Large model training

On the morning of September 19th, the Chinese-language Internet base 3.0 was officially released to the public at the National Cyber Security Awareness Week, 2025, Forum on Artificial Intelligence Security Governance in Kunming. Under the guidance of the relevant departments of the Central Network Office, the China Cyberspace Safety Association, in conjunction with the National Internet Emergency Response Centre and others, has continued to build on the synergy of business, higher education and scientific research units, building on the language sharing mechanism established by the Networking Networking Association ' s Ad Hoc Committee on Artificial Intelligence and Security, bringing together a new set of high-quality..

Recently, Moore Thread and the full-discipline education AI big model "Shizhe AI" jointly announced that the two parties have completed the big model training test. Relying on Moore Thread's KUAE Qianka Intelligent Computing Cluster, Shizhe AI successfully completed the high-intensity training of the 7 billion parameter big model in one week, and the training efficiency reached the expected level, fully demonstrating the capabilities of the domestic full-featured GPU Qianka 100 billion training platform. Source Note: The image is generated by AI, and the image is authorized by Midjourney. It is understood that "Shizhe AI" was established in 2020. The core team comes from Tsinghua University and focuses on the full-discipline education big model...