Inicio Explorar Ayuda
Registro Iniciar sesión
david
/
flash-attention
espejo de https://github.com/Dao-AILab/flash-attention
1
0
Fork 0
Archivos Incidencias 0 Wiki
Rama: ipiszy/used_q
Ramas Etiquetas
flash-attention
/
csrc
Tri Dao 65f723bb9a Split bwd into more .cu files to speed up compilation hace 6 meses
..
composable_kernel @ 8182976c37 d8f104e97a Support AMD ROCm on FlashAttention 2 (#1010) hace 6 meses
cutlass @ 756c351b49 74b0761ff7 [FA3] BF16 forward hace 6 meses
flash_attn 65f723bb9a Split bwd into more .cu files to speed up compilation hace 6 meses
flash_attn_ck d8f104e97a Support AMD ROCm on FlashAttention 2 (#1010) hace 6 meses
ft_attention 50896ec574 Make nvcc threads configurable via environment variable (#885) hace 10 meses
fused_dense_lib 50896ec574 Make nvcc threads configurable via environment variable (#885) hace 10 meses
fused_softmax 50896ec574 Make nvcc threads configurable via environment variable (#885) hace 10 meses
layer_norm 50896ec574 Make nvcc threads configurable via environment variable (#885) hace 10 meses
rotary 50896ec574 Make nvcc threads configurable via environment variable (#885) hace 10 meses
xentropy 50896ec574 Make nvcc threads configurable via environment variable (#885) hace 10 meses