[ci][test] adjust max wait time for cpu offloading test (#7709)

tests/utils.py

@@ -56,16 +56,14 @@ VLLM_PATH = Path(__file__).parent.parent
 
 class RemoteOpenAIServer:
     DUMMY_API_KEY = "token-abc123"  # vLLM's OpenAI server does not need API key
-    MAX_START_WAIT_S = 240  # wait for server to start for 240 seconds
 
-    def __init__(
-        self,
-        model: str,
-        cli_args: List[str],
-        *,
-        env_dict: Optional[Dict[str, str]] = None,
-        auto_port: bool = True,
-    ) -> None:
+    def __init__(self,
+                 model: str,
+                 cli_args: List[str],
+                 *,
+                 env_dict: Optional[Dict[str, str]] = None,
+                 auto_port: bool = True,
+                 max_wait_seconds: Optional[float] = None) -> None:
         if auto_port:
             if "-p" in cli_args or "--port" in cli_args:
                 raise ValueError("You have manually specified the port"
@@ -90,8 +88,9 @@ class RemoteOpenAIServer:
                                      env=env,
                                      stdout=sys.stdout,
                                      stderr=sys.stderr)
+        max_wait_seconds = max_wait_seconds or 240
         self._wait_for_server(url=self.url_for("health"),
-                              timeout=self.MAX_START_WAIT_S)
+                              timeout=max_wait_seconds)
 
     def __enter__(self):
         return self
@@ -145,7 +144,8 @@ def compare_two_settings(model: str,
                          arg1: List[str],
                          arg2: List[str],
                          env1: Optional[Dict[str, str]] = None,
-                         env2: Optional[Dict[str, str]] = None):
+                         env2: Optional[Dict[str, str]] = None,
+                         max_wait_seconds: Optional[float] = None) -> None:
     """
     Launch API server with two different sets of arguments/environments
     and compare the results of the API calls.
@@ -164,7 +164,10 @@ def compare_two_settings(model: str,
     token_ids = tokenizer(prompt)["input_ids"]
     results = []
     for args, env in ((arg1, env1), (arg2, env2)):
-        with RemoteOpenAIServer(model, args, env_dict=env) as server:
+        with RemoteOpenAIServer(model,
+                                args,
+                                env_dict=env,
+                                max_wait_seconds=max_wait_seconds) as server:
             client = server.get_client()
 
             # test models list