david/flash-attention

mirror da https://github.com/Dao-AILab/flash-attention

Autore	SHA1 Messaggio	Data
Tri Dao	7f5d73a162 Add env var to disable specific hdim	1 mese fa
Tri Dao	3e5d77a102 Group instantiations for different hdims together	1 mese fa
Tri Dao	fc2fd95a18 Renable FP8 kernels	2 mesi fa
Tri Dao	64d92bce53 Split PagedKV into separate .cu files to speed up compilation	2 mesi fa
Tri Dao	586ba914bb Move fwd tile size to a separate file	2 mesi fa
Tri Dao	018b9af683 Move .cu files to instantiations, use generate_kernels.py	2 mesi fa