Huawei announces open-sourcing of Pangu models to accelerate AI application, value creation

Chinese technology company Huawei announced on Monday the open sourcing of its Pangu models and model inference technologies based on Ascend, an initiative the company described as another key step toward advancing research and innovation in large model technologies and accelerating the application and value creation of artificial intelligence (AI) across industries.

According to the statement posted on its official website, Huawei has officially announced the open-sourcing of its Pangu dense model with 7 billion parameters, the Pangu Pro MoE (Mixture-of-Experts) model with 72 billion parameters, and its model inference technology based on Ascend, which serves as the platform for AI infrastructure.

The weights of the Pangu Pro MoE 72B model, the basic inference code, and large-scale MoE model inference code based on Ascend have been officially released on the open-source platform. The model weights and inference code for the Pangu 7B will be launched on the open-source platform soon, per the statement.

The company invited developers, corporate partners and researchers around the world to download and use it, offer feedback and improve it together, the statement said.

Huawei said that this move is another key step in implementing its Ascend ecosystem strategy, aimed at promoting research and innovation in large model technologies and accelerating the application and value creation of AI across various industries.