diff --git a/ci/L0_backend_vllm/metrics_test/vllm_metrics_test.py b/ci/L0_backend_vllm/metrics_test/vllm_metrics_test.py
index ad0cf990..8284835b 100644
--- a/ci/L0_backend_vllm/metrics_test/vllm_metrics_test.py
+++ b/ci/L0_backend_vllm/metrics_test/vllm_metrics_test.py
@@ -70,12 +70,10 @@ def get_metrics(self):
 
         return vllm_dict
 
-    def vllm_async_stream_infer(
+    def vllm_infer(
         self,
         prompts,
         sampling_parameters,
-        stream,
-        send_parameters_as_tensor,
         model_name,
     ):
         """
@@ -89,15 +87,15 @@ def vllm_async_stream_infer(
             request_data = create_vllm_request(
                 prompts[i],
                 i,
-                stream,
+                False,
                 sampling_parameters,
                 model_name,
-                send_parameters_as_tensor,
+                True,
             )
             self.triton_client.async_stream_infer(
                 model_name=model_name,
-                request_id=request_data["request_id"],
                 inputs=request_data["inputs"],
+                request_id=request_data["request_id"],
                 outputs=request_data["outputs"],
                 parameters=sampling_parameters,
             )
@@ -121,11 +119,9 @@ def test_vllm_metrics(self):
         }
 
         # Test vLLM metrics
-        self.vllm_async_stream_infer(
+        self.vllm_infer(
             prompts=self.prompts,
             sampling_parameters=self.sampling_parameters,
-            stream=False,
-            send_parameters_as_tensor=True,
             model_name=self.vllm_model_name,
         )
         expected_metrics_dict["vllm:prompt_tokens_total"] = 18
diff --git a/src/model.py b/src/model.py
index 650faccd..5e77e602 100644
--- a/src/model.py
+++ b/src/model.py
@@ -168,7 +168,7 @@ def init_engine(self):
         )
 
         # Add vLLM custom metrics
-        if not self.metrics:
+        if self.metrics:
             self.llm_engine.add_logger("triton", self.metrics)
 
     def setup_lora(self):