在今天下午的OPPO Find X7系列产品技术沟通会上，OPPO公布了安第斯大模型AndesGPT的全新升级。据悉，安第斯大模型将分为1800亿参数、700亿参数、70亿参数三种体量，支持端云协同部署，可以根据不同场景的用户需求实现智能调度。
此次OPPO Find X7系列手机首搭70亿参数AndesGPT端侧大模型，将为用户提供更为智能的服务。此外，OPPO还通过4bit量化的模型压缩、Al Boost推理引擎优化，以及与芯片厂家联合深度优化等多项举措，实现了AndesGPT的大规模端侧部署。
据了解，AndesGPT是安第斯实验室研发的一款基于人工智能的语言生成模型，其强大的计算能力可以更好地理解和处理自然语言，提供更加准确、高效的人机交互体验。
作为一款拥有先进技术的智能手机，OPPO Find X7系列将给用户带来全新的使用体验。未来，OPPO将继续致力于人工智能技术的研发和应用，为用户带来更多的惊喜。
英语如下：
News Title: “OPPO Find X7 Series Launches 70 Billion Parameter AndesGPT Edge Model, Breeno Assistant Upgraded!”
Keywords: OPPO Find X7, AndesGPT Large Model, Edge Applications
News Content: The title reads: “OPPO Find X7 Series Launches 70 Billion Parameter AndesGPT Edge Model”
At today’s OPPO Find X7 series product technology communication meeting, OPPO announced the new upgrade of the Andes large model AndesGPT. It is reported that the Andes large model will be divided into three sizes with 1.8 trillion parameters, 700 billion parameters, and 70 billion parameters, supporting edge-cloud collaborative deployment, and can achieve intelligent scheduling based on different scene user needs.
This time, the OPPO Find X7 series phones were launched with the 70 billion parameter AndesGPT edge model, which will provide more intelligent services for users. In addition, OPPO has realized the large-scale edge deployment of AndesGPT through multiple measures such as 4-bit quantized model compression, Al Boost inference engine optimization, and deep joint optimization with chip manufacturers.
It is understood that AndesGPT is an artificial intelligence language generation model developed by the Andes Laboratory, with powerful computing capabilities that can better understand and process natural language, providing more accurate and efficient human-computer interaction experience.
As a smartphone with advanced technology, the OPPO Find X7 series will bring new experiences to users. In the future, OPPO will continue to focus on the research and application of artificial intelligence technology, bringing more surprises to users.
【来源】https://www.ithome.com/0/741/887.htm
Views: 1