Abu Dhabi-based artificial intelligence (AI) technology company Microsoft-backed G42 on Tuesday unveiled a Hindi large language model (LLM) for India. The AI model called Nanda is said to be trained with a huge amount of datasets in both Hindi, English and Hinglish languages. The generic AI model was developed in collaboration with MBZUAI and Cerebras Systems. At this time, the company has not announced use cases for the LLM. Last year, the AI firm released the Jais AI model that was trained on Arabic and English languages.
G42 unveils a Hindi macro language model
In a post on X (formerly known as Twitter), G42 India CEO Manu Kumar Jain unveiled the Hindi language generative AI model Nanda. It was unveiled at the UAE-India Business Forum in Mumbai in the presence of Crown Prince of Abu Dhabi Sheikh Khalid bin Mohammed bin Zayed Al Nahyan and Union Minister of Commerce and Industry Piyush Goyal. The LLM is named after Nanda Devi, the second highest mountain peak in India.
Jain also highlighted some specialties of LLM. According to the post, NANDA is a 13-billion-parameter model that was trained on approximately 2.13 trillion tokens of the language dataset, including Hindi, English, and Hinglish. The CEO also said that Nanda is a bilingual model and is also proficient in Hinglish, a hybrid language that blends Romanized letters and Devanagari pronunciation.
As of now, 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 use. Jain said, “We really believe that Nanda can integrate into the fabric of India. It meets all sovereign requirements and can help take large-scale technology initiatives to new heights. This will provide an opportunity to more than half a billion Hindi language speakers to harness the potential of Generative AI.
Last year, the AI firm launched Jais, an open-source Arabic LLM with Arabic natural language processing (NLP) capabilities. Several AI models were released with up to 70 billion parameters. In April, Microsoft became a major backer and invested $1.5 billion (about Rs 12,600 crore) in the company. Windows 10 also took up space on the G42’s board.
