Minor fix

triton-inference-server · Aug 7, 2024 · e9d0dbb · e9d0dbb
1 parent bf7669e
commit e9d0dbb
Showing 1 changed file with 1 addition and 12 deletions.
diff --git a/ci/L0_backend_vllm/metrics_test/vllm_metrics_test.py b/ci/L0_backend_vllm/metrics_test/vllm_metrics_test.py
@@ -66,7 +66,7 @@ def get_metrics(self):
 
         for match in matches:
             key, value = match
-            vllm_dict[key] = int(value)
+            vllm_dict[key] = float(value) if "." in value else int(value)
 
         return vllm_dict
 
@@ -132,17 +132,6 @@ def test_vllm_metrics(self):
         expected_metrics_dict["vllm:generation_tokens_total"] = 48
         self.assertEqual(self.get_metrics(), expected_metrics_dict)
 
-        self.vllm_async_stream_infer(
-            prompts=self.prompts,
-            sampling_parameters=self.sampling_parameters,
-            stream=False,
-            send_parameters_as_tensor=False,
-            model_name=self.vllm_model_name,
-        )
-        expected_metrics_dict["vllm:prompt_tokens_total"] = 36
-        expected_metrics_dict["vllm:generation_tokens_total"] = 96
-        self.assertEqual(self.get_metrics(), expected_metrics_dict)
-
     def tearDown(self):
         self.triton_client.close()