data in AI development, noting that the initiative will enable the collection of new datasets in indigenous Nigerian languages, empowering the nation to develop AI tools with a profound understanding of its cultural and linguistic intricacies.

“Data is central to the development of all AI models and applications. The launch will enable us to capture entirely new data sets in indigenous Nigerian languages allowing us as a nation to build AI tools with a deep understanding of Nigeria’s cultural and linguistic nuances.”