david/flash-attention

miroir de https://github.com/Dao-AILab/flash-attention

Auteur	SHA1 Message	Date
Ying Zhang	496fdc4f6c Add seqused_q in fwd / bwd and seqused_k in bwd.	il y a 5 mois
Cameron Shinn	3cea2fb6ee Add ArchTag to pre/postprocess bwd kernels (#1180)	il y a 5 mois
Tri Dao	bafe253042 [FA3] Bwd	il y a 6 mois