david/flash-attention

kopia lustrzana https://github.com/Dao-AILab/flash-attention

Autor	SHA1 Wiadomość	Data
Tri Dao	dfe29f5e2b [Gen] Don't use ft_attention, use flash_attn_with_kvcache instead	1 rok temu
Tri Dao	a01d1213d7 [Gen] Add kernel from FasterTransformer for benchmarking	2 lat temu