Huawei open-sources openPangu-2.0 training, SFT and RL code

Huawei's Ascend Tribe has released pretraining, supervised fine-tuning and reinforcement-learning post-training code for its openPangu-2.0 model family. The projects openPangu-2.0-Training and openPangu-2.0-RL target Huawei's Ascend-based training ecosystem.

• openPangu-2.0-Training and openPangu-2.0-RL code released
• Training covers pretraining and SFT; RL covers post-training
• Code targets Huawei's Ascend training ecosystem