Microsoft-Backed G42 Introduces Nanda, a Hindi Large Language Model for India

Nanda AI model was introduced at the UAE-India Business Forum in Mumbai.

Written by Akash Dutta, Edited by Siddharth Suvarna | Updated: 11 September 2024 14:54 IST

Microsoft-Backed G42 Introduces Nanda, a Hindi Large Language Model for India

Photo Credit: X/Manu Kumar Jain

G42 India CEO said the LLM was named after Indian mountain peak Nanda Devi

Highlights

Nanda is a 13-billion-parameter AI model
G42’s Nanda is trained on 2.13 trillion tokens of language datasets
Microsoft-backed G42 said the LLM knows English, Hindi, and Hinglish

Microsoft-backed G42, an Abu Dhabi-based artificial intelligence (AI) technology company, unveiled a Hindi large language model (LLM) for India on Tuesday. Dubbed Nanda, the AI model is said to be trained with large volumes of datasets in both Hindi, English, and Hinglish languages. The generative AI model was developed in collaboration with MBZUAI and Cerebras Systems. Currently, the company has not announced the use cases for the LLM. Last year, the AI firm released the Jais AI model which was trained on Arabic and English languages.

G42 Unveils a Hindi Large Language Model

In a post on X (formerly known as Twitter), G42 India CEO Manu Kumar Jain unveiled Nanda, the Hindi language generative AI model. It was unveiled at the UAE-India Business Forum in Mumbai in the presence of the Crown Prince of Abu Dhabi Sheikh Khaled bin Mohamed bin Zayed Al Nahyan and the Union Minister of Commerce and Industry Piyush Goyal. The LLM is named after the second-highest mountain peak in India, Nanda Devi.

Jain also highlighted some of the specifications of the LLM. According to the post, Nanda is a 13-billion-parameter model which was trained on approximately 2.13 trillion tokens of language datasets including Hindi, English and Hinglish. The CEO also added that Nanda is a bi-lingual model and is also proficient in Hinglish, the hybrid language that blends Romanised letters and Devanagari pronunciation.

OpenAI ‘Strawberry’ AI Model Might Be Launched Before End of September

Currently, the company has not shared any release timeline for the AI model. It is not known whether Nanda will be available in the public domain or it will be reserved for government usage. Jain said, “We genuinely believe that #Nanda can get integrated into the fabric of India. It meets all the sovereign requirements and can help take large-scale technology initiatives to newer heights. It will offer over half a billion Hindi language speakers the opportunity to harness the potential of generative AI.”

Last year, the AI firm launched Jais, an open-source Arabic LLM with capabilities for Arabic natural language processing (NLP). Multiple AI models were released with up to 70 billion parameters. In April, Microsoft invested $1.5 billion (roughly 12,600 crores) in the company becoming a major backer. The Windows maker also took a seat on G42's board.

Comments

For the latest tech news and reviews, follow Gadgets 360 on X, Facebook, WhatsApp, Threads and Google News. For the latest videos on gadgets and tech, subscribe to our YouTube channel. If you want to know everything about top influencers, follow our in-house Who'sThat360 on Instagram and YouTube.

Further reading: AI, Artificial Intelligence, LLM, AI Model, India

Akash Dutta Email Akash Dutta

Akash Dutta is a Senior Sub Editor at Gadgets 360. He is particularly interested in the social impact of technological developments and loves reading about emerging fields such as AI, metaverse, and fediverse. In his free time, he can be seen supporting his favourite football club - Chelsea, watching movies and anime, and sharing passionate opinions on food. More