[Misc] Fix test_sleep to use query parameters (vllm-project#14373)

Signed-off-by: Lize Cai <lize.cai@sap.com>
Signed-off-by: youkaichao <youkaichao@gmail.com>
Co-authored-by: youkaichao <youkaichao@gmail.com>

Author: 2 people

tests/entrypoints/openai/test_sleep.py

@@ -25,8 +25,9 @@ def test_sleep_mode():
                                 "VLLM_SERVER_DEV_MODE": "1",
                                 "CUDA_VISIBLE_DEVICES": "0"
                             }) as remote_server:
+
         response = requests.post(remote_server.url_for("/sleep"),
-                                 data={"level": "1"})
+                                 params={"level": "1"})
         assert response.status_code == 200
         response = requests.get(remote_server.url_for("/is_sleeping"))
         assert response.status_code == 200