diff --git a/.lightning/workflows/tests.yaml b/.lightning/workflows/tests.yaml
index 8604cac59c..344dc1f477 100644
--- a/.lightning/workflows/tests.yaml
+++ b/.lightning/workflows/tests.yaml
@@ -43,7 +43,9 @@ run: |
   python -c "import torch ; gpus = torch.cuda.device_count() ; assert gpus >= 2, f'GPU: {gpus}'"
   python -c "from torch import __version__ as ver ; assert str(ver).split('+')[0] == '$TORCH_VERSION', f'PyTorch: installed {ver} but expected $TORCH_VERSION'"
 
-  pytest -v --durations=100
+  pytest -v --durations=100 tests/ --ignore=tests/test_serve.py
+  pytest -v --durations=100 tests/test_serve.py
+
 
   wget https://raw.githubusercontent.com/Lightning-AI/utilities/main/scripts/run_standalone_tests.sh
   PL_RUN_STANDALONE_TESTS=1 bash run_standalone_tests.sh "tests"
diff --git a/tests/test_serve.py b/tests/test_serve.py
index 81d51dcd27..dc1dff1e78 100644
--- a/tests/test_serve.py
+++ b/tests/test_serve.py
@@ -208,7 +208,9 @@ def run_server():
     try:
         # Test server health
         response = requests.get("http://127.0.0.1:8000/health")
-        assert response.status_code == 200, f"Server health check failed with status code {response.status_code}"
+        assert response.status_code == 200, (
+            f"Server health check failed with status code {response.status_code}: {response.text}"
+        )
         assert response.text == "ok", "Server did not respond as expected."
 
         # Test non-streaming chat completion
@@ -220,7 +222,7 @@ def run_server():
             },
         )
         assert response.status_code == 200, (
-            f"Non-streaming chat completion failed with status code {response.status_code}"
+            f"Non-streaming chat completion failed with status code {response.status_code}: {response.text}"
         )
         response_json = response.json()
         assert "choices" in response_json, "Response JSON does not contain 'choices'."