[Core] Graduate MiniCPM-o 4.5 and PersonaPlex full-duplex serving out of experimental (#6196)

Signed-off-by: chickeyton <ngton2014@gmail.com>
Signed-off-by: Gao Han <hgaoaf@connect.ust.hk>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Co-authored-by: Gao Han <hgaoaf@connect.ust.hk>

tests/benchmarks/patch/test_patch.py

@@ -28,7 +28,7 @@ from vllm_omni.benchmarks.patch.patch import (
     async_request_openai_realtime_duplex,
     should_request_stage_metrics,
 )
-from vllm_omni.experimental.fullduplex.client import RealtimeEventCollector
+from vllm_omni.clients.duplex import EventCollector
 
 pytestmark = [pytest.mark.core_model, pytest.mark.benchmark, pytest.mark.cpu]
 
@@ -131,7 +131,7 @@ async def test_seed_tts_realtime_duplex_exports_per_request_metrics(monkeypatch)
 
         def __init__(self, url):
             assert url == "ws://localhost:8000/v1/realtime?duplex=1"
-            self.events = RealtimeEventCollector()
+            self.events = EventCollector()
             self.configure_kwargs = None
             self.sent = []
             self.response_count = 0
@@ -206,7 +206,7 @@ async def test_seed_tts_realtime_duplex_exports_per_request_metrics(monkeypatch)
             return None
 
     monkeypatch.setattr(
-        "vllm_omni.benchmarks.patch.patch.RealtimeDuplexClient",
+        "vllm_omni.benchmarks.patch.patch._RealtimeTTSProbe",
         FakeRealtimeClient,
     )
     request_input = RequestFuncInput(