squall/flash-attention
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
main
flash-attention/csrc
History
rocking 88d1657a14
[AMD ROCm] Fix KVcache bug and improve performance (#1328) 
* update ck

* update ck

* update ck again

* update ck

* use pointer as seed and offset

* update CK

* Remove useless "else"

* Fix page-attn block table read out-of-bound

---------

Co-authored-by: Po Yen, Chen <PoYen.Chen@amd.com>
2024-11-12 11:32:11 -08:00
..
composable_kernel@13332998a4 [AMD ROCm] Fix KVcache bug and improve performance (#1328) 2024-11-12 11:32:11 -08:00
cutlass@756c351b49 [FA3] BF16 forward 2024-07-14 23:39:46 -07:00
flash_attn Add custom ops for compatibility with PT Compile (#1139) 2024-09-17 19:49:26 -07:00
flash_attn_ck [AMD ROCm] Fix KVcache bug and improve performance (#1328) 2024-11-12 11:32:11 -08:00
ft_attention Make nvcc threads configurable via environment variable (#885) 2024-03-13 20:46:57 -07:00
fused_dense_lib Make nvcc threads configurable via environment variable (#885) 2024-03-13 20:46:57 -07:00
fused_softmax Make nvcc threads configurable via environment variable (#885) 2024-03-13 20:46:57 -07:00
layer_norm Make nvcc threads configurable via environment variable (#885) 2024-03-13 20:46:57 -07:00
rotary Make nvcc threads configurable via environment variable (#885) 2024-03-13 20:46:57 -07:00
xentropy Make nvcc threads configurable via environment variable (#885) 2024-03-13 20:46:57 -07:00