近日，印度信实工业与印度理工学院孟买分校等机构支持的研究联盟BharatGPT宣布，计划于3月发布首个印度语言大型语言模型服务，该服务以印度教神猴哈努曼的名字命名，即“Hanuman”。这一项目旨在促进印度本土语言的数字化发展，目前该模型已能支持11种印度语言，包括印地语、泰米尔语、泰卢固语、马拉雅拉姆语和马拉地语，研发团队计划未来将支持的语言数量扩展至22种。
BharatGPT表示，该语言模型的研发不仅限于语言处理，还涉及电信和零售等行业的应用模型。信实工业作为该项目的支持者，将利用其在电信和零售领域的强大影响力，推动这一语言模型的商业化应用。此举有望加速印度本土语言在数字领域的普及，同时也为信实工业在相关行业的数字化转型提供技术支持。
印度是一个多语言、多文化的国家，拥有超过22种官方语言和数百种方言。然而，在数字领域，英语一直是主导语言，这限制了非英语母语用户的数字参与。BharatGPT的这一项目，旨在打破这一语言壁垒，让印度本土语言在数字世界中得到更广泛的应用。
随着“Hanuman”模型的发布，印度有望在语言技术领域迈出重要一步，这不仅有助于保护和促进印度本土语言文化的传承，也将为全球多语言处理技术的发展贡献新的视角和经验。
英文标题：India Launches Native Language Model “Hanuman”
英文关键词：Indian Language Model, Reliance Industries, BharatGPT
英文新闻内容：
India’s BharatGPT, backed by Reliance Industries and the Indian Institute of Technology Bombay, has announced plans to launch a large-scale Indian language model service named “Hanuman” next month. The initiative aims to boost the digital presence of Indian languages, with the current model supporting 11 languages including Hindi, Tamil, Telugu, Malayalam, and Marathi. The team behind the project aims to expand this to 22 languages.
The “Hanuman” model is not limited to language processing but also includes industry-specific applications in sectors like telecommunications and retail. Reliance Industries, as a supporter of the project, will leverage its strong presence in these sectors to promote the commercial application of the language model. This move is expected to accelerate the digital adoption of Indian languages and provide technological support for Reliance Industries’ own digital transformation in relevant industries.
India, a country with over 22 official languages and numerous dialects, has been dominated by English in the digital space, limiting the digital participation of non-English speakers. BharatGPT’s initiative aims to break this linguistic barrier and widely apply Indian languages in the digital world.
The launch of the “Hanuman” model marks a significant step for India in the field of language technology. It not only helps in protecting and promoting the heritage of Indian languages but also contributes new perspectives and experiences to the development of multilingual processing technologies globally.
【来源】https://www.jiemian.com/article/10819925.html
Views: 2