[Full Duplex] Feat: Support Full-Duplex realtime runtime & add MiniCPM-o 4.5 demo (#3907)

Signed-off-by: Sy03 <1370724210@qq.com>
Signed-off-by: lishunyang12 <lishunyang03@gmail.com>
Signed-off-by: Yueqian Lin <linyueqian@outlook.com>
Co-authored-by: lishunyang12 <lishunyang03@gmail.com>
Co-authored-by: Yueqian Lin <linyueqian@outlook.com>
Co-authored-by: StormMelody <22371184@zju.edu.cn>
Co-authored-by: Gao Han <hgaoaf@connect.ust.hk>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>

tests/engine/test_async_omni_engine_input.py

@@ -1,4 +1,5 @@
 import pytest
+import torch
 from pytest_mock import MockerFixture
 from vllm.sampling_params import SamplingParams
 from vllm.v1.engine import EngineCoreRequest
@@ -64,6 +65,48 @@ def test_build_add_request_message_preserves_additional_information(mocker: Mock
     output_processor.add_request.assert_not_called()
 
 
+def test_build_add_request_message_preserves_model_intermediate_buffer(mocker: MockerFixture):
+    engine = object.__new__(AsyncOmniEngine)
+    params = SamplingParams(max_tokens=8)
+    engine.default_sampling_params_list = [params]
+    engine.stage_metadata = [StageRuntimeInfo(final_output=False, final_output_type=None, stage_type="llm")]
+    engine.supported_tasks = ("speech",)
+
+    input_processor = mocker.Mock()
+    input_processor.process_inputs.return_value = _make_engine_core_request()
+    engine.input_processor = input_processor
+
+    output_processor = mocker.Mock()
+    engine.output_processors = [output_processor]
+
+    hidden = torch.arange(6, dtype=torch.float32).reshape(2, 3)
+    prompt = {
+        "prompt_token_ids": [1, 1, 1],
+        "model_intermediate_buffer": {
+            "ids": {"tts": [11, 12]},
+            "hidden_states": {"tts": hidden},
+        },
+    }
+
+    msg = engine._build_add_request_message(
+        request_id="req-1",
+        prompt=prompt,
+        sampling_params_list=[params],
+        final_stage_id=0,
+        arrival_time=0.0,
+    )
+
+    request = msg.prompt
+    assert isinstance(request, OmniEngineCoreRequest)
+    assert request.additional_information is not None
+    assert request.additional_information.entries["global_request_id"].list_data == ["req-1"]
+    assert request.additional_information.entries["omni_final_stage_id"].scalar_data == 0
+    assert isinstance(request.model_intermediate_buffer, dict)
+    info = request.model_intermediate_buffer
+    assert info["ids"]["tts"] == [11, 12]
+    assert torch.equal(info["hidden_states"]["tts"], hidden)
+
+
 def test_build_add_request_message_with_resumable_streaming(mocker: MockerFixture):
     engine = object.__new__(AsyncOmniEngine)
     params = SamplingParams(max_tokens=8)