Commit History

Author SHA1 Message Date
  Tri Dao 82c1aa3514 Move PackGQA epilogue code to pack_gqa.h 1 month ago
  Tri Dao df96486c31 Decode: varlen, paged KV, leftpad 1 month ago
  Ying Zhang 3669b25206 bwd benchmark + small fixes (#1129) 4 months ago
  Ying Zhang c7f20a2d31 add cudnn benchmark for var-len 4 months ago
  Ying Zhang dfe1a59e4b Add var-seq-len to FA3 fp16 / bf16 fwd (#1072) 4 months ago