Huawei is set to unveil its highly anticipated Pangu AI model at the upcoming Huawei Developer Conference 2023 (Cloud) on July 7th. This new model is expected to bring significant technological advancements to the cloud community.

Huawei Developer Conference 2023

Clear Positioning: Pangu Model’s Focus on Industry Applications

During the event, several keynote speakers will address the audience. Zhang Pingan, Executive Director of Huawei and CEO of Huawei Cloud will deliver a keynote speech titled “Everything is a Service, AI Reshape Thousands of Industries.” Additionally, You Peng, Director of Huawei Cloud EI Service Product Department, Tian Qi, Chief Scientist of Huawei Cloud Artificial Intelligence, and Yao Jun, Director of Huawei Noah’s Ark Laboratory, will share insights on the Pangu model.

Huawei Developer Conference speakers

The Pangu model will be positioned as “born for the industry,” marking a clear focus on industry-specific applications. Huawei will also announce the full-stack innovation of the Pangu basic model, along with the introduction of the “refining technique” for the industry model.

Notably, the Pangu AI model boasts impressive specifications. It is the first Chinese pre-training large model with 100 billion parameters in the field of natural language processing (NLP). Furthermore, the computer vision (CV) large model reaches a groundbreaking milestone of 3 billion parameters, enabling both discrimination and generation capabilities. The CV large model also demonstrates exceptional small sample learning ability on ImageNet, making it a pioneering achievement in the industry.

Moreover, Huawei’s Pangu meteorological large model provides second-level weather forecasts, while Zidong Taichu stands as the world’s first map, text, and audio three-modal large model.

The unveiling of the Pangu AI model at the Huawei Developer Conference 2023 (Cloud) is expected to bring forth significant advancements in AI technology, setting the stage for enhanced applications across various industries.

RELATED:

(Via 1,2)