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

Thanks to CTOnews.com netizens Wu Yanzu in South China for the delivery of clues! CTOnews.com news on June 6, at the Wenxin Big Model Technology Exchange (Chengdu), Baidu Smart Cloud launched the "Comate" coding intelligent recommendation tool (CTOnews.com Note: Comate is taken from Coding Mate, meaning everyone's AI coding partner), and officially opened the invitation for testing.

Hou Zhenyu, vice president of ▲ Comate "coding Intelligent recommendation" Multi-line Code recommendation Baidu Group, said that the "Comate" code assistant is a new generation of coding aids based on large models and has conducted a large number of tests within Baidu. The test results show that nearly 50% of the suggested code written with the aid of "Comate" has been adopted by developers.

According to Baidu insiders, auxiliary code writing is the main product function of the first stage. In the second stage, "Comate" will achieve natural language code generation in specific areas and scenarios, and the third stage will achieve full-domain natural language development.

According to reports, Comate supports more than 20 mainstream R & D languages and frameworks, such as Java, C / C++, Python, Go, and JavaScript, and supports mainstream IDE, such as VSCode and Jetbrains series, with the ability to write a single line, generate complete logic blocks, select multiple items, and generate code comments. At the same time, Comate also combines Baidu technology selection and coding specifications, which can not only help engineers complete coding tasks faster, but also have better code quality.

In the future, Comate will also launch the capabilities of code understanding and description, annotation generation, unit test generation, defect identification and automatic repair, code refactoring optimization, automatic pile insertion Debug, etc., as well as natural language generation of code and instructions and technical knowledge question and answer through Chat interaction, as well as the ability to generate applications or products directly in some scenarios, reducing the threshold of software development. Improve the quality and efficiency of R & D code.

Baidu Intelligent Cloud announced at a technology exchange conference in Chengdu that its reasoning performance of "one word" has improved tenfold in the two months since internal testing began in March.

At the same time, based on the complete tool chain provided by the Wenxin Qianfan model platform, in the enterprise application high-frequency core scenarios, the performance of the high-performance mode "Wenxin-Turbo" reasoning service has been improved by 50 times, which will greatly improve the model use effect and efficiency of customers.

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.