biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[V1] Support any head size for FlexAttention backend (#20467)

Browse Source 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
This commit is contained in:
Cyrus Leung
2025-07-07 00:54:36 +08:00
committed by GitHub
parent e202dd2736
commit 9fb52e523a
20 changed files with 202 additions and 118 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
vllm/config.py
View File

@@ -2319,7 +2319,7 @@ class SchedulerConfig:
if self.max_num_batched_tokens > self.max_num_seqs * self.max_model_len:
logger.warning(
"max_num_batched_tokens (%d) exceeds max_num_seqs"
"max_num_batched_tokens (%d) exceeds max_num_seqs "
"* max_model_len (%d). This may lead to unexpected behavior.",
self.max_num_batched_tokens,
self.max_num_seqs * self.max_model_len)