xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2025-12-10 06:35:00 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fix an endless loop issue when engine_step throws a RuntimeError (#339)

Browse Source
This commit is contained in:
coolcloudcol 2023-07-04 06:22:28 +08:00 committed by GitHub
parent 42e0c1df78
commit 7717d0838b
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
vllm/engine/async_llm_engine.py
View File

@ -144,7 +144,11 @@ class AsyncLLMEngine:
# Kick the engine if the engine is not running.
if not self.is_engine_running:
try:
await self.engine_step(request_id)
except RuntimeError as e:
await self.abort(request_id)
raise e
# Wait for new output. The group_event will be set in engine_step
# when there is new output available for the sequence group.