xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-09-06 08:27:00 +08:00
Code Issues Packages Projects Releases Wiki Activity

Update scheduler.py

Browse Source 
Signed-off-by: Pr0Wh1teGivee <calvin_zhu0210@outlook.com>
Signed-off-by: weichen <calvin_zhu0210@outlook.com>
This commit is contained in:
Pr0Wh1teGivee 2025-12-05 17:15:33 +08:00 committed by weichen
parent 779769ea97
commit 6413793466
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
vllm/config/scheduler.py
View File

@ -21,7 +21,11 @@ logger = init_logger(__name__)
RunnerType = Literal["generate", "pooling", "draft"] RunnerType = Literal["generate", "pooling", "draft"]
SchedulerPolicy = Literal["fcfs", "priority", "sjf"] SchedulerPolicy = Literal["fcfs", "priority", "sjf"]
""" SJF Scheduling Policy:
It stands for shortest-job-first — requests are scheduled by total prompt +
output length (shorter first), with aging to prevent starvation. For more
information, please check: https://github.com/vllm-project/vllm/issues/29406
"""
@config @config
@dataclass @dataclass