Implements dual-chunk-flash-attn backend for dual chunk attention with sparse attention support (#11844)
This commit is contained in:
@@ -230,6 +230,7 @@ set(VLLM_EXT_SRC
|
||||
"csrc/attention/paged_attention_v1.cu"
|
||||
"csrc/attention/paged_attention_v2.cu"
|
||||
"csrc/attention/merge_attn_states.cu"
|
||||
"csrc/attention/vertical_slash_index.cu"
|
||||
"csrc/pos_encoding_kernels.cu"
|
||||
"csrc/activation_kernels.cu"
|
||||
"csrc/layernorm_kernels.cu"
|
||||
|
||||
Reference in New Issue
Block a user