On May 20th, WAVE SUMMIT 2021 Deep Learning Developer SUMMIT, co-hosted by National Engineering Laboratory for Deep Learning Technology and Applications and Baidu, was held in Beijing. Wang Haifeng, chief technology officer of Baidu and director of the National Engineering Laboratory for Deep Learning Technology and Applications, delivered a speech at the summit. Wang Haifeng revealed that the flying paddle has gathered 3.2 million developers, compared with a year ago, an increase of nearly 70%; The number of enterprises and institutions served reached 120,000. Wang Haifeng also shared two thoughts on ai technology and industrial development at the present stage:

First, integrated innovation. From the perspective of technology, the combination of knowledge and deep learning breaks through the deep semantic understanding enhanced by knowledge. Multi-technology integration innovation, through the knowledge graph correlation across modal information, using natural language semantics to represent the semantic space of different modes such as language, voice, vision, breakthrough across modal semantic understanding. From the perspective of platform, deep learning platform and chip hardware integration innovation, to meet the diverse needs of different computing power, power consumption, delay and so on, to achieve the best effect of AI applications. Aiming at the environment where many kinds of chips coexist, the heterogeneous parameter server training technology is developed to break through the problem of efficient training of super-large model. At present, Feoar has carried out adaptation and joint optimization with 22 domestic and foreign hardware manufacturers, and the hardware ecosystem is booming. From the perspective of industry, ARTIFICIAL intelligence technology is more and more deeply integrated with the industry. Driven by industrial demand, AI technology and platform capabilities are continuously polished and innovated with application scenarios.

Second, lower the threshold. With the penetration of artificial intelligence technology in various industries, it is crucial to continuously reduce the threshold for different application scenarios and efficiently meet the needs of different developers. Fly blade was due to the depth of the industrial practice learning open source open platform, has been committed to reduce the threshold, both support movement unified core framework, there are industrial grade model library components, development kit and tools, and improve enterprise efficiency fly blade enterprise edition, etc., to meet different industries, different stages, the needs of different levels of developers. Based on the flyblade platform, 3.2 million developers no longer need to write artificial intelligence algorithm code from scratch, and can efficiently carry out technological innovation and business expansion. The substantial reduction of the threshold has accelerated the diversification and scale of artificial intelligence applications and accelerated the process of industrial intelligence.

At this summit, Baidu analyzed the realization path of AI industrial mass production under the trend of integration innovation. In the application process of AI, enterprises should start from AI pioneers to explore the way, and after completing model verification and generating benefits, enterprises should set up AI teams and enter the application stage of AI workshops. When enterprises develop a large number of AI applications, When multi-person and multi-task cooperative AI production is carried out, it is the large-scale production of AI industry. The flying Oar team explained in detail the new release and important upgrades of the flying oar deep learning platform, including: the latest open source framework of flying oar, which makes development more flexible and convenient; Large scale image retrieval engine, supporting terabyte distributed image storage and retrieval; ERNIE pre-training model, further in the knowledge enhanced semantic understanding, cross modal semantic understanding and other breakthroughs, empower developers, as well as reasoning deployment navigation chart, help developers get through the AI application “the last kilometer”, and so on. At the same time, Flying oar will jointly launch the great sailing plan of flying oar with academia and industry to help AI talent training, industrial intelligent upgrading and frontier exploration. In addition, the partners of flying OARS shared their industrial innovation practices and joint talent training plans. Feoar continues to lead the development of deep learning technology and the integration of science and technology innovation, while constantly lowering the application threshold, accelerating the large-scale production of AI industry, and promoting the process of industrial intelligence.