1 paper · 1 atlas page
FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
Dao, Fu, Ermon, Rudra et al. · NeurIPS 2022 2022