[Model] Keep CosyVoice3 defaults compatible without CUDA MPS

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

docs/user_guide/examples/offline_inference/text_to_speech.md

@@ -130,7 +130,7 @@ python examples/offline_inference/text_to_speech/cosyvoice3/end2end.py \
 ### Notes
 
 - Stage 0 (`talker`) emits speech tokens; stage 1 (`code2wav`) runs flow matching + HiFiGAN to synthesize waveform.
-- H200 automatically selects `cosyvoice3_packed_streaming_optimized_standard.yaml`; other devices use `cosyvoice3.yaml`. See the [serving defaults and tradeoffs](../online_serving/text_to_speech.md#cosyvoice3). Pass `--deploy-config cosyvoice3.yaml` for the previous profile or `--deploy-config <path>` to override. The example uses the resolved profile's sampling parameters.
+- H200 automatically selects `cosyvoice3_packed_streaming_optimized_standard.yaml` when the CUDA MPS control tool is available; other devices and environments use `cosyvoice3.yaml`. See the [serving defaults and tradeoffs](../online_serving/text_to_speech.md#cosyvoice3). Pass `--deploy-config cosyvoice3.yaml` for the previous profile or `--deploy-config <path>` to override. The example uses the resolved profile's sampling parameters.
 - `async_chunk: true` is the default; pass `--no-async-chunk` to switch to the legacy synchronous path.
 
 ---

docs/user_guide/examples/online_serving/text_to_speech.md

@@ -155,7 +155,9 @@ vllm serve FunAudioLLM/Fun-CosyVoice3-0.5B-2512 --omni --port 8091 --trust-remot
 ./cosyvoice3/run_server.sh
 ```
 
-Streaming is on by default. On Hopper GPUs with at least 140 GiB of device memory (H200), serving automatically selects `cosyvoice3_packed_streaming_optimized_standard.yaml`: AR32 / codec32, standard sampling, packed Flow, cached ISTFT, HiFT decode graphs, and local CUDA MPS. No `--hf-overrides` is needed. Other devices use `cosyvoice3.yaml` with RAS sampling and AR8 / codec8.
+Streaming is on by default. On CUDA Hopper GPUs with at least 140 GiB of device memory (H200), serving automatically selects `cosyvoice3_packed_streaming_optimized_standard.yaml` when `nvidia-cuda-mps-control` is available on `PATH`: AR32 / codec32, standard sampling, packed Flow, cached ISTFT, HiFT decode graphs, and local CUDA MPS. No `--hf-overrides` is needed. Other devices and environments without the MPS control tool use `cosyvoice3.yaml` with RAS sampling and AR8 / codec8.
+
+The optimized profile is not a universal hardware configuration. Its packed kernels are Hopper-only; the talker's Model Runner V2 is unsupported on NPU/XPU, and CUDA MPS requires a local NVIDIA GPU and its control tool. Larger batches, 25-frame chunks, and graph allocations also require per-device validation on smaller or slower GPUs. Automatic selection retains the generic profile outside the validated device class; an explicit deploy configuration takes precedence.
 
 The H200 default targets streaming throughput at C32/C64. It changes sampling from RAS to standard (temperature 0.7, top-p 0.8, top-k 20, repetition penalty 1.21); output lengths and quality can differ. HiFT captures a recurring exact shape on its third use, which adds latency to that request, and its graph pool needs memory headroom beyond the engine reservations. This profile is not a claim of optimal first-audio latency at low concurrency. Pass `--deploy-config cosyvoice3.yaml` to retain the previous configuration, or `--deploy-config <path>` to select another profile.
 

tests/config/test_config_factory.py

@@ -572,17 +572,20 @@ class TestPipelineDiscovery:
         assert "qwen3_tts" in OMNI_PIPELINES
 
     @pytest.mark.parametrize(
-        "cuda,major,memory_gib,optimized",
+        "cuda,major,memory_gib,mps,optimized",
         [
-            (False, None, 0, False),
-            (True, None, 141, False),
-            (True, 8, 141, False),
-            (True, 9, 80, False),
-            (True, 10, 180, False),
-            (True, 9, 141, True),
+            (False, None, 0, True, False),
+            (True, None, 141, True, False),
+            (True, 8, 141, True, False),
+            (True, 9, 80, True, False),
+            (True, 10, 180, True, False),
+            (True, 9, 141, False, False),
+            (True, 9, 141, True, True),
         ],
     )
-    def test_cosyvoice3_device_default_and_explicit_overrides(self, monkeypatch, cuda, major, memory_gib, optimized):
+    def test_cosyvoice3_device_default_and_explicit_overrides(
+        self, monkeypatch, cuda, major, memory_gib, mps, optimized
+    ):
         from vllm.platforms import current_platform
         from vllm.platforms.interface import DeviceCapability
 
@@ -590,6 +593,10 @@ class TestPipelineDiscovery:
         capability = DeviceCapability(major, 0) if major is not None else None
         monkeypatch.setattr(current_platform, "get_device_capability", lambda: capability)
         monkeypatch.setattr(current_platform, "get_device_total_memory", lambda: memory_gib * 1024**3)
+        monkeypatch.setattr(
+            "vllm_omni.model_executor.models.cosyvoice3.pipeline.shutil.which",
+            lambda command: "/usr/bin/nvidia-cuda-mps-control" if mps else None,
+        )
         pipeline = resolve_pipeline_config("cosyvoice3")
         assert pipeline is not None
         config = VllmOmniConfig.from_pipeline_config(pipeline)

vllm_omni/deploy/cosyvoice3_packed_streaming_optimized_standard.yaml

@@ -1,5 +1,6 @@
 # Single-Hopper streaming throughput profile: AR32 / codec32, Flow10.
-# Automatically selected on Hopper devices with at least 140 GiB (H200).
+# Automatically selected on Hopper devices with at least 140 GiB (H200)
+# when nvidia-cuda-mps-control is available on PATH.
 # Standard sampling is configured below; request stream=true for first audio.
 # F0 and HiFT stay FP32; only deterministic HiFT decode is graphed.
 # HiFT graphs capture on the third use of each exact shape, adding latency then.

vllm_omni/model_executor/models/cosyvoice3/pipeline.py

@@ -12,6 +12,7 @@ Stage 1: Code2Wav — flow-matching decoder → acoustic features → waveform.
     through the shared-memory connector.
 """
 
+import shutil
 from dataclasses import replace
 
 from transformers import PretrainedConfig
@@ -61,7 +62,7 @@ COSYVOICE3_PIPELINE = PipelineConfig(
 
 
 def resolve_cosyvoice3_pipeline(hf_config: PretrainedConfig | None = None) -> PipelineConfig | None:
-    """Use the validated throughput profile on H200-class Hopper devices.
+    """Select a streaming default compatible with the device and runtime.
 
     Query the platform through NVML without initializing CUDA in the parent.
     Explicit deploy configs still take precedence over this pipeline default.
@@ -76,6 +77,7 @@ def resolve_cosyvoice3_pipeline(hf_config: PretrainedConfig | None = None) -> Pi
             capability is not None
             and capability.major == 9
             and current_platform.get_device_total_memory() >= 140 * 1024**3
+            and shutil.which("nvidia-cuda-mps-control") is not None
         ):
             return replace(
                 COSYVOICE3_PIPELINE,