今日下午,OPPO 举办了 Find X7 系列产品技术沟通会,会上官方公布了安第斯大模型 AndesGPT 的全新升级。此次升级涵盖了 1800 亿参数、700 亿参数、70 亿参数三种体量的安第斯大模型,支持端云协同部署,能够根据不同场景的用户需求实现智能调度。
值得一提的是,OPPO 在 Find X7 系列上首次实际应用了 70 亿参数的大模型。通过 4bit 量化的模型压缩,OPPO 的 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,OPPO 成功地将大模型应用于端侧。
此次全新升级的安第斯大模型,将为用户带来更智能、更高效的使用体验。OPPO 持续致力于技术创新,为用户提供更优质的产品和服务。
News content:
At the Find X7 series product technology communication meeting held today, OPPO announced the new upgrade of the AndesGPT large model. The upgrade covers three types of AndesGPT large models with 1800 billion, 700 billion, and 70 billion parameters, supporting end-cloud collaborative deployment and intelligent scheduling according to different user scenarios.
It is worth noting that OPPO has first applied the 70-billion-parameter large model on the Find X7 series. Through 4bit quantitative model compression, Al Boost inference engine optimization, and deep collaboration with chip manufacturers, OPPO has successfully applied the large model to the end side.
The newly upgraded AndesGPT large model will bring a more intelligent and efficient experience to users. OPPO continues to strive for technological innovation to provide users with better products and services.
【来源】https://www.ithome.com/0/741/887.htm
Views: 1