Skip to content

Commit 994d6d8

Browse files
author
weijinqian_v1
committed
[Refactor] add fia_v3 attention & remove other attention operator.
Signed-off-by: weijinqian_v1 <[email protected]>
1 parent af59fa2 commit 994d6d8

File tree

1 file changed

+2
-2
lines changed

1 file changed

+2
-2
lines changed

vllm_ascend/attention/attention_v1.py

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -560,8 +560,8 @@ def forward(
560560
value = value.contiguous()
561561

562562
if self.attn_type == AttentionType.ENCODER_ONLY:
563-
attn_output = self._forward_encode(query, key, value, attn_metadata,
564-
output)
563+
attn_output = self._forward_encode(query, key, value,
564+
attn_metadata, output)
565565
output[:num_tokens] = attn_output[:num_tokens]
566566
return output
567567

0 commit comments

Comments
 (0)