Shulou(Shulou.com)11/24 Report--

Thanks to CTOnews.com netizen Xiao Zhan for the clue delivery! CTOnews.com, April 11, 2023, March 17, 2023, the national standard "Information Technology Neural Network representation and Model Compression part 1: convolution Neural Network" (GB / T 42382.1-2023), jointly compiled by Huawei, Baidu, Peking University, Pengcheng Laboratory and other units, was officially released, becoming the first national standard for artificial intelligence (AI) model representation and compression technology in China. It will be officially implemented on October 1, 2023.

According to reports, the national standard is led by the Yunnao project team led by Professor Tian Yonghong, deputy director of the Network Intelligence Research Department of Pengcheng Laboratory, and is an important achievement of the project team's work in promoting domestic standards in the systematic layout of global standards in the field of AI models.

In December 2017, Yunnao project team and industry core stakeholders officially launched the development of the standard after the concentrated research and development of key technologies such as AI model representation, compression, deployment and so on. At the same time, in terms of the standard promotion strategy, the Yunnao project team adopts the international and domestic "two legs" model to iteratively optimize the standard according to the feedback of different standardization organizations in the development of the standard. The CTOnews.com summary is as follows:

On the international front, the Yunnao project team led the successful release of the "Standard for Artificial Intelligence (AI) Model Representation, Compression, Distribution and Management" standard ("artificial Intelligence (AI) Model representation, Compression, Distribution and Management Standard") at the Institute of Electrical and Electronic Engineers at IEEE last year, which further combines broader opinions in the global field on the basis of previous work. It has formed the first international standard of AI model in the world, which is more adaptable to the international market, and has won two standard awards of IEEE SA. With the gradual deepening of the research, the work of the Yunnao project team has received great attention from the industry, and led the successful establishment of a new generation AI model to represent technical standards in ITU-T, Telecommunication Standards Branch of the International Telecommunication Union last year.

On the domestic side, the Yunnao project team took the lead in issuing the group standard "Information Technology Neural Network representation and Model Compression part 1: convolution Neural Network" in the new generation artificial intelligence industry technology innovation strategic alliance AITISA on October 21, 2021, and carried out iterative optimization on the basis of this group standard, and finally formed the national standard of this release. So far, the layout of the standard system of the first global AI model representation and compression technology has basically taken shape.

According to the current requirements of AI industry in the deployment of resource-constrained devices and AI models, this series of standards define the representation specification of AI model and the distribution, compression and deployment process to meet the requirements of multi-scenario, break down the barriers between computing architecture and algorithm framework, and improve the efficiency of AI model in network distribution, learning and application deployment.

At the same time, some of this series of standards have been widely used in the fields of AI development framework (Baidu), AI chip (Huawei Haisi) and AI open platform (Haekangwei), and the released series of standards have radiated to international AI leading companies such as AMD Xilinx Sailings.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.