diff --git a/fastchat/serve/sglang_worker.py b/fastchat/serve/sglang_worker.py index c5b8db43a..b30668433 100644 --- a/fastchat/serve/sglang_worker.py +++ b/fastchat/serve/sglang_worker.py @@ -200,7 +200,7 @@ async def api_generate_stream(request: Request): await acquire_worker_semaphore() generator = worker.generate_stream_gate(params) background_tasks = create_background_tasks() - return StreamingResponse(generator) + return StreamingResponse(generator, background=background_tasks) @app.post("/worker_generate")