File tree Expand file tree Collapse file tree 1 file changed +6
-1
lines changed
vllm/v1/attention/backends Expand file tree Collapse file tree 1 file changed +6
-1
lines changed Original file line number Diff line number Diff line change 9
9
from vllm .attention .ops .paged_attn import PagedAttention
10
10
from vllm .attention .ops .prefix_prefill import context_attention_fwd
11
11
from vllm .logger import init_logger
12
- from vllm .v1 .attention .backends .flash_attn import FlashAttentionMetadata
12
+ from vllm .v1 .attention .backends .flash_attn import (
13
+ FlashAttentionMetadata , FlashAttentionMetadataBuilder )
13
14
14
15
logger = init_logger (__name__ )
15
16
@@ -49,6 +50,10 @@ def get_kv_cache_shape(
49
50
def use_cascade_attention (* args , ** kwargs ) -> bool :
50
51
return False
51
52
53
+ @staticmethod
54
+ def get_builder_cls () -> Type ["FlashAttentionMetadataBuilder" ]:
55
+ return FlashAttentionMetadataBuilder
56
+
52
57
53
58
class ROCmAttentionImpl (AttentionImpl ):
54
59
You can’t perform that action at this time.
0 commit comments