Kimi K3 Is Here: Efficient Day-0 Support on vLLM
vLLM delivers day-0 Kimi K3 serving with hybrid KDA prefix caching, DSpark speculative decoding, production-scale disaggregation, and optimized kernels across N
Topic
#LatentMoE
1 篇笔记
Topic
1 篇笔记
vLLM delivers day-0 Kimi K3 serving with hybrid KDA prefix caching, DSpark speculative decoding, production-scale disaggregation, and optimized kernels across N