Multimodal Capabilities: Hanooman AI Model Supports Text, Speech, and Video Responses, Report Indicates
The unveiling of Hanooman, a groundbreaking series of language models tailored for the Indian demographic, marks a significant milestone in the realm of artificial intelligence (AI). Developed by the Reliance Industries-backed BharatGPT group, led by the esteemed Indian Institute of Technology (IIT) Bombay and seven other IITs, Hanooman represents a collaborative effort to harness AI technology for the benefit of Indian society.
The AI models were showcased at a recent technology event in Mumbai, garnering attention and anticipation from industry experts and stakeholders alike. This unveiling comes hot on the heels of Reliance Jio’s announcement of Jio Brain, an AI and machine learning (ML) platform geared towards enterprises, signaling a broader push towards AI-driven innovation within the Reliance ecosystem.
According to reports by Bloomberg, the consortium behind Hanooman aims to launch the AI models next month, with plans to support 11 Indian regional languages and operate across various sectors, including healthcare, governance, financial services, and education. This strategic approach underscores the versatility and adaptability of the Hanooman series, which has been trained on data relevant to these domains, ensuring its effectiveness and relevance in real-world applications.
Moreover, Hanooman is not merely limited to text-based interactions; it boasts multimodal capabilities, including support for speech and video formats. With speech-to-text functionality and a focus on accessibility, Hanooman aims to cater to a diverse audience, including those who may be unable to read or write, thus democratizing access to AI-driven solutions.
In addition to its broad applicability, Hanooman offers specialized versions tailored to specific sectors, such as VizzhyGPT, a customized variant designed for healthcare applications. Trained on relevant data and equipped with domain-specific knowledge, VizzhyGPT exemplifies the potential of AI to revolutionize healthcare delivery and improve patient outcomes. As Hanooman prepares to make its debut next month, it holds the promise of empowering organizations and individuals across India with advanced AI capabilities, driving innovation, and fostering socioeconomic development in the digital age.
The publication revealed that during the event, a demo video was shown where a mechanic asked queries to a chatbot built using Hanooman in the Tamil language, a bank employee spoke with it in the Hindi language, and a developer used it to write codes.
A separate report by the Times of India touts it to be open source. It highlighted that the first four language models in the Hanooman series, carrying 1.5 billion, seven billion, 13 billion, and 40 billion parameters each, will be available to developers and researchers. It will likely be made available to build specific AI tools such as summarising chatbots, and more. Commercial rights will likely not be given for the models since Reliance Jio is reported to be building specific AI tools based on the language model.