david/flash-attention

zrkadlo https://github.com/Dao-AILab/flash-attention

Autor	SHA1 Správa	Dátum
Tri Dao	1e3208566a Tune tile sizes for compilation	1 mesiac pred
Tri Dao	df5fe55264 Change tile sizes for Sm8x to reduce stack frame	1 mesiac pred
Tri Dao	5f525322ec Only pass sm_90a compile flag to Sm90 kernels, same w Sm89 kernels	1 mesiac pred
Tri Dao	180ff782dd Template for Sm86	1 mesiac pred
Tri Dao	7a802796e1 Big refactor and update	1 mesiac pred