[Feat] Streaming TTS word-level timestamps via forced aligner (#4034)

Signed-off-by: wjinxu <1299461899@qq.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>

vllm_omni/entrypoints/openai/api_server.py

@@ -144,6 +144,7 @@ from vllm_omni.entrypoints.openai.video_api_utils import decode_audio_url, decod
 from vllm_omni.entrypoints.openpi.serving import ServingRealtimeRobotOpenPI
 from vllm_omni.errors import OmniClientError
 from vllm_omni.inputs.data import OmniDiffusionSamplingParams, OmniTextPrompt
+from vllm_omni.utils.forced_aligner import build_forced_aligner_config
 from vllm_omni.utils.tracking_parser import TrackingArgumentParser, TrackingNamespace
 
 logger = init_logger(__name__)
@@ -1058,7 +1059,11 @@ async def omni_init_app_state(
     )
 
     state.openai_serving_speech = OmniOpenAIServingSpeech(
-        engine_client, state.openai_serving_models, request_logger=request_logger, model_name=model_name
+        engine_client,
+        state.openai_serving_models,
+        request_logger=request_logger,
+        model_name=model_name,
+        forced_aligner_config=build_forced_aligner_config(args),
     )
 
     # Warm up speech pipeline (CUDA Graph capture, torch.compile) so the first