dont worry kitten
We're open-sourcing Mixture-of-Kittens (MoK), our MoE training megakernel for NVL72s. It fuses all Mixture-of-Experts communication and computation into a singl...
AI 摘要
别担心，小猫咪 [引用 @cursor_ai]：我们正在开源 Mixture-of-Kittens (MoK)，这是我们面向 NVL72 的 MoE 训练超内核。 它将所有混合专家通信与计算融合到单个完全确定性的内核中，运行速度比最强公开基线快达 2.37 倍。
dont worry kitten
We're open-sourcing Mixture-of-Kittens (MoK), our MoE training megakernel for NVL72s. It fuses all Mixture-of-Experts communication and computation into a singl...