印度发射BharatGen号卫星,这是印度第一个支持22种语言和多种数据类型的主权AI模型。 India launches BharatGen, its first sovereign AI model supporting 22 languages and multiple data types.

联合部长Jitendra Singh(Jitendra Singh)宣布BharatGen是印度第一个主权、多语言和多式AI大语言模式, Union Minister Jitendra Singh announced BharatGen, India’s first sovereign, multilingual, and multimodal AI large language model, developed at IIT Bombay with ₹235 crore funding. 该项目是国家跨学科网络-物理系统特派团的一部分,为超过22种印度语言提供支持,并整合了文字、言论和文件愿景。 The project, part of the National Mission on Interdisciplinary Cyber-Physical Systems, supports over 22 Indian languages and integrates text, speech, and document vision. 它包括诸如Param-1、Shrutam、Sooktam和Patram等基础模型,得到Bharat Data Sagar(国家数据倡议)的支持。 It includes foundational models like Param-1, Shrutam, Sooktam, and Patram, backed by Bharat Data Sagar, a national data initiative. 该部长强调向私营部门领导过渡,并强调了在农业、保健和治理方面现实世界的应用,例如Krishi Sathi和Docbodh。 The minister emphasized transitioning to private sector leadership and highlighted real-world applications in agriculture, healthcare, and governance, such as Krishi Sathi and Docbodh.