fix test

tests/spec_decode/e2e/conftest.py

@@ -185,7 +185,8 @@ def generator_inner():
 
         # Override logging interval to 0 for spec decode test run to
         # log all metrics in time.
-        if baseline_or_test == "test" and not use_async:
+        if (baseline_or_test == "test" and not use_async
+                and llm.llm_engine.log_stats):
             for sate_logger in llm.llm_engine.stat_loggers.values():
                 sate_logger.local_interval = 0
         set_random_seed(seed)
@@ -228,10 +229,13 @@ def get_output_from_llm_generator(
         outputs = llm.generate(prompts, sampling_params, use_tqdm=True)
         token_ids = [output.outputs[0].token_ids for output in outputs]
         tokens = [output.outputs[0].text for output in outputs]
-        stat_logger = llm.llm_engine.stat_loggers["prometheus"]
-        acceptance_rate = (
-            stat_logger.metrics.gauge_spec_decode_draft_acceptance_rate.labels(
-                **stat_logger.labels)._value.get())
+
+        # Fetch acceptance rate if logging is enabled.
+        if llm.llm_engine.log_stats:
+            stat_logger = llm.llm_engine.stat_loggers["prometheus"]
+            acceptance_rate = (stat_logger.metrics.
+                               gauge_spec_decode_draft_acceptance_rate.labels(
+                                   **stat_logger.labels)._value.get())
         del llm
 
     return tokens, token_ids, acceptance_rate