deepseek-ai

deepseek-ai / FlashMLA

#14

12,6741,045+1 todayC++

FlashMLA: Efficient Multi-head Latent Attention Kernels

📊 Project Info

Language: C++
Stars: ⭐ 12,674
Forks: 1,045
Today: +1
Ranking: #14
Collection: Language
Trending Date: May 30, 2026
Last Push: 4/30/2026