LongCat-2.0: 1.6 trillion parameter AI model trained on Chinese cluster

Original: The industry's first trillion-parameter model to complete end-to-end training and inference on a 50,000-GPU Chinese computing cluster 👀

Source: x.com ↗