Researchers at Swiss universities are set to release a groundbreaking large language model (LLM) this summer, trained on the “Alps” supercomputer at the Swiss National Supercomputing Centre (CSCS). This open-source AI model boasts multilingual capabilities, supporting over 1,500 languages, and represents a significant advancement in transparent AI development. The initiative, launched by ETH Zurich and EPFL, emphasizes collaboration among international organizations focused on trustworthy AI. The LLM, currently in its final testing phase, aims to enhance accessibility and accountability in various sectors, including scientific and governmental fields. Unlike commercial models primarily developed in the U.S. and China, this model prioritizes linguistic diversity, with approximately 60% of its training data in English and 40% in other languages, fostering global applicability. Full documentation will accompany the model’s release, ensuring transparency in its architecture and training processes, thereby promoting continuous development and compliance with regulatory standards.
Source link
Share
Read more