xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-09-04 07:36:59 +08:00
Code Issues Packages Projects Releases Wiki Activity

minor change

Browse Source 
Signed-off-by: Sage Moore <sage@neuralmagic.com>
This commit is contained in:
Sage Moore 2025-07-08 19:12:45 +00:00
parent 90330563c6
commit 376e7eb838
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/v1/attention/backends/flash_attn.py
View File

@ -266,7 +266,7 @@ class FlashAttentionMetadataBuilder(
use_cascade = common_prefix_len > 0 use_cascade = common_prefix_len > 0
if use_cascade: if use_cascade:
cu_prefix_query_lens = torch.tensor([0,num_actual_tokens], cu_prefix_query_lens = torch.tensor([0, num_actual_tokens],
dtype=torch.int32, dtype=torch.int32,
device=self.runner.device) device=self.runner.device)
prefix_kv_lens = torch.tensor([common_prefix_len], prefix_kv_lens = torch.tensor([common_prefix_len],