China Telecom Artificial Intelligence Technology Co., Ltd. andChina TelecomTeleAI has released a newStar Model・Software Factory (hereinafter referred to as Xingchen Software Factory) and claimed that this isThe first groundbreaking AI development tool in China.
Xingchen Software Factory uses the latest multi-modal large model technology.Automatically generate front-end and back-end code, and conductSelf-testing and debugging, forming the software developmentFully automatic production line. Users only need to write down the product documents, and the big model can generate code based on the document content, and deploy and launch it in the software factory with one click.
Xingchen Big Model Software Factory has built three core capabilities from development process innovation, controllable code generation, and interactive mode upgrade:
-
1. A new collaborative process for document development, using project documentation that is closer to everyday language as the core of software development, which allows non-software developers to participate.
-
2. Self-developed multimodal code generation model to decouple the relationships between code modules and achieve stable generation from project documents to project code.
-
3. In the new interactive mode of large-model programming, users only need to care about product requirements and description document writing, and follow the step-by-step instructions of the software factory to complete software development, testing and deployment.
According to reports, China Telecom has accumulated more than 500TB of text data, 1.2 billion pictures and text data, and PB-level video data, and has jointly built large models covering more than 20 industries such as education, government affairs, and emergency response with the leading ecosystem.Covering more than 500 application scenarios across the industry.
In May, China Telecom released the Xingchen multi-dialect speech recognition model, which supports the simultaneous recognition and understanding of more than 30 dialects including Cantonese, Shanghainese, Sichuanese, Wenzhounese, etc.It is the largest speech recognition model in China that supports the most dialects., to solve the problem that information services cannot reach the elderly, old people, children, remote and poor areas.