Skip to content

Commit e24b3fa

Browse files
CLFutureXusberkeley
authored andcommitted
[Bugfix][Core] running queue index leakage exception (vllm-project#26754)
Signed-off-by: CLFutureX <chenyongqyl@163.com>
1 parent 7b301ad commit e24b3fa

File tree

1 file changed

+1
-0
lines changed

1 file changed

+1
-0
lines changed

vllm/v1/core/sched/scheduler.py

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -278,6 +278,7 @@ def schedule(self) -> SchedulerOutput:
278278
token_budget += num_scheduled_tokens[preempted_req.request_id]
279279
req_to_new_blocks.pop(preempted_req.request_id)
280280
num_scheduled_tokens.pop(preempted_req.request_id)
281+
req_index -= 1
281282
else:
282283
preempted_req = self.running.pop()
283284

0 commit comments

Comments
 (0)