Meta 在俄亥俄州 New Albany 的吉瓦级超算集群 Prometheus 上训练了 Muse 系列基础模型，包括 Muse Wonder。该集群拥有数十万块 GPU，预训练 goodput 达 85%+，强化学习 goodput 达 90%+，Muse Wonder 训练期间平均每约 47K GPU 小时出现一次中断。
原文给出 Prometheus 超算集群在预训练与强化学习上的 goodput 和中断间隔数据，可作为大规模训练基础设施的参考。
We trained our Muse foundational models, including Muse Wonder, at Prometheus, our gigawatt-scale supercluster in New Albany, Ohio.
Across a supercluster with hundreds of thousands of GPUs, even one underperforming host can slow an entire training run. We engineered Prometheus for resilience at scale, achieving 85%+ goodput on pretraining, 90%+ on reinforcement learning and ~47K GPU hours mean time between interruptions during Muse Wonder training.
Take a look at the tech powering Prometheus: https://go.meta.me/4yXXApZ
来源：AI at Meta · x.com