Mumbai, February 21:  Reliance Jio-backed mega group BharatGPT has set the stage to launch its AI model in India in March 2024. SML (Seetha Mahalaxmi Healthcare, in partnership with BharatGPT ecosystem, led by the IIT Bombay, have reportedly unveiled a series of Indic LLMs called 'Hanooman'. The report said that these models are trained in 22 Indian language; however will support 11 local languages.

The Hanooman AI model will reportedly help India advance in the accelerating race to develop transformative AI technology. According to the report by Financial Express, during an event, a teaser about the AI model gave was dropped, giving the audience a sneak peek at how it would work. The report highlighted that a Tamilian bike mechanic was shown asking a question in his native language, and a banker talked in Hindi with an AI bot. PM Narendra Modi Has a Dream That AI Startups Should Come From Tier 2 and Tier 3 Cities, Says MoS IT Rajeev Chandrasekhar.

The report further said that the developer was also seen writing computer codes using the AI model. The Hanooman AI model backed by Reliance will reportedly be functional in nearly 11 local languages however could be expanded in future. The report said that the Hanooman AI model will focus on four major areas - governance, education, health care and financial services.

According to the report by Times of India, the size of the family of these LLMs ranges up to 40 billion parameters. The report further highlighted that the first four large language models included the following parameters - 1.5 billion, 7 billion, 13 billion and 40 billion. These Hanooman AI LLMs will be introduced as open-source models accessible to all. The report said that the models will be released in March 2024. Bharti Airtel in Partnership With Ericsson Successfully Demonstrated mmWave 5G Functionality on Airtel Network.

Hanooman AI Capabilities

The Times of India report said that the Hanooman AI model will have multimodal AI capabilities such as generating text-to-text, text-to-video, text-to-speech and vice versa. The report highlighted major challenges faced by the creators in building the large Indian language models, including sourcing quality datasets in the form of text, video or audio. The model will reportedly compete with other LLMs such as Kurtrim by Ola, the Airavata model developed by IIT-Madras and OpenHathi developed by Sarvam AI.

(The above story first appeared on LatestLY on Feb 21, 2024 04:04 PM IST. For more news and updates on politics, world, sports, entertainment and lifestyle, log on to our website latestly.com).