Home Verkennen Help
Registreren Inloggen
david
/
flash-attention
spiegel van https://github.com/Dao-AILab/flash-attention
1
0
Vork 0
Bestanden Issues 0 Wiki
Aftakking: ipiszy/local_attn
Aftakkingen Labels
flash-attention
/
csrc
rocking e2182cc21d Support page kvcache in AMD ROCm (#1198) 4 maanden geleden
..
composable_kernel @ a9b170b541 e2182cc21d Support page kvcache in AMD ROCm (#1198) 4 maanden geleden
cutlass @ 756c351b49 74b0761ff7 [FA3] BF16 forward 6 maanden geleden
flash_attn 65f723bb9a Split bwd into more .cu files to speed up compilation 5 maanden geleden
flash_attn_ck e2182cc21d Support page kvcache in AMD ROCm (#1198) 4 maanden geleden
ft_attention 50896ec574 Make nvcc threads configurable via environment variable (#885) 10 maanden geleden
fused_dense_lib 50896ec574 Make nvcc threads configurable via environment variable (#885) 10 maanden geleden
fused_softmax 50896ec574 Make nvcc threads configurable via environment variable (#885) 10 maanden geleden
layer_norm 50896ec574 Make nvcc threads configurable via environment variable (#885) 10 maanden geleden
rotary 50896ec574 Make nvcc threads configurable via environment variable (#885) 10 maanden geleden
xentropy 50896ec574 Make nvcc threads configurable via environment variable (#885) 10 maanden geleden