fix web_ui infer error (#5106)

Author: Jintao-Huang

examples/deploy/server/sglang.sh

@@ -0,0 +1,18 @@
+CUDA_VISIBLE_DEVICES=0,1 \
+swift deploy \
+    --model Qwen/Qwen3-8B \
+    --infer_backend sglang \
+    --max_new_tokens 2048 \
+    --sglang_context_length 8192 \
+    --sglang_tp_size 2 \
+    --served_model_name Qwen3-8B
+
+# After the server-side deployment above is successful, use the command below to perform a client call test.
+
+# curl http://localhost:8000/v1/chat/completions \
+# -H "Content-Type: application/json" \
+# -d '{
+# "model": "Qwen3-8B",
+# "messages": [{"role": "user", "content": "What is your name?"}],
+# "temperature": 0
+# }'

examples/deploy/server/demo.sh renamed to examples/deploy/server/vllm.sh

@@ -72,7 +72,7 @@ async def _run_async_iter():
             else:
                 queue.put(None)
 
-        loop = asyncio.get_event_loop()
+        loop = asyncio.new_event_loop()
         thread = Thread(target=lambda: loop.run_until_complete(_run_async_iter()))
         thread.start()
         pre_output = None
@@ -81,6 +81,7 @@ async def _run_async_iter():
             if output is None or isinstance(output, Exception):
                 prog_bar.update()
                 self._update_metrics(pre_output, metrics)
+                loop.close()
                 return
             pre_output = output
             yield output
@@ -255,10 +256,13 @@ def func(target, queue, args, kwargs):
 
     @staticmethod
     def safe_asyncio_run(coro):
-        loop = asyncio.get_event_loop()
+        loop = asyncio.new_event_loop()
 
         def asyncio_run(core):
-            return loop.run_until_complete(core)
+            try:
+                return loop.run_until_complete(core)
+            finally:
+                loop.close()
 
         return InferEngine.thread_run(asyncio_run, args=(coro, ))