[Perf][Qwen3-TTS] Enable event-driven orchestration by default

Signed-off-by: Sy03 <1370724210@qq.com>

docs/configuration/environment_variables.md

@@ -83,7 +83,7 @@ depends on the installed kernels and model path.
 | `SPEAKER_SAMPLES_DIR` | Filesystem path; default `~/.cache/vllm-omni/speakers` | Speech server; read when speaker storage initializes | Environment-only setting. The directory is created; filesystem errors propagate. | Stable |
 | `SPEAKER_MAX_UPLOADED` | Integer; default `1000` | Speech server; read when speaker storage initializes | Environment-only setting. A non-integer logs a warning and uses `1000`; range is not otherwise validated. | Stable |
 | `VLLM_OMNI_ASYNC_OUTPUT_TIMEOUT` | Float seconds; default `600` | Diffusion engine async-output wait in `step_streaming`; resolved per call on the request path, not at import | Environment-only setting. A non-float or `<=0` value warns once and uses the default. | Experimental |
-| `VLLM_OMNI_EVENT_DRIVEN_ORCH` | `1`, `true`, `yes` or `on` enables; default `0` (off) | Orchestration loop and the serving-side final-output drain; read once when the `Orchestrator` is constructed | Environment-only setting. Values are stripped and case-normalized; any unrecognized value leaves the legacy poll loop selected. | Experimental |
+| `VLLM_OMNI_EVENT_DRIVEN_ORCH` | `1`, `true`, `yes` or `on` enables; default `0` (off), except Qwen3-TTS defaults to on | Orchestration loop and the serving-side final-output drain; read once when the `Orchestrator` is constructed | Environment-only setting. An explicit value always wins. Values are stripped and case-normalized; any unrecognized value leaves the legacy poll loop selected. | Experimental |
 | `VLLM_OMNI_INPUT_WAIT_TIMEOUT_S` | Float seconds; default `600`; `<=0` disables | Full-payload input coordinator, not async-chunk transfer; read when the scheduler module imports in each worker | Environment-only setting. A non-float logs a warning and uses `600`. | Stable operational control |
 | `VLLM_OMNI_ORCH_MONITOR_PATH` | Filesystem path; default `<current-working-directory>/vllm_omni_orch_monitor_<timestamp>.json` | Orchestrator monitor enabled by `--enable-orch-monitor`; read when the monitor is created | Environment-only path override. Parent directories are created; write errors are logged. | Diagnostic |
 | `VLLM_OMNI_VIDEO_SYNC_TIMEOUT` | Float seconds; default `600` | Synchronous Videos API; read when the API server module imports | Environment-only setting. A non-float raises `ValueError` during import. | Experimental |

docs/design/module/engine_orchestration.md

@@ -72,7 +72,7 @@ the boundary affected by the in-flight stage client/process refactor in
 responsibilities proposed only by that PR are not current contracts.
 
 The orchestration loop also has an opt-in event-driven mode
-(`VLLM_OMNI_EVENT_DRIVEN_ORCH=1`, default off) proposed in
+(`VLLM_OMNI_EVENT_DRIVEN_ORCH=1`, default off except for Qwen3-TTS) proposed in
 [#5221](https://github.com/vllm-project/vllm-omni/pull/5221). It changes poll
 cadence only: the routing, ordering, and terminal-state contracts below hold
 identically on both loops.

tests/engine/test_orchestrator_event_driven.py

@@ -19,7 +19,10 @@ import pytest
 
 from vllm_omni.engine.async_omni_engine import AsyncOmniEngine
 from vllm_omni.engine.messages import ShutdownRequestMessage
-from vllm_omni.engine.orchestrator import _event_driven_orch_enabled
+from vllm_omni.engine.orchestrator import (
+    _event_driven_orch_default_for_pipeline,
+    _event_driven_orch_enabled,
+)
 
 from . import test_orchestrator as legacy
 from . import test_orchestrator_error_handling as legacy_errors
@@ -125,6 +128,19 @@ def test_flag_parsing(monkeypatch) -> None:
         assert _event_driven_orch_enabled() is False
 
 
+def test_only_qwen3_tts_has_a_pipeline_default() -> None:
+    assert _event_driven_orch_default_for_pipeline("qwen3_tts") is True
+    for model_type in (None, "qwen3_omni_moe", "minicpmo_4_5", "moss_tts_delay"):
+        assert _event_driven_orch_default_for_pipeline(model_type) is False
+
+
+def test_explicit_event_driven_value_overrides_pipeline_default(monkeypatch) -> None:
+    monkeypatch.delenv("VLLM_OMNI_EVENT_DRIVEN_ORCH", raising=False)
+    assert _event_driven_orch_enabled(default=True) is True
+    monkeypatch.setenv("VLLM_OMNI_EVENT_DRIVEN_ORCH", "0")
+    assert _event_driven_orch_enabled(default=True) is False
+
+
 def test_default_is_legacy_loop(monkeypatch) -> None:
     """Without the env flag, the harness runs the legacy poll loop."""
     monkeypatch.delenv("VLLM_OMNI_EVENT_DRIVEN_ORCH", raising=False)

vllm_omni/engine/async_omni_engine.py

@@ -13,6 +13,8 @@ from __future__ import annotations
 import asyncio
 import concurrent.futures
 import copy
+import json
+import os
 import queue
 import shutil
 import threading
@@ -69,7 +71,11 @@ from vllm_omni.engine.messages import (
     OutputMessage,
     StageSubmissionMessage,
 )
-from vllm_omni.engine.orchestrator import Orchestrator
+from vllm_omni.engine.orchestrator import (
+    _EVENT_DRIVEN_ORCH_ENV,
+    Orchestrator,
+    _event_driven_orch_default_for_pipeline,
+)
 from vllm_omni.engine.rpc_result_router import CorrelatedRpcClient
 from vllm_omni.engine.serialization import deserialize_additional_information
 from vllm_omni.engine.stage_client import StageClient
@@ -212,6 +218,40 @@ class AsyncOmniEngine:
         # retaining the richer resolver result for pipeline-wide settings.
         # Overrides of that private seam fall back to the factory below.
         deploy_config_path = kwargs.get("deploy_config")
+        # ``trust_remote_code`` is tri-state (bool | None): ``None`` means "not
+        # specified" so stage-config resolution can defer to the deploy yaml's
+        # per-stage value (see ``with_trust_remote_code_override``). The
+        # restriction path below loads the top-level HF config via vLLM's
+        # ``get_config``, which needs a real bool, so collapse ``None`` to the
+        # default ``False`` here (#5495).
+        pipeline_config = StageConfigFactory.get_pipeline_config(
+            model=model,
+            trust_remote_code=bool(trust_remote_code),
+            deploy_config_path=deploy_config_path,
+        )
+        self.endpoint_restrictions = pipeline_config.endpoint_restrictions if pipeline_config is not None else ()
+        self._duplex_runtime_extension_path = (
+            pipeline_config.duplex_runtime_extension if pipeline_config is not None else None
+        )
+        self.duplex_serving_adapter_path = (
+            pipeline_config.duplex_serving_adapter if pipeline_config is not None else None
+        )
+        self._duplex_control_enabled = bool(pipeline_config and pipeline_config.duplex_control_enabled)
+        if (
+            _EVENT_DRIVEN_ORCH_ENV not in os.environ
+            and pipeline_config is not None
+            and _event_driven_orch_default_for_pipeline(pipeline_config.model_type)
+        ):
+            os.environ[_EVENT_DRIVEN_ORCH_ENV] = "1"
+        self.duplex_session_config = DuplexSessionRuntimeConfig()
+        if deploy_config_path is not None:
+            self.duplex_session_config = load_deploy_config(deploy_config_path).duplex_session
+
+        # Tri-state: None means "not specified" — the deploy yaml's per-stage
+        # trust_remote_code stays in effect. An explicit True/False here is a
+        # global override (precedence: caller > deploy yaml > default False);
+        # the merge rule lives in with_trust_remote_code_override.
+        kwargs = with_trust_remote_code_override(kwargs, trust_remote_code)
         self._config_resolution: OmniConfigResolution | None = None
         self.config_path, self.stage_configs = self._resolve_stage_configs(
             model,

vllm_omni/engine/orchestrator.py

@@ -96,7 +96,7 @@ def cleanup_request_artifact_dirs(artifact_dirs: set[str] | list[str]) -> None:
 # 1 ms poll cadence to event-driven wakeups: one reader task per live LLM stage
 # replica awaits `client.get_output_async()` directly — the same pattern vLLM's
 # own AsyncLLM output handler uses — and feeds a single serial dispatch queue.
-# Default off; the legacy poll loop remains the fallback.
+# Default is off except for pipelines with an explicit validated default.
 _EVENT_DRIVEN_ORCH_ENV = "VLLM_OMNI_EVENT_DRIVEN_ORCH"
 
 # How often the event-driven loop reconciles its reader-task set against
@@ -104,8 +104,16 @@ _EVENT_DRIVEN_ORCH_ENV = "VLLM_OMNI_EVENT_DRIVEN_ORCH"
 _ORCH_READER_RECONCILE_INTERVAL_S = 0.5
 
 
-def _event_driven_orch_enabled() -> bool:
-    return os.environ.get(_EVENT_DRIVEN_ORCH_ENV, "0").strip().lower() in ("1", "true", "yes", "on")
+def _event_driven_orch_enabled(*, default: bool = False) -> bool:
+    value = os.environ.get(_EVENT_DRIVEN_ORCH_ENV)
+    if value is None:
+        return default
+    return value.strip().lower() in ("1", "true", "yes", "on")
+
+
+def _event_driven_orch_default_for_pipeline(pipeline_model_type: str | None) -> bool:
+    """Return whether a pipeline has a validated event-driven default."""
+    return pipeline_model_type == "qwen3_tts"
 
 
 def _build_terminal_empty_output(