Domovská stránka Prehľadávať Pomoc
Registrovať Prihlásiť sa
david
/
flash-attention
zrkadlo https://github.com/Dao-AILab/flash-attention
1
0
Fork 0
Súbory Issues 0 Wiki
Branch: changes_for_fp8
Branche Tagy
flash-attention
/
csrc
Tri Dao 65f723bb9a Split bwd into more .cu files to speed up compilation 4 mesiacov pred
..
composable_kernel @ 8182976c37 d8f104e97a Support AMD ROCm on FlashAttention 2 (#1010) 4 mesiacov pred
cutlass @ 756c351b49 74b0761ff7 [FA3] BF16 forward 5 mesiacov pred
flash_attn 65f723bb9a Split bwd into more .cu files to speed up compilation 4 mesiacov pred
flash_attn_ck d8f104e97a Support AMD ROCm on FlashAttention 2 (#1010) 4 mesiacov pred
ft_attention 50896ec574 Make nvcc threads configurable via environment variable (#885) 9 mesiacov pred
fused_dense_lib 50896ec574 Make nvcc threads configurable via environment variable (#885) 9 mesiacov pred
fused_softmax 50896ec574 Make nvcc threads configurable via environment variable (#885) 9 mesiacov pred
layer_norm 50896ec574 Make nvcc threads configurable via environment variable (#885) 9 mesiacov pred
rotary 50896ec574 Make nvcc threads configurable via environment variable (#885) 9 mesiacov pred
xentropy 50896ec574 Make nvcc threads configurable via environment variable (#885) 9 mesiacov pred