fix: switch to Qwen2.5-72B-Instruct (7B not available on free inference API)

app.py

@@ -35,9 +35,9 @@ USER_PROMPT_TEMPLATE = """【参考资料】
 请基于资料用自然语言回答。"""
 
 client = InferenceClient(
-    model="Qwen/Qwen2.5-7B-Instruct",
+    model="Qwen/Qwen2.5-72B-Instruct",
     token=os.environ.get("HF_TOKEN"),
-    timeout=15,
+    timeout=20,
 )