News
AI Summary
28 Sept 202617 Rabiʻ II 1448 AH
Huawei Open-Sources openPangu-2.0 Pretrain, SFT and RL Training Code on Ascend

Huawei Open-Sources openPangu-2.0 Pretrain, SFT and RL Training Code on Ascend

On September 28, 2026, Huawei announced the release of training code for the openPangu-2.0 model, including pretraining and reinforcement learning components. This launch enhances developers' ability to reproduce and extend the training toolkit associated with the model. The new code includes mechanisms such as supervised fine-tuning and reinforcement learning, providing developers access to Ascend-native tools. The openPangu-2.0-Pro model features approximately 505 billion parameters, while the openPangu-2.0-Flash model has about 92 billion parameters.

Follow these topics

Sign in to follow the topics that matter to you

Sign in to follow

This summary is generated with AI and receives periodic editorial review. Refer to the original source for full details.

0
0 reading now

Insight Score

Rate to unlock

Sign in to react, rate, and save. Sign In