Kazakh Language Gets its National LLM with a Groundbreaking Partnership of Kazakh Research Institutions and VEON's QazCode

Kaz-LLM Development and Launch - VEON Ltd, in partnership with the Ministry of Digital, Innovations, and Aerospace Industry of Kazakhstan, ISSAI NU, Beeline Kazakhstan, and Astana Hub, launched the open-source Kazakh-language large language model (Kaz-LLM) [1] - Kaz-LLM was developed with over 150 billion tokens and is capable of interacting in Kazakh, Turkish, English, and Russian [2] - The model has two versions with 8 billion and 70 billion parameters, aimed at accelerating AI-powered product and service adoption in Kazakhstan [2] - Kaz-LLM has been published on the Hugging Face platform ahead of its full launch [2] VEON's Role and Strategy - VEON's QazCode is the only private sector partner in the Kaz-LLM initiative, aligning with VEON's mission to provide augmented intelligence tools for low-resource languages [3] - Kazakhstan is VEON's most advanced market in terms of augmented intelligence capabilities [3] - VEON aims to bridge the linguistic gap in AI, ensuring equal opportunities for the half a billion people in its markets [5] - VEON Group CEO highlighted the potential of AI to enhance human skills in various sectors, including healthcare, education, and agriculture [5] Beeline Kazakhstan and QazCode Contributions - Beeline Kazakhstan and QazCode have already launched several AI products, including the open-source Kaz-RoBERTA-conversational model with 2 billion parameters [7] - Kaz-RoBERTA-conversational is used for customer service interactions on Beeline Kazakhstan's digital platforms and has been downloaded over three thousand times on Hugging Face [7] - QazCode, with a team of 700 including 350 developers, focuses on telecoms process automation, gamification, entertainment, and IT productivity using AI [11] - Beeline Kazakhstan serves 11 million mobile customers and 2 million fixed internet customers, with a digital ecosystem of 60 products and 11.6 million monthly active users as of June 2024 [10] International Collaboration and Expertise - Beeline Kazakhstan and QazCode contribute to international know-how on LLM development for low-resource languages, collaborating with GSMA Foundry and the Barcelona Supercomputing Center [8] - QazCode shares expertise on LLM development, including the Catalan language model developed by the Barcelona Supercomputing Center [8] ISSAI and National AI Ecosystem - ISSAI, founded in September 2019, drives AI research and innovation in Kazakhstan, collaborating with national and international partners in education, industry, and government [9] - The launch of Kaz-LLM is a pivotal step in developing Kazakhstan's AI ecosystem, reflecting the country's commitment to innovation and technological progress [4]