FlashMLA/csrc/flash_kernel
zhan3916 30f7561775 Performance: Add new tile for small bs
Signed-off-by: zhan3916 <zhan3916@gmail.com>
2025-03-25 11:54:33 +08:00
..
feature Performance improvement 2025-03-07 16:44:52 +08:00
flash_fwd_split_kernel.h Performance: Add new tile for small bs 2025-03-25 11:54:33 +08:00
flash_fwd_split_kernel_k64_16x16_4waves.h Performance: Add new tile for small bs 2025-03-25 11:54:33 +08:00
flash_fwd_split_kernel_k64_32x16_4waves.h Fix: fix large input length offset overflow 2025-03-19 17:29:06 +08:00
flash_fwd_split_kernel_k64_64x16_8waves.h Fix: fix large input length offset overflow 2025-03-19 17:29:06 +08:00
flash_fwd_split_kernel_k64_V1x8.h Performance improvement 2025-03-07 16:44:52 +08:00
kernel_traits.h Performance: Add new tile for small bs 2025-03-25 11:54:33 +08:00