background
With the widespread application of artificial intelligence technology in recent years, various fields and industries have explored how to use AI to assist production and development. Software developers are also working hard to improve the development experience and use AI capabilities for coding, debugging, testing and review. Github Copilot, Codeium, Amazon CodeWhisperer, etc. are all outstanding representatives in this field. In order to satisfy developers' continuous pursuit of high-quality development experience, major manufacturers have set their sights on the field of AI code assistants. Recently, Tencent Cloud AI Code Assistant was announced at the Tencent Global Digital Ecosystem Conference . Since this product is not yet open to the public, in this article today , let us learn more about it based on the information disclosed at the conference and the introduction on the official website . Magic bar.
introduce
Tencent Cloud AI Code Assistant is a programming assistance tool announced at the Tencent Global Digital Ecology Conference on September 7, 2023. For convenience of description, it is hereinafter referred to as the AI Code Assistant. It was the first to be integrated into Tencent Cloud's Cloud Studio and Microsoft's VS Code. It can program online, assist development, and use AI to code in the IDE. The core value is to release mechanical work and improve development productivity. It can play an important role in developers' daily development iterations. Specific performance includes communication, coding, troubleshooting, review, and testing. Let's take a look at how it works in these business scenarios.
communicate
After installing the AI code assistant plug-in, you can conduct technical communication with AI in the IDE. Stop switching, stop traditional search engines, and chat with the AI directly in the IDE to code. You can ask it questions directly. Please use JavaScript to implement a bubble sort algorithm. If you encounter code that you don't understand, you can let it help you answer it. It has learned and mastered more than 20 programming languages and massive development paradigms. There is a demonstration video below for you to experience its charm.
coding
In addition to answering programming questions, it also performs well in code completion and automatic generation of project files.
It supports more than 20 kinds of syntax deep learning, supports annotation description completion, variable completion, function completion, etc., and can automatically generate README, Git configuration, scaffolding and other files according to the directory structure of the project. The video below shows its code completion, which is very fast.
Troubleshooting
In daily development work, we often need to debug a certain program and need to track a variable or method. The AI code assistant also performs very well in this regard. It can give reasonable suggestions when project compilation errors occur, so that users do not have to rely on Error code Go to search engines to find answers. Use it to find code problems with one click and push fixes.
Review
The AI code assistant can generate submission information with one click based on the code to be submitted when submitting code , and can also do proofreading before MR with one click, which is very helpful for team review.
test
After the business logic code is written, you can select a function and enter /tests (tests) in the input box, and the plug-in will automatically generate test cases for you.
Advantages and Highlights
- Code capabilities based on Tencent's Hunyuan large model: Tencent Cloud AI Assistant is based on Tencent's independently developed Hunyuan large model and has powerful natural language processing and understanding capabilities. This enables the AI assistant to more accurately understand the user's needs and provide smarter and more humane services.
- Polishing Tencent’s internal massive business data: Tencent Cloud AI Assistant was designed with full consideration of Tencent’s internal massive business data. These data have been deeply mined and analyzed, allowing the AI assistant to better understand and handle various complex business scenarios.
- Covering the entire development process from communication to testing: Tencent Cloud AI Assistant not only provides a rich API interface, but also provides a complete set of development tool chains, including code editors, debugging tools, version control, etc., to help developers meet their needs Efficiently manage the entire process from analysis, design, development to testing.
- Supports multiple programming languages, mainstream development frameworks, and commonly used IDEs: Tencent Cloud AI Assistant supports multiple programming languages, such as Python, Java, C++, etc., as well as mainstream development frameworks such as TensorFlow, PyTorch, etc., and is also compatible with commonly used IDEs such as VSCode , IDEA, Eclipse, etc., provide developers with flexible choices.
- Open and compatible, supporting enterprise development and personalized construction: Tencent Cloud AI Assistant will provide an open API interface, allowing developers to carry out personalized customization and secondary development according to their own business needs. It is also compatible with various hardware devices and operating systems, providing Enterprises offer highly flexible solutions.
Conclusion
Tencent Cloud has carried out internal testing for several months, and tens of thousands of people use the AI code assistant for development every day. They continue to tune the model and optimize interactions based on user feedback, especially for completion scenarios that are very practical in production, and various indicators have gradually improved. I believe that after using the AI code assistant, users' development capabilities can be greatly improved.
Interested friends can also learn more about Tencent Global Digital Ecology Conference: https://qdrl.qq.com/PIrEk5OT