[Bugfix] Fix SHM cache initialization (#26427)

Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>

tests/entrypoints/openai/test_lora_resolvers.py

@@ -113,15 +113,17 @@ def mock_serving_setup():
     mock_engine.generate.reset_mock()
     mock_engine.add_lora.reset_mock()
 
-    mock_model_config = MockModelConfig()
+    mock_engine.model_config = MockModelConfig()
+    mock_engine.processor = MagicMock()
+    mock_engine.io_processor = MagicMock()
+
     models = OpenAIServingModels(
         engine_client=mock_engine,
         base_model_paths=BASE_MODEL_PATHS,
-        model_config=mock_model_config,
     )
 
     serving_completion = OpenAIServingCompletion(
-        mock_engine, mock_model_config, models, request_logger=None
+        mock_engine, models, request_logger=None
     )
 
     serving_completion._process_inputs = AsyncMock(

tests/entrypoints/openai/test_serving_chat.py

@@ -245,17 +245,13 @@ class MockModelConfig:
         return self.diff_sampling_param or {}
 
 
-def _build_serving_chat(
+def _build_serving_chat(engine: AsyncLLM) -> OpenAIServingChat:
-    engine: AsyncLLM, model_config: MockModelConfig
-) -> OpenAIServingChat:
     models = OpenAIServingModels(
         engine_client=engine,
         base_model_paths=BASE_MODEL_PATHS,
-        model_config=model_config,
     )
     serving_chat = OpenAIServingChat(
         engine,
-        model_config,
         models,
         response_role="assistant",
         chat_template=CHAT_TEMPLATE,
@@ -280,18 +276,17 @@ def _build_serving_chat(
 
 @dataclass
 class MockEngine:
-    async def get_model_config(self):
+    model_config: MockModelConfig = field(default_factory=MockModelConfig)
-        return MockModelConfig()
+    processor: MagicMock = field(default_factory=MagicMock)
+    io_processor: MagicMock = field(default_factory=MagicMock)
 
 
 async def _async_serving_chat_init():
     engine = MockEngine()
-    model_config = await engine.get_model_config()
 
-    models = OpenAIServingModels(engine, model_config, BASE_MODEL_PATHS)
+    models = OpenAIServingModels(engine, BASE_MODEL_PATHS)
     serving_completion = OpenAIServingChat(
         engine,
-        model_config,
         models,
         response_role="assistant",
         chat_template=CHAT_TEMPLATE,
@@ -311,8 +306,11 @@ async def test_serving_chat_returns_correct_model_name():
     mock_engine = MagicMock(spec=AsyncLLM)
     mock_engine.get_tokenizer.return_value = get_tokenizer(MODEL_NAME)
     mock_engine.errored = False
+    mock_engine.model_config = MockModelConfig()
+    mock_engine.processor = MagicMock()
+    mock_engine.io_processor = MagicMock()
 
-    serving_chat = _build_serving_chat(mock_engine, MockModelConfig())
+    serving_chat = _build_serving_chat(mock_engine)
     messages = [{"role": "user", "content": "what is 1+1?"}]
 
     async def return_model_name(*args):
@@ -338,8 +336,11 @@ async def test_serving_chat_should_set_correct_max_tokens():
     mock_engine = MagicMock(spec=AsyncLLM)
     mock_engine.get_tokenizer.return_value = get_tokenizer(MODEL_NAME)
     mock_engine.errored = False
+    mock_engine.model_config = MockModelConfig()
+    mock_engine.processor = MagicMock()
+    mock_engine.io_processor = MagicMock()
 
-    serving_chat = _build_serving_chat(mock_engine, MockModelConfig())
+    serving_chat = _build_serving_chat(mock_engine)
 
     req = ChatCompletionRequest(
         model=MODEL_NAME,
@@ -368,9 +369,12 @@ async def test_serving_chat_should_set_correct_max_tokens():
     mock_engine = MagicMock(spec=AsyncLLM)
     mock_engine.get_tokenizer.return_value = get_tokenizer(MODEL_NAME)
     mock_engine.errored = False
+    mock_engine.model_config = mock_model_config
+    mock_engine.processor = MagicMock()
+    mock_engine.io_processor = MagicMock()
 
     # Initialize the serving chat
-    serving_chat = _build_serving_chat(mock_engine, mock_model_config)
+    serving_chat = _build_serving_chat(mock_engine)
 
     # Test Case 1: No max_tokens specified in request
     req = ChatCompletionRequest(
@@ -410,9 +414,12 @@ async def test_serving_chat_should_set_correct_max_tokens():
     mock_engine = MagicMock(spec=AsyncLLM)
     mock_engine.get_tokenizer.return_value = get_tokenizer(MODEL_NAME)
     mock_engine.errored = False
+    mock_engine.model_config = mock_model_config
+    mock_engine.processor = MagicMock()
+    mock_engine.io_processor = MagicMock()
 
     # Initialize the serving chat
-    serving_chat = _build_serving_chat(mock_engine, mock_model_config)
+    serving_chat = _build_serving_chat(mock_engine)
 
     # Test case 1: No max_tokens specified, defaults to context_window
     req = ChatCompletionRequest(
@@ -453,9 +460,12 @@ async def test_serving_chat_could_load_correct_generation_config():
     mock_engine = MagicMock(spec=AsyncLLM)
     mock_engine.get_tokenizer.return_value = get_tokenizer(MODEL_NAME)
     mock_engine.errored = False
+    mock_engine.model_config = mock_model_config
+    mock_engine.processor = MagicMock()
+    mock_engine.io_processor = MagicMock()
 
     # Initialize the serving chat
-    serving_chat = _build_serving_chat(mock_engine, mock_model_config)
+    serving_chat = _build_serving_chat(mock_engine)
 
     req = ChatCompletionRequest(
         model=MODEL_NAME,
@@ -496,8 +506,11 @@ async def test_serving_chat_did_set_correct_cache_salt(model_type):
     mock_engine = MagicMock(spec=AsyncLLM)
     mock_engine.get_tokenizer.return_value = get_tokenizer(MODEL_NAME)
     mock_engine.errored = False
+    mock_engine.model_config = mock_model_config
+    mock_engine.processor = MagicMock()
+    mock_engine.io_processor = MagicMock()
 
-    serving_chat = _build_serving_chat(mock_engine, mock_model_config)
+    serving_chat = _build_serving_chat(mock_engine)
 
     # Test cache_salt
     req = ChatCompletionRequest(

tests/entrypoints/openai/test_serving_engine.py

@@ -22,10 +22,12 @@ def serving() -> OpenAIServing:
     model_config = Mock(spec=ModelConfig)
     model_config.max_model_len = 32768
     models = Mock(spec=OpenAIServingModels)
+    models.model_config = model_config
+    models.processor = Mock()
+    models.io_processor = Mock()
 
     serving = OpenAIServing(
         engine_client=engine_client,
-        model_config=model_config,
         models=models,
         request_logger=None,
     )

tests/entrypoints/openai/test_serving_models.py

@@ -25,15 +25,17 @@ LORA_UNLOADING_SUCCESS_MESSAGE = (
 
 
 async def _async_serving_models_init() -> OpenAIServingModels:
-    mock_model_config = MagicMock(spec=ModelConfig)
     mock_engine_client = MagicMock(spec=EngineClient)
     # Set the max_model_len attribute to avoid missing attribute
+    mock_model_config = MagicMock(spec=ModelConfig)
     mock_model_config.max_model_len = 2048
+    mock_engine_client.model_config = mock_model_config
+    mock_engine_client.processor = MagicMock()
+    mock_engine_client.io_processor = MagicMock()
 
     serving_models = OpenAIServingModels(
         engine_client=mock_engine_client,
         base_model_paths=BASE_MODEL_PATHS,
-        model_config=mock_model_config,
         lora_modules=None,
     )
     await serving_models.init_static_loras()

tests/entrypoints/openai/test_serving_responses.py

@@ -2,7 +2,7 @@
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
 from contextlib import AsyncExitStack
-from unittest.mock import AsyncMock, MagicMock
+from unittest.mock import MagicMock
 
 import pytest
 import pytest_asyncio
@@ -70,11 +70,14 @@ class TestInitializeToolSessions:
         """Create a real OpenAIServingResponses instance for testing"""
         # Create minimal mocks for required dependencies
         engine_client = MagicMock()
-        engine_client.get_model_config = AsyncMock()
 
         model_config = MagicMock()
         model_config.hf_config.model_type = "test"
         model_config.get_diff_sampling_param.return_value = {}
+        engine_client.model_config = model_config
+
+        engine_client.processor = MagicMock()
+        engine_client.io_processor = MagicMock()
 
         models = MagicMock()
 
@@ -83,7 +86,6 @@ class TestInitializeToolSessions:
         # Create the actual instance
         instance = OpenAIServingResponses(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=None,
             chat_template=None,
@@ -132,18 +134,20 @@ class TestValidateGeneratorInput:
         """Create a real OpenAIServingResponses instance for testing"""
         # Create minimal mocks for required dependencies
         engine_client = MagicMock()
-        engine_client.get_model_config = AsyncMock()
 
         model_config = MagicMock()
         model_config.hf_config.model_type = "test"
         model_config.get_diff_sampling_param.return_value = {}
+        engine_client.model_config = model_config
+
+        engine_client.processor = MagicMock()
+        engine_client.io_processor = MagicMock()
 
         models = MagicMock()
 
         # Create the actual instance
         instance = OpenAIServingResponses(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=None,
             chat_template=None,

tests/test_inputs.py

@@ -7,6 +7,7 @@ from vllm.config import ModelConfig
 from vllm.inputs import zip_enc_dec_prompts
 from vllm.inputs.parse import parse_raw_prompts
 from vllm.inputs.preprocess import InputPreprocessor
+from vllm.transformers_utils.tokenizer import init_tokenizer_from_configs
 
 pytestmark = pytest.mark.cpu_test
 
@@ -106,7 +107,8 @@ def test_zip_enc_dec_prompts(mm_processor_kwargs, expected_mm_kwargs):
 )
 def test_preprocessor_text_no_mm_inputs(model_id, prompt):
     model_config = ModelConfig(model=model_id)
-    input_preprocessor = InputPreprocessor(model_config)
+    tokenizer = init_tokenizer_from_configs(model_config)
+    input_preprocessor = InputPreprocessor(model_config, tokenizer)
 
     with pytest.raises(ValueError, match="does not support multimodal inputs"):
         input_preprocessor.preprocess(prompt)
@@ -127,8 +129,8 @@ def test_preprocessor_text_no_mm_inputs(model_id, prompt):
 )
 def test_preprocessor_always_mm_code_path(model_id, prompt):
     model_config = ModelConfig(model=model_id)
-    input_preprocessor = InputPreprocessor(model_config)
+    tokenizer = init_tokenizer_from_configs(model_config)
-    tokenizer = input_preprocessor.tokenizer
+    input_preprocessor = InputPreprocessor(model_config, tokenizer)
 
     # HF processor adds sep token
     sep_token_id = tokenizer.vocab[tokenizer.sep_token]

tests/v1/engine/test_processor_multi_modal_uuids.py

@@ -65,7 +65,7 @@ def _mk_processor(
         device_config=DeviceConfig(device="cpu"),
     )
 
-    return Processor(vllm_config)
+    return Processor(vllm_config, tokenizer=None)
 
 
 def test_multi_modal_uuids_length_mismatch_raises(monkeypatch):

tests/v1/sample/test_logprobs.py

@@ -459,7 +459,7 @@ def test_all_logprobs(example_prompts):
     results_logprobs_all = runner.llm.generate(
         example_prompts, sampling_params=sampling_params_logprobs_all
     )
-    vocab_size = runner.llm.llm_engine.get_model_config().get_vocab_size()
+    vocab_size = runner.llm.llm_engine.model_config.get_vocab_size()
 
     for i in range(len(results_logprobs_all)):
         logprobs = results_logprobs_all[i].outputs[0].logprobs

vllm/benchmarks/throughput.py

@@ -186,7 +186,7 @@ async def run_vllm_async(
         engine_args,
         disable_frontend_multiprocessing=disable_frontend_multiprocessing,
     ) as llm:
-        model_config = await llm.get_model_config()
+        model_config = llm.model_config
         assert all(
             model_config.max_model_len
             >= (request.prompt_len + request.expected_output_len)

vllm/engine/protocol.py

@@ -1,26 +1,23 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
 
-import asyncio
 from abc import ABC, abstractmethod
 from collections.abc import AsyncGenerator, Iterable, Mapping
 from typing import Any, Optional, Union
 
-from vllm.beam_search import BeamSearchSequence, create_sort_beams_key_function
 from vllm.config import ModelConfig, VllmConfig
-from vllm.inputs.data import PromptType, TokensPrompt
+from vllm.inputs.data import PromptType
-from vllm.inputs.parse import is_explicit_encoder_decoder_prompt
-from vllm.inputs.preprocess import InputPreprocessor
 from vllm.logger import init_logger
 from vllm.lora.request import LoRARequest
-from vllm.outputs import CompletionOutput, PoolingRequestOutput, RequestOutput
+from vllm.outputs import PoolingRequestOutput, RequestOutput
-from vllm.plugins.io_processors.interface import IOProcessor
+from vllm.plugins.io_processors import IOProcessor
 from vllm.pooling_params import PoolingParams
-from vllm.sampling_params import BeamSearchParams, SamplingParams
+from vllm.sampling_params import SamplingParams
 from vllm.tasks import SupportedTask
 from vllm.transformers_utils.tokenizer import AnyTokenizer
-from vllm.utils import Device, collect_from_async_generator, random_uuid
+from vllm.utils import Device
 from vllm.v1.engine import EngineCoreRequest
+from vllm.v1.engine.processor import Processor
 
 logger = init_logger(__name__)
 
@@ -28,6 +25,11 @@ logger = init_logger(__name__)
 class EngineClient(ABC):
     """Protocol class for Clients to Engine"""
 
+    vllm_config: VllmConfig
+    model_config: ModelConfig
+    processor: Processor
+    io_processor: Optional[IOProcessor]
+
     @property
     @abstractmethod
     def is_running(self) -> bool: ...
@@ -61,180 +63,6 @@ class EngineClient(ABC):
         """Generate outputs for a request."""
         ...
 
-    async def beam_search(
-        self,
-        prompt: PromptType,
-        request_id: str,
-        params: BeamSearchParams,
-        lora_request: Optional[LoRARequest] = None,
-    ) -> AsyncGenerator[RequestOutput, None]:
-        beam_width = params.beam_width
-        max_tokens = params.max_tokens
-        ignore_eos = params.ignore_eos
-        temperature = params.temperature
-        length_penalty = params.length_penalty
-        include_stop_str_in_output = params.include_stop_str_in_output
-
-        preprocessor = await self.get_input_preprocessor()
-        tokenizer = preprocessor.get_tokenizer()
-        eos_token_id = tokenizer.eos_token_id
-
-        if is_explicit_encoder_decoder_prompt(prompt):
-            raise NotImplementedError
-        else:
-            processed_inputs = preprocessor._prompt_to_llm_inputs(prompt)
-
-        if processed_inputs["type"] == "embeds":
-            raise NotImplementedError
-
-        # This is a workaround to fix multimodal beam search; this is a
-        # bandaid fix for 2 small problems:
-        # 1. Multi_modal_data on the processed_inputs currently resolves to
-        #    `None`.
-        # 2. preprocessing above expands the multimodal placeholders. However,
-        #    this happens again in generation, so the double expansion causes
-        #    a mismatch.
-        # TODO - would be ideal to handle this more gracefully.
-        if isinstance(prompt, str):
-            prompt_text = prompt
-            prompt_token_ids = []
-            multi_modal_data = None
-        else:
-            prompt_text = prompt.get("prompt")
-            prompt_token_ids = prompt.get("prompt_token_ids", [])
-            multi_modal_data = prompt.get("multi_modal_data")
-
-        mm_processor_kwargs = processed_inputs.get("mm_processor_kwargs")
-
-        tokenized_length = len(prompt_token_ids)
-
-        sort_beams_key = create_sort_beams_key_function(eos_token_id, length_penalty)
-
-        beam_search_params = SamplingParams(
-            logprobs=2 * beam_width,
-            max_tokens=1,
-            temperature=temperature,
-        )
-        all_beams = [
-            BeamSearchSequence(
-                tokens=prompt_token_ids,
-                cum_logprob=0,
-                logprobs=[],
-                multi_modal_data=multi_modal_data,
-                mm_processor_kwargs=mm_processor_kwargs,
-                lora_request=lora_request,
-            )
-        ]
-        completed = []
-
-        for _ in range(max_tokens):
-            prompts_batch, lora_req_batch = zip(
-                *[
-                    (
-                        TokensPrompt(
-                            prompt_token_ids=beam.tokens,
-                            multi_modal_data=beam.multi_modal_data,
-                            mm_processor_kwargs=beam.mm_processor_kwargs,
-                        ),
-                        beam.lora_request,
-                    )
-                    for beam in all_beams
-                ]
-            )
-
-            tasks = []
-
-            request_id = f"beam_search-{random_uuid()}"
-            for i, (individual_prompt, lora_req) in enumerate(
-                zip(prompts_batch, lora_req_batch)
-            ):
-                request_id_item = f"{request_id}-{i}"
-                task = asyncio.create_task(
-                    collect_from_async_generator(
-                        self.generate(
-                            individual_prompt,
-                            beam_search_params,
-                            request_id_item,
-                            lora_request=lora_req,
-                        )
-                    )
-                )
-                tasks.append(task)
-
-            output = await asyncio.gather(*tasks)
-
-            output = [x[0] for x in output]
-
-            new_beams = []
-            for i, current_beam in enumerate(all_beams):
-                result = output[i]
-
-                if result.outputs[0].logprobs is not None:
-                    logprobs = result.outputs[0].logprobs[0]
-                    for token_id, logprob_obj in logprobs.items():
-                        if token_id == eos_token_id and not ignore_eos:
-                            completed.append(
-                                BeamSearchSequence(
-                                    tokens=current_beam.tokens + [token_id]
-                                    if include_stop_str_in_output
-                                    else current_beam.tokens,
-                                    logprobs=current_beam.logprobs + [logprobs],
-                                    cum_logprob=current_beam.cum_logprob
-                                    + logprob_obj.logprob,
-                                    finish_reason="stop",
-                                    stop_reason=eos_token_id,
-                                )
-                            )
-                        else:
-                            new_beams.append(
-                                BeamSearchSequence(
-                                    tokens=current_beam.tokens + [token_id],
-                                    logprobs=current_beam.logprobs + [logprobs],
-                                    lora_request=current_beam.lora_request,
-                                    cum_logprob=current_beam.cum_logprob
-                                    + logprob_obj.logprob,
-                                    multi_modal_data=current_beam.multi_modal_data,
-                                    mm_processor_kwargs=current_beam.mm_processor_kwargs,
-                                )
-                            )
-
-            sorted_beams = sorted(new_beams, key=sort_beams_key, reverse=True)
-            all_beams = sorted_beams[:beam_width]
-
-        completed.extend(all_beams)
-        sorted_completed = sorted(completed, key=sort_beams_key, reverse=True)
-        best_beams = sorted_completed[:beam_width]
-
-        for beam in best_beams:
-            if beam.tokens[-1] == eos_token_id and not ignore_eos:
-                # Skip the eos token in the text.
-                tokens = beam.tokens[tokenized_length:-1]
-            else:
-                tokens = beam.tokens[tokenized_length:]
-            beam.text = tokenizer.decode(tokens)
-
-        yield RequestOutput(
-            request_id=request_id,
-            prompt=prompt_text,
-            outputs=[
-                CompletionOutput(
-                    text=beam.text,
-                    cumulative_logprob=beam.cum_logprob,
-                    token_ids=beam.tokens[tokenized_length:],
-                    index=i,
-                    logprobs=beam.logprobs,
-                    finish_reason=beam.finish_reason
-                    if beam.finish_reason is not None
-                    else "length",
-                    stop_reason=beam.stop_reason,
-                )
-                for (i, beam) in enumerate(best_beams)
-            ],
-            finished=True,
-            prompt_token_ids=prompt_token_ids,
-            prompt_logprobs=None,
-        )
-
     @abstractmethod
     def encode(
         self,
@@ -259,29 +87,11 @@ class EngineClient(ABC):
         """
         ...
 
-    @abstractmethod
-    async def get_vllm_config(self) -> VllmConfig:
-        """Get the vllm configuration of the vLLM engine."""
-        ...
-
-    @abstractmethod
-    async def get_model_config(self) -> ModelConfig:
-        """Get the model configuration of the vLLM engine."""
-        ...
-
-    @abstractmethod
-    async def get_input_preprocessor(self) -> InputPreprocessor:
-        """Get the input processor of the vLLM engine."""
-        ...
-
     @abstractmethod
     async def get_tokenizer(self) -> AnyTokenizer:
         """Get the tokenizer"""
         ...
 
-    async def get_io_processor(self) -> IOProcessor:
-        raise NotImplementedError
-
     @abstractmethod
     async def is_tracing_enabled(self) -> bool: ...
 

vllm/entrypoints/llm.py

@@ -66,7 +66,6 @@ from vllm.outputs import (
     RequestOutput,
     ScoringRequestOutput,
 )
-from vllm.plugins.io_processors import get_io_processor
 from vllm.pooling_params import PoolingParams
 from vllm.sampling_params import BeamSearchParams, RequestOutputKind, SamplingParams
 from vllm.tasks import PoolingTask
@@ -79,7 +78,6 @@ from vllm.usage.usage_lib import UsageContext
 from vllm.utils import Counter, Device, as_iter, is_list_of
 from vllm.v1.engine import EngineCoreRequest
 from vllm.v1.engine.llm_engine import LLMEngine
-from vllm.v1.engine.processor import Processor
 from vllm.v1.sample.logits_processor import LogitsProcessor
 
 if TYPE_CHECKING:
@@ -335,21 +333,13 @@ class LLM:
         self.request_counter = Counter()
         self.default_sampling_params: Union[dict[str, Any], None] = None
 
-        supported_tasks = self.llm_engine.get_supported_tasks()  # type: ignore
+        supported_tasks = self.llm_engine.get_supported_tasks()
-
+        logger.info("Supported tasks: %s", supported_tasks)
-        logger.info("Supported_tasks: %s", supported_tasks)
-
         self.supported_tasks = supported_tasks
 
-        # Load the Input/Output processor plugin if any
+        self.model_config = self.llm_engine.model_config
-        io_processor_plugin = self.llm_engine.model_config.io_processor_plugin
+        self.processor = self.llm_engine.processor
-        self.io_processor = get_io_processor(
+        self.io_processor = self.llm_engine.io_processor
-            self.llm_engine.vllm_config, io_processor_plugin
-        )
-
-    @property
-    def model_config(self):
-        return self.llm_engine.model_config
 
     def get_tokenizer(self) -> AnyTokenizer:
         return self.llm_engine.get_tokenizer()
@@ -364,18 +354,9 @@ class LLM:
         else:
             self.llm_engine.tokenizer = get_cached_tokenizer(tokenizer)
 
-    def _get_processor(self) -> Processor:
-        if not hasattr(self, "_processor"):
-            vllm_config = self.llm_engine.vllm_config
-            self._processor = Processor(vllm_config)
-
-        return self._processor
-
     def get_default_sampling_params(self) -> SamplingParams:
         if self.default_sampling_params is None:
-            self.default_sampling_params = (
+            self.default_sampling_params = self.model_config.get_diff_sampling_param()
-                self.llm_engine.model_config.get_diff_sampling_param()
-            )
         if self.default_sampling_params:
             return SamplingParams.from_optional(**self.default_sampling_params)
         return SamplingParams()
@@ -423,7 +404,7 @@ class LLM:
             considered legacy and may be deprecated in the future. You should
             instead pass them via the `inputs` parameter.
         """
-        model_config = self.llm_engine.model_config
+        model_config = self.model_config
         runner_type = model_config.runner_type
         if runner_type != "generate":
             raise ValueError(
@@ -463,7 +444,7 @@ class LLM:
         # isn't multimodal, leave the lora as is.
         if (
             lora_config is None
-            or not self.llm_engine.model_config.is_multimodal_model
+            or not self.model_config.is_multimodal_model
             or (lora_config and lora_config.default_mm_loras is None)
         ):
             return lora_request
@@ -495,15 +476,13 @@ class LLM:
         if (
             not default_mm_loras
             or not isinstance(prompt, dict)
-            or "multi_modal_data" not in prompt
+            or not (mm_data := prompt.get("multi_modal_data") or {})
         ):
             return lora_request
 
-        prompt = cast(Union[TextPrompt, TokensPrompt], prompt)
+        intersection = set(
-
+            mm_data.keys()  # type: ignore
-        intersection = set(prompt["multi_modal_data"].keys()).intersection(
+        ).intersection(default_mm_loras.keys())
-            default_mm_loras.keys()
-        )
         if not intersection:
             return lora_request
         if len(intersection) > 1:
@@ -819,7 +798,7 @@ class LLM:
             list_of_messages = [cast(list[ChatCompletionMessageParam], messages)]
 
         tokenizer = self.get_tokenizer()
-        model_config = self.llm_engine.get_model_config()
+        model_config = self.model_config
         resolved_content_format = resolve_chat_template_content_format(
             chat_template,
             tools,
@@ -1031,7 +1010,7 @@ class LLM:
                 pooling_task,
             )
 
-        model_config = self.llm_engine.model_config
+        model_config = self.model_config
         runner_type = model_config.runner_type
         if runner_type != "pooling":
             raise ValueError(
@@ -1276,7 +1255,7 @@ class LLM:
         pooling_params: Optional[PoolingParams] = None,
         lora_request: Optional[Union[list[LoRARequest], LoRARequest]] = None,
     ) -> list[ScoringRequestOutput]:
-        model_config = self.llm_engine.model_config
+        model_config = self.model_config
 
         if isinstance(tokenizer, MistralTokenizer):
             raise ValueError("Score API is not supported for Mistral tokenizer")
@@ -1287,7 +1266,6 @@ class LLM:
         if pooling_params is None:
             pooling_params = PoolingParams(task="score")
 
-        model_config = self.llm_engine.model_config
         pooling_params.verify("score", model_config)
         pooling_params_list = list[PoolingParams]()
 
@@ -1301,8 +1279,6 @@ class LLM:
 
         input_pairs = [(t1, t2) for t1, t2 in zip(data_1, data_2)]
 
-        model_config = self.llm_engine.model_config
-
         for q, d in input_pairs:
             _, engine_prompt = get_score_prompt(
                 model_config=model_config,
@@ -1380,7 +1356,7 @@ class LLM:
             A list of `ScoringRequestOutput` objects containing the
             generated scores in the same order as the input prompts.
         """
-        model_config = self.llm_engine.model_config
+        model_config = self.model_config
         runner_type = model_config.runner_type
         if runner_type != "pooling":
             raise ValueError(
@@ -1658,8 +1634,7 @@ class LLM:
             tokenization_kwargs,
         )
 
-        processor = self._get_processor()
+        engine_request = self.processor.process_inputs(
-        engine_request = processor.process_inputs(
             request_id,
             engine_prompt,
             params,

vllm/entrypoints/openai/api_server.py

@@ -1601,10 +1601,11 @@ def build_app(args: Namespace) -> FastAPI:
 
 async def init_app_state(
     engine_client: EngineClient,
-    vllm_config: VllmConfig,
     state: State,
     args: Namespace,
 ) -> None:
+    vllm_config = engine_client.vllm_config
+
     if args.served_model_name is not None:
         served_model_names = args.served_model_name
     else:
@@ -1622,11 +1623,9 @@ async def init_app_state(
     state.engine_client = engine_client
     state.log_stats = not args.disable_log_stats
     state.vllm_config = vllm_config
-    model_config = vllm_config.model_config
 
     supported_tasks = await engine_client.get_supported_tasks()
-
+    logger.info("Supported tasks: %s", supported_tasks)
-    logger.info("Supported_tasks: %s", supported_tasks)
 
     resolved_chat_template = load_chat_template(args.chat_template)
     if resolved_chat_template is not None:
@@ -1688,7 +1687,6 @@ async def init_app_state(
 
     state.openai_serving_models = OpenAIServingModels(
         engine_client=engine_client,
-        model_config=model_config,
         base_model_paths=base_model_paths,
         lora_modules=lora_modules,
     )
@@ -1696,7 +1694,6 @@ async def init_app_state(
     state.openai_serving_responses = (
         OpenAIServingResponses(
             engine_client,
-            model_config,
             state.openai_serving_models,
             request_logger=request_logger,
             chat_template=resolved_chat_template,
@@ -1717,7 +1714,6 @@ async def init_app_state(
     state.openai_serving_chat = (
         OpenAIServingChat(
             engine_client,
-            model_config,
             state.openai_serving_models,
             args.response_role,
             request_logger=request_logger,
@@ -1740,7 +1736,6 @@ async def init_app_state(
     state.openai_serving_completion = (
         OpenAIServingCompletion(
             engine_client,
-            model_config,
             state.openai_serving_models,
             request_logger=request_logger,
             return_tokens_as_token_ids=args.return_tokens_as_token_ids,
@@ -1754,7 +1749,6 @@ async def init_app_state(
     state.openai_serving_pooling = (
         OpenAIServingPooling(
             engine_client,
-            vllm_config,
             state.openai_serving_models,
             request_logger=request_logger,
             chat_template=resolved_chat_template,
@@ -1768,7 +1762,6 @@ async def init_app_state(
     state.openai_serving_embedding = (
         OpenAIServingEmbedding(
             engine_client,
-            model_config,
             state.openai_serving_models,
             request_logger=request_logger,
             chat_template=resolved_chat_template,
@@ -1782,7 +1775,6 @@ async def init_app_state(
     state.openai_serving_classification = (
         ServingClassification(
             engine_client,
-            model_config,
             state.openai_serving_models,
             request_logger=request_logger,
             log_error_stack=args.log_error_stack,
@@ -1793,7 +1785,6 @@ async def init_app_state(
     state.openai_serving_scores = (
         ServingScores(
             engine_client,
-            model_config,
             state.openai_serving_models,
             request_logger=request_logger,
             log_error_stack=args.log_error_stack,
@@ -1803,7 +1794,6 @@ async def init_app_state(
     )
     state.openai_serving_tokenization = OpenAIServingTokenization(
         engine_client,
-        model_config,
         state.openai_serving_models,
         request_logger=request_logger,
         chat_template=resolved_chat_template,
@@ -1814,7 +1804,6 @@ async def init_app_state(
     state.openai_serving_transcription = (
         OpenAIServingTranscription(
             engine_client,
-            model_config,
             state.openai_serving_models,
             request_logger=request_logger,
             log_error_stack=args.log_error_stack,
@@ -1825,7 +1814,6 @@ async def init_app_state(
     state.openai_serving_translation = (
         OpenAIServingTranslation(
             engine_client,
-            model_config,
             state.openai_serving_models,
             request_logger=request_logger,
             log_error_stack=args.log_error_stack,
@@ -1946,12 +1934,11 @@ async def run_server_worker(
         maybe_register_tokenizer_info_endpoint(args)
         app = build_app(args)
 
-        vllm_config = await engine_client.get_vllm_config()
+        await init_app_state(engine_client, app.state, args)
-        await init_app_state(engine_client, vllm_config, app.state, args)
 
         logger.info(
             "Starting vLLM API server %d on %s",
-            vllm_config.parallel_config._api_process_rank,
+            engine_client.vllm_config.parallel_config._api_process_rank,
             listen_address,
         )
         shutdown_task = await serve_http(

vllm/entrypoints/openai/run_batch.py

@@ -14,7 +14,6 @@ import torch
 from prometheus_client import start_http_server
 from tqdm import tqdm
 
-from vllm.config import VllmConfig
 from vllm.engine.arg_utils import AsyncEngineArgs, optional_type
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.logger import RequestLogger
@@ -328,7 +327,6 @@ async def run_request(
 
 async def run_batch(
     engine_client: EngineClient,
-    vllm_config: VllmConfig,
     args: Namespace,
 ) -> None:
     if args.served_model_name is not None:
@@ -345,22 +343,19 @@ async def run_batch(
         BaseModelPath(name=name, model_path=args.model) for name in served_model_names
     ]
 
-    model_config = vllm_config.model_config
+    model_config = engine_client.model_config
-
     supported_tasks = await engine_client.get_supported_tasks()
-    logger.info("Supported_tasks: %s", supported_tasks)
+    logger.info("Supported tasks: %s", supported_tasks)
 
     # Create the openai serving objects.
     openai_serving_models = OpenAIServingModels(
         engine_client=engine_client,
-        model_config=model_config,
         base_model_paths=base_model_paths,
         lora_modules=None,
     )
     openai_serving_chat = (
         OpenAIServingChat(
             engine_client,
-            model_config,
             openai_serving_models,
             args.response_role,
             request_logger=request_logger,
@@ -374,7 +369,6 @@ async def run_batch(
     openai_serving_embedding = (
         OpenAIServingEmbedding(
             engine_client,
-            model_config,
             openai_serving_models,
             request_logger=request_logger,
             chat_template=None,
@@ -392,7 +386,6 @@ async def run_batch(
     openai_serving_scores = (
         ServingScores(
             engine_client,
-            model_config,
             openai_serving_models,
             request_logger=request_logger,
         )
@@ -509,9 +502,7 @@ async def main(args: Namespace):
         usage_context=UsageContext.OPENAI_BATCH_RUNNER,
         disable_frontend_multiprocessing=False,
     ) as engine_client:
-        vllm_config = await engine_client.get_vllm_config()
+        await run_batch(engine_client, args)
-
-        await run_batch(engine_client, vllm_config, args)
 
 
 if __name__ == "__main__":

vllm/entrypoints/openai/serving_chat.py

@@ -15,7 +15,6 @@ from fastapi import Request
 from openai_harmony import Message as OpenAIMessage
 from pydantic import TypeAdapter
 
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.chat_utils import (
     ChatTemplateContentFormatOption,
@@ -81,7 +80,6 @@ class OpenAIServingChat(OpenAIServing):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         response_role: str,
         *,
@@ -101,7 +99,6 @@ class OpenAIServingChat(OpenAIServing):
     ) -> None:
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             return_tokens_as_token_ids=return_tokens_as_token_ids,
@@ -138,7 +135,7 @@ class OpenAIServingChat(OpenAIServing):
         self.tool_parser: Optional[Callable[[AnyTokenizer], ToolParser]] = None
         if self.enable_auto_tools:
             try:
-                if tool_parser == "pythonic" and model_config.model.startswith(
+                if tool_parser == "pythonic" and self.model_config.model.startswith(
                     "meta-llama/Llama-3.2"
                 ):
                     logger.warning(
@@ -169,7 +166,7 @@ class OpenAIServingChat(OpenAIServing):
         else:
             self.tool_call_id_type = "random"
 
-        self.use_harmony = model_config.hf_config.model_type == "gpt_oss"
+        self.use_harmony = self.model_config.hf_config.model_type == "gpt_oss"
         if self.use_harmony:
             if "stop_token_ids" not in self.default_sampling_params:
                 self.default_sampling_params["stop_token_ids"] = []
@@ -338,7 +335,7 @@ class OpenAIServingChat(OpenAIServing):
                 )
 
                 if isinstance(sampling_params, BeamSearchParams):
-                    generator = self.engine_client.beam_search(
+                    generator = self.beam_search(
                         prompt=engine_prompt,
                         request_id=request_id,
                         params=sampling_params,

vllm/entrypoints/openai/serving_classification.py

@@ -8,7 +8,6 @@ import numpy as np
 from fastapi import Request
 from typing_extensions import override
 
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.logger import RequestLogger
 from vllm.entrypoints.openai.protocol import (
@@ -128,7 +127,6 @@ class ServingClassification(ClassificationMixin):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -136,7 +134,6 @@ class ServingClassification(ClassificationMixin):
     ) -> None:
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             log_error_stack=log_error_stack,

vllm/entrypoints/openai/serving_completion.py

@@ -10,7 +10,6 @@ from typing import Optional, Union, cast
 import jinja2
 from fastapi import Request
 
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.logger import RequestLogger
 from vllm.entrypoints.openai.protocol import (
@@ -44,7 +43,6 @@ class OpenAIServingCompletion(OpenAIServing):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -55,7 +53,6 @@ class OpenAIServingCompletion(OpenAIServing):
     ):
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             return_tokens_as_token_ids=return_tokens_as_token_ids,
@@ -201,7 +198,7 @@ class OpenAIServingCompletion(OpenAIServing):
                 # but pre-commit in CI fails without it.
                 engine_prompt = cast(Union[EmbedsPrompt, TokensPrompt], engine_prompt)
                 if isinstance(sampling_params, BeamSearchParams):
-                    generator = self.engine_client.beam_search(
+                    generator = self.beam_search(
                         prompt=engine_prompt,
                         request_id=request_id,
                         params=sampling_params,

vllm/entrypoints/openai/serving_embedding.py

@@ -10,7 +10,6 @@ import torch
 from fastapi import Request
 from typing_extensions import assert_never, override
 
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.chat_utils import ChatTemplateContentFormatOption
 from vllm.entrypoints.logger import RequestLogger
@@ -597,7 +596,6 @@ class OpenAIServingEmbedding(EmbeddingMixin):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -608,7 +606,6 @@ class OpenAIServingEmbedding(EmbeddingMixin):
     ) -> None:
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             log_error_stack=log_error_stack,

vllm/entrypoints/openai/serving_engine.py

@@ -1,5 +1,6 @@
 # SPDX-License-Identifier: Apache-2.0
 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project
+import asyncio
 import json
 import sys
 import time
@@ -15,17 +16,13 @@ from pydantic import BaseModel, ConfigDict, Field
 from starlette.datastructures import Headers
 from typing_extensions import TypeIs
 
-from vllm.entrypoints.utils import _validate_truncation_size
-from vllm.v1.engine import EngineCoreRequest
-from vllm.v1.engine.processor import Processor
-
 if sys.version_info >= (3, 12):
     from typing import TypedDict
 else:
     from typing_extensions import TypedDict
 
 import vllm.envs as envs
-from vllm.config import ModelConfig
+from vllm.beam_search import BeamSearchSequence, create_sort_beams_key_function
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.chat_utils import (
     ChatCompletionMessageParam,
@@ -68,9 +65,14 @@ from vllm.entrypoints.openai.protocol import (
 from vllm.entrypoints.openai.serving_models import OpenAIServingModels
 from vllm.entrypoints.openai.tool_parsers import ToolParser
 from vllm.entrypoints.renderer import BaseRenderer, CompletionRenderer, RenderConfig
+from vllm.entrypoints.utils import _validate_truncation_size
 from vllm.inputs.data import PromptType
 from vllm.inputs.data import TokensPrompt as EngineTokensPrompt
-from vllm.inputs.parse import PromptComponents, get_prompt_components
+from vllm.inputs.parse import (
+    PromptComponents,
+    get_prompt_components,
+    is_explicit_encoder_decoder_prompt,
+)
 from vllm.logger import init_logger
 from vllm.logprobs import Logprob, PromptLogprobs
 from vllm.lora.request import LoRARequest
@@ -78,7 +80,7 @@ from vllm.multimodal import (  # noqa: F401 - Required to resolve Pydantic error
     MultiModalDataDict,
     MultiModalUUIDDict,
 )
-from vllm.outputs import PoolingRequestOutput, RequestOutput
+from vllm.outputs import CompletionOutput, PoolingRequestOutput, RequestOutput
 from vllm.pooling_params import PoolingParams
 from vllm.sampling_params import BeamSearchParams, SamplingParams
 from vllm.tracing import (
@@ -89,11 +91,13 @@ from vllm.tracing import (
 from vllm.transformers_utils.tokenizer import AnyTokenizer, MistralTokenizer
 from vllm.utils import (
     AsyncMicrobatchTokenizer,
+    collect_from_async_generator,
     is_list_of,
     make_async,
     merge_async_iterators,
     random_uuid,
 )
+from vllm.v1.engine import EngineCoreRequest
 
 logger = init_logger(__name__)
 
@@ -240,7 +244,6 @@ class OpenAIServing:
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -251,8 +254,6 @@ class OpenAIServing:
         super().__init__()
 
         self.engine_client = engine_client
-        self.model_config = model_config
-        self.max_model_len = model_config.max_model_len
 
         self.models = models
 
@@ -268,12 +269,194 @@ class OpenAIServing:
         self._async_tokenizer_pool: dict[AnyTokenizer, AsyncMicrobatchTokenizer] = {}
         self.log_error_stack = log_error_stack
 
-    async def _get_processor(self) -> Processor:
+        self.processor = self.models.processor
-        if not hasattr(self, "_processor"):
+        self.io_processor = self.models.io_processor
-            vllm_config = await self.engine_client.get_vllm_config()
+        self.model_config = self.models.model_config
-            self._processor = Processor(vllm_config)
+        self.max_model_len = self.model_config.max_model_len
 
-        return self._processor
+    async def beam_search(
+        self,
+        prompt: PromptType,
+        request_id: str,
+        params: BeamSearchParams,
+        lora_request: Optional[LoRARequest] = None,
+    ) -> AsyncGenerator[RequestOutput, None]:
+        beam_width = params.beam_width
+        max_tokens = params.max_tokens
+        ignore_eos = params.ignore_eos
+        temperature = params.temperature
+        length_penalty = params.length_penalty
+        include_stop_str_in_output = params.include_stop_str_in_output
+
+        processor = self.processor
+        tokenizer = processor.tokenizer
+        if tokenizer is None:
+            raise ValueError(
+                "You cannot use beam search when `skip_tokenizer_init` is True"
+            )
+
+        eos_token_id: int = tokenizer.eos_token_id  # type: ignore
+
+        if is_explicit_encoder_decoder_prompt(prompt):
+            raise NotImplementedError
+        else:
+            processed_inputs = processor.input_preprocessor._prompt_to_llm_inputs(
+                prompt
+            )
+
+        if processed_inputs["type"] == "embeds":
+            raise NotImplementedError
+
+        # This is a workaround to fix multimodal beam search; this is a
+        # bandaid fix for 2 small problems:
+        # 1. Multi_modal_data on the processed_inputs currently resolves to
+        #    `None`.
+        # 2. preprocessing above expands the multimodal placeholders. However,
+        #    this happens again in generation, so the double expansion causes
+        #    a mismatch.
+        # TODO - would be ideal to handle this more gracefully.
+        prompt_text: Optional[str]
+        prompt_token_ids: list[int]
+        multi_modal_data: Optional[MultiModalDataDict]
+        if isinstance(prompt, str):
+            prompt_text = prompt
+            prompt_token_ids = []
+            multi_modal_data = None
+        else:
+            prompt_text = prompt.get("prompt")  # type: ignore
+            prompt_token_ids = prompt.get("prompt_token_ids", [])  # type: ignore
+            multi_modal_data = prompt.get("multi_modal_data")  # type: ignore
+
+        mm_processor_kwargs: Optional[dict[str, Any]] = processed_inputs.get(
+            "mm_processor_kwargs"
+        )  # type: ignore
+
+        tokenized_length = len(prompt_token_ids)
+
+        sort_beams_key = create_sort_beams_key_function(eos_token_id, length_penalty)
+
+        beam_search_params = SamplingParams(
+            logprobs=2 * beam_width,
+            max_tokens=1,
+            temperature=temperature,
+        )
+        all_beams = [
+            BeamSearchSequence(
+                tokens=prompt_token_ids,
+                cum_logprob=0,
+                logprobs=[],
+                multi_modal_data=multi_modal_data,
+                mm_processor_kwargs=mm_processor_kwargs,
+                lora_request=lora_request,
+            )
+        ]
+        completed = []
+
+        for _ in range(max_tokens):
+            prompts_batch, lora_req_batch = zip(
+                *[
+                    (
+                        EngineTokensPrompt(
+                            prompt_token_ids=beam.tokens,
+                            multi_modal_data=beam.multi_modal_data,
+                            mm_processor_kwargs=beam.mm_processor_kwargs,
+                        ),
+                        beam.lora_request,
+                    )
+                    for beam in all_beams
+                ]
+            )
+
+            tasks = []
+            request_id_batch = f"{request_id}-{random_uuid()}"
+
+            for i, (individual_prompt, lora_req) in enumerate(
+                zip(prompts_batch, lora_req_batch)
+            ):
+                request_id_item = f"{request_id_batch}-beam-{i}"
+                task = asyncio.create_task(
+                    collect_from_async_generator(
+                        self.engine_client.generate(
+                            individual_prompt,
+                            beam_search_params,
+                            request_id_item,
+                            lora_request=lora_req,
+                        )
+                    )
+                )
+                tasks.append(task)
+
+            output = [x[0] for x in await asyncio.gather(*tasks)]
+
+            new_beams = []
+            for i, current_beam in enumerate(all_beams):
+                result = output[i]
+
+                if result.outputs[0].logprobs is not None:
+                    logprobs = result.outputs[0].logprobs[0]
+                    for token_id, logprob_obj in logprobs.items():
+                        if token_id == eos_token_id and not ignore_eos:
+                            completed.append(
+                                BeamSearchSequence(
+                                    tokens=current_beam.tokens + [token_id]
+                                    if include_stop_str_in_output
+                                    else current_beam.tokens,
+                                    logprobs=current_beam.logprobs + [logprobs],
+                                    cum_logprob=current_beam.cum_logprob
+                                    + logprob_obj.logprob,
+                                    finish_reason="stop",
+                                    stop_reason=eos_token_id,
+                                )
+                            )
+                        else:
+                            new_beams.append(
+                                BeamSearchSequence(
+                                    tokens=current_beam.tokens + [token_id],
+                                    logprobs=current_beam.logprobs + [logprobs],
+                                    lora_request=current_beam.lora_request,
+                                    cum_logprob=current_beam.cum_logprob
+                                    + logprob_obj.logprob,
+                                    multi_modal_data=current_beam.multi_modal_data,
+                                    mm_processor_kwargs=current_beam.mm_processor_kwargs,
+                                )
+                            )
+
+            sorted_beams = sorted(new_beams, key=sort_beams_key, reverse=True)
+            all_beams = sorted_beams[:beam_width]
+
+        completed.extend(all_beams)
+        sorted_completed = sorted(completed, key=sort_beams_key, reverse=True)
+        best_beams = sorted_completed[:beam_width]
+
+        for beam in best_beams:
+            if beam.tokens[-1] == eos_token_id and not ignore_eos:
+                # Skip the eos token in the text.
+                tokens = beam.tokens[tokenized_length:-1]
+            else:
+                tokens = beam.tokens[tokenized_length:]
+            beam.text = tokenizer.decode(tokens)
+
+        yield RequestOutput(
+            request_id=request_id,
+            prompt=prompt_text,
+            outputs=[
+                CompletionOutput(
+                    text=beam.text,  # type: ignore
+                    cumulative_logprob=beam.cum_logprob,
+                    token_ids=beam.tokens[tokenized_length:],
+                    index=i,
+                    logprobs=beam.logprobs,
+                    finish_reason=beam.finish_reason
+                    if beam.finish_reason is not None
+                    else "length",
+                    stop_reason=beam.stop_reason,
+                )
+                for (i, beam) in enumerate(best_beams)
+            ],
+            finished=True,
+            prompt_token_ids=prompt_token_ids,
+            prompt_logprobs=None,
+        )
 
     def _get_renderer(self, tokenizer: Optional[AnyTokenizer]) -> BaseRenderer:
         """
@@ -938,8 +1121,7 @@ class OpenAIServing:
             self.max_model_len, params.truncate_prompt_tokens, tokenization_kwargs
         )
 
-        processor = await self._get_processor()
+        engine_request = self.processor.process_inputs(
-        engine_request = processor.process_inputs(
             request_id,
             engine_prompt,
             params,

vllm/entrypoints/openai/serving_models.py

@@ -7,7 +7,6 @@ from dataclasses import dataclass
 from http import HTTPStatus
 from typing import Optional, Union
 
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.openai.protocol import (
     ErrorInfo,
@@ -51,18 +50,14 @@ class OpenAIServingModels:
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         base_model_paths: list[BaseModelPath],
         *,
         lora_modules: Optional[list[LoRAModulePath]] = None,
     ):
         super().__init__()
 
-        self.base_model_paths = base_model_paths
-
-        self.max_model_len = model_config.max_model_len
         self.engine_client = engine_client
-        self.model_config = model_config
+        self.base_model_paths = base_model_paths
 
         self.static_lora_modules = lora_modules
         self.lora_requests: dict[str, LoRARequest] = {}
@@ -75,6 +70,11 @@ class OpenAIServingModels:
             )
         self.lora_resolver_lock: dict[str, Lock] = defaultdict(Lock)
 
+        self.processor = self.engine_client.processor
+        self.io_processor = self.engine_client.io_processor
+        self.model_config = self.engine_client.model_config
+        self.max_model_len = self.model_config.max_model_len
+
     async def init_static_loras(self):
         """Loads all static LoRA modules.
         Raises if any fail to load"""

vllm/entrypoints/openai/serving_pooling.py

@@ -13,7 +13,6 @@ import torch
 from fastapi import Request
 from typing_extensions import assert_never
 
-from vllm.config import VllmConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.chat_utils import ChatTemplateContentFormatOption
 from vllm.entrypoints.logger import RequestLogger
@@ -34,7 +33,6 @@ from vllm.entrypoints.renderer import RenderConfig
 from vllm.entrypoints.utils import _validate_truncation_size
 from vllm.logger import init_logger
 from vllm.outputs import PoolingOutput, PoolingRequestOutput
-from vllm.plugins.io_processors import get_io_processor
 from vllm.utils import merge_async_iterators
 
 logger = init_logger(__name__)
@@ -60,7 +58,6 @@ class OpenAIServingPooling(OpenAIServing):
     def __init__(
         self,
         engine_client: EngineClient,
-        vllm_config: VllmConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -71,7 +68,6 @@ class OpenAIServingPooling(OpenAIServing):
     ) -> None:
         super().__init__(
             engine_client=engine_client,
-            model_config=vllm_config.model_config,
             models=models,
             request_logger=request_logger,
             log_error_stack=log_error_stack,
@@ -80,8 +76,6 @@ class OpenAIServingPooling(OpenAIServing):
         self.chat_template = chat_template
         self.chat_template_content_format: Final = chat_template_content_format
         self.trust_request_chat_template = trust_request_chat_template
-        io_processor_plugin = self.model_config.io_processor_plugin
-        self.io_processor = get_io_processor(vllm_config, io_processor_plugin)
 
     async def create_pooling(
         self,

vllm/entrypoints/openai/serving_responses.py

@@ -49,7 +49,6 @@ from openai.types.responses.response_reasoning_item import (
 from openai_harmony import Message as OpenAIHarmonyMessage
 
 from vllm import envs
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.chat_utils import (
     ChatCompletionMessageParam,
@@ -109,7 +108,6 @@ class OpenAIServingResponses(OpenAIServing):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -127,7 +125,6 @@ class OpenAIServingResponses(OpenAIServing):
     ) -> None:
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             return_tokens_as_token_ids=return_tokens_as_token_ids,
@@ -176,7 +173,7 @@ class OpenAIServingResponses(OpenAIServing):
                 "the store."
             )
 
-        self.use_harmony = model_config.hf_config.model_type == "gpt_oss"
+        self.use_harmony = self.model_config.hf_config.model_type == "gpt_oss"
         if self.use_harmony:
             logger.warning(
                 "For gpt-oss, we ignore --enable-auto-tool-choice "

vllm/entrypoints/openai/serving_score.py

@@ -7,7 +7,6 @@ from typing import Any, Optional, Union
 
 from fastapi import Request
 
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.logger import RequestLogger
 from vllm.entrypoints.openai.protocol import (
@@ -47,7 +46,6 @@ class ServingScores(OpenAIServing):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -55,7 +53,6 @@ class ServingScores(OpenAIServing):
     ) -> None:
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             log_error_stack=log_error_stack,

vllm/entrypoints/openai/serving_tokenization.py

@@ -6,7 +6,6 @@ from typing import Any, Final, Optional, Union
 import jinja2
 from fastapi import Request
 
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.chat_utils import ChatTemplateContentFormatOption
 from vllm.entrypoints.logger import RequestLogger
@@ -32,7 +31,6 @@ class OpenAIServingTokenization(OpenAIServing):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -43,7 +41,6 @@ class OpenAIServingTokenization(OpenAIServing):
     ) -> None:
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             log_error_stack=log_error_stack,

vllm/entrypoints/openai/serving_transcription.py

@@ -5,7 +5,6 @@ from typing import Optional, Union
 
 from fastapi import Request
 
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.logger import RequestLogger
 from vllm.entrypoints.openai.protocol import (
@@ -34,7 +33,6 @@ class OpenAIServingTranscription(OpenAISpeechToText):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -43,7 +41,6 @@ class OpenAIServingTranscription(OpenAISpeechToText):
     ):
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             return_tokens_as_token_ids=return_tokens_as_token_ids,
@@ -95,7 +92,6 @@ class OpenAIServingTranslation(OpenAISpeechToText):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -104,7 +100,6 @@ class OpenAIServingTranslation(OpenAISpeechToText):
     ):
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             return_tokens_as_token_ids=return_tokens_as_token_ids,

vllm/entrypoints/openai/speech_to_text.py

@@ -12,7 +12,6 @@ import numpy as np
 from fastapi import Request
 
 import vllm.envs as envs
-from vllm.config import ModelConfig
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.logger import RequestLogger
 from vllm.entrypoints.openai.protocol import (
@@ -53,7 +52,6 @@ class OpenAISpeechToText(OpenAIServing):
     def __init__(
         self,
         engine_client: EngineClient,
-        model_config: ModelConfig,
         models: OpenAIServingModels,
         *,
         request_logger: Optional[RequestLogger],
@@ -63,7 +61,6 @@ class OpenAISpeechToText(OpenAIServing):
     ):
         super().__init__(
             engine_client=engine_client,
-            model_config=model_config,
             models=models,
             request_logger=request_logger,
             return_tokens_as_token_ids=return_tokens_as_token_ids,
@@ -74,7 +71,7 @@ class OpenAISpeechToText(OpenAIServing):
         self.task_type = task_type
 
         self.asr_config = self.model_cls.get_speech_to_text_config(
-            model_config, task_type
+            self.model_config, task_type
         )
 
         self.max_audio_filesize_mb = envs.VLLM_MAX_AUDIO_CLIP_FILESIZE_MB

vllm/inputs/data.py

@@ -20,13 +20,13 @@ class TextPrompt(TypedDict):
     prompt: str
     """The input text to be tokenized before passing to the model."""
 
-    multi_modal_data: NotRequired["MultiModalDataDict"]
+    multi_modal_data: NotRequired[Optional["MultiModalDataDict"]]
     """
     Optional multi-modal data to pass to the model,
     if the model supports it.
     """
 
-    mm_processor_kwargs: NotRequired[dict[str, Any]]
+    mm_processor_kwargs: NotRequired[Optional[dict[str, Any]]]
     """
     Optional multi-modal processor kwargs to be forwarded to the
     multimodal input mapper & processor. Note that if multiple modalities
@@ -61,13 +61,13 @@ class TokensPrompt(TypedDict):
     token_type_ids: NotRequired[list[int]]
     """A list of token type IDs to pass to the cross encoder model."""
 
-    multi_modal_data: NotRequired["MultiModalDataDict"]
+    multi_modal_data: NotRequired[Optional["MultiModalDataDict"]]
     """
     Optional multi-modal data to pass to the model,
     if the model supports it.
     """
 
-    mm_processor_kwargs: NotRequired[dict[str, Any]]
+    mm_processor_kwargs: NotRequired[Optional[dict[str, Any]]]
     """
     Optional multi-modal processor kwargs to be forwarded to the
     multimodal input mapper & processor. Note that if multiple modalities

vllm/inputs/preprocess.py

@@ -17,7 +17,7 @@ from vllm.multimodal.inputs import (
     MultiModalUUIDDict,
 )
 from vllm.multimodal.processing import BaseMultiModalProcessor
-from vllm.transformers_utils.tokenizer import AnyTokenizer, init_tokenizer_from_configs
+from vllm.transformers_utils.tokenizer import AnyTokenizer
 from vllm.utils.jsontree import json_iter_leaves
 
 from .data import (
@@ -45,20 +45,17 @@ class InputPreprocessor:
     def __init__(
         self,
         model_config: ModelConfig,
+        tokenizer: Optional[AnyTokenizer],
         mm_registry: MultiModalRegistry = MULTIMODAL_REGISTRY,
         mm_processor_cache: Optional[BaseMultiModalProcessorCache] = None,
     ) -> None:
         super().__init__()
 
         self.model_config = model_config
+        self.tokenizer = tokenizer
         self.mm_registry = mm_registry
         self.mm_processor_cache = mm_processor_cache
 
-        if model_config.skip_tokenizer_init:
-            self.tokenizer = None
-        else:
-            self.tokenizer = init_tokenizer_from_configs(model_config)
-
     def get_tokenizer(self) -> AnyTokenizer:
         if self.tokenizer is None:
             raise ValueError(
@@ -351,8 +348,8 @@ class InputPreprocessor:
         if self.model_config.is_multimodal_model:
             inputs = self._process_multimodal(
                 prompt_token_ids,
-                parsed_content.get("multi_modal_data", {}),
+                parsed_content.get("multi_modal_data") or {},
-                parsed_content.get("mm_processor_kwargs"),
+                parsed_content.get("mm_processor_kwargs") or {},
                 tokenization_kwargs=tokenization_kwargs,
                 mm_uuids=mm_uuids,
             )
@@ -380,8 +377,8 @@ class InputPreprocessor:
         if self.model_config.is_multimodal_model:
             inputs = self._process_multimodal(
                 prompt_text,
-                parsed_content.get("multi_modal_data", {}),
+                parsed_content.get("multi_modal_data") or {},
-                parsed_content.get("mm_processor_kwargs"),
+                parsed_content.get("mm_processor_kwargs") or {},
                 tokenization_kwargs=tokenization_kwargs,
                 mm_uuids=mm_uuids,
             )

vllm/v1/engine/async_llm.py

@@ -12,23 +12,23 @@ import numpy as np
 import torch
 
 import vllm.envs as envs
-from vllm.config import ModelConfig, VllmConfig
+from vllm.config import VllmConfig
 from vllm.engine.arg_utils import AsyncEngineArgs
 from vllm.engine.protocol import EngineClient
 from vllm.entrypoints.utils import _validate_truncation_size
 from vllm.envs import VLLM_V1_OUTPUT_PROC_CHUNK_SIZE
 from vllm.inputs import PromptType
-from vllm.inputs.preprocess import InputPreprocessor
 from vllm.logger import init_logger
 from vllm.lora.request import LoRARequest
 from vllm.multimodal import MULTIMODAL_REGISTRY, MultiModalRegistry
 from vllm.outputs import PoolingRequestOutput, RequestOutput
+from vllm.plugins.io_processors import get_io_processor
 from vllm.pooling_params import PoolingParams
 from vllm.sampling_params import SamplingParams
 from vllm.tasks import SupportedTask
 from vllm.tracing import init_tracer
 from vllm.transformers_utils.config import maybe_register_config_serialize_by_value
-from vllm.transformers_utils.tokenizer import AnyTokenizer
+from vllm.transformers_utils.tokenizer import AnyTokenizer, init_tokenizer_from_configs
 from vllm.usage.usage_lib import UsageContext
 from vllm.utils import Device, as_list, cancel_task_threadsafe, cdiv, deprecate_kwargs
 from vllm.v1.engine import EngineCoreRequest
@@ -104,8 +104,16 @@ class AsyncLLM(EngineClient):
                 "logger list; enabling logging without default stat loggers"
             )
 
-        # Processor (converts Inputs --> EngineCoreRequests).
+        if self.model_config.skip_tokenizer_init:
-        self.processor = Processor(vllm_config, mm_registry=mm_registry)
+            tokenizer = None
+        else:
+            tokenizer = init_tokenizer_from_configs(self.model_config)
+
+        self.processor = Processor(self.vllm_config, tokenizer)
+        self.io_processor = get_io_processor(
+            self.vllm_config,
+            self.model_config.io_processor_plugin,
+        )
 
         # OutputProcessor (converts EngineCoreOutputs --> RequestOutput).
         self.output_processor = OutputProcessor(
@@ -245,10 +253,6 @@ class AsyncLLM(EngineClient):
 
         cancel_task_threadsafe(getattr(self, "output_handler", None))
 
-    @property
-    def tokenizer(self) -> Optional[AnyTokenizer]:
-        return self.processor.tokenizer
-
     async def get_supported_tasks(self) -> tuple[SupportedTask, ...]:
         return await self.engine_core.get_supported_tasks_async()
 
@@ -615,14 +619,13 @@ class AsyncLLM(EngineClient):
                 logger.info("Request %s failed.", request_id)
             raise EngineGenerateError() from e
 
-    async def get_vllm_config(self) -> VllmConfig:
+    @property
-        return self.vllm_config
+    def tokenizer(self) -> Optional[AnyTokenizer]:
+        return self.processor.tokenizer
 
-    async def get_model_config(self) -> ModelConfig:
+    @tokenizer.setter
-        return self.model_config
+    def tokenizer(self, tokenizer: Optional[AnyTokenizer]) -> None:
-
+        self.processor.tokenizer = tokenizer
-    async def get_input_preprocessor(self) -> InputPreprocessor:
-        return self.processor.input_preprocessor
 
     async def get_tokenizer(self) -> AnyTokenizer:
         if self.tokenizer is None:

vllm/v1/engine/llm_engine.py

@@ -19,11 +19,12 @@ from vllm.logger import init_logger
 from vllm.lora.request import LoRARequest
 from vllm.multimodal import MULTIMODAL_REGISTRY, MultiModalRegistry
 from vllm.outputs import PoolingRequestOutput, RequestOutput
+from vllm.plugins.io_processors import get_io_processor
 from vllm.pooling_params import PoolingParams
 from vllm.sampling_params import SamplingParams
 from vllm.tasks import SupportedTask
 from vllm.tracing import init_tracer
-from vllm.transformers_utils.tokenizer import AnyTokenizer
+from vllm.transformers_utils.tokenizer import AnyTokenizer, init_tokenizer_from_configs
 from vllm.usage.usage_lib import UsageContext
 from vllm.utils import Device
 from vllm.v1.engine import EngineCoreRequest
@@ -95,8 +96,16 @@ class LLMEngine:
             self.dp_group = None
         self.should_execute_dummy_batch = False
 
-        # Processor (convert Inputs --> EngineCoreRequests)
+        if self.model_config.skip_tokenizer_init:
-        self.processor = Processor(vllm_config, mm_registry=mm_registry)
+            tokenizer = None
+        else:
+            tokenizer = init_tokenizer_from_configs(self.model_config)
+
+        self.processor = Processor(self.vllm_config, tokenizer)
+        self.io_processor = get_io_processor(
+            self.vllm_config,
+            self.model_config.io_processor_plugin,
+        )
 
         # OutputProcessor (convert EngineCoreOutputs --> RequestOutput).
         self.output_processor = OutputProcessor(
@@ -204,14 +213,6 @@ class LLMEngine:
     def validate_outputs(cls, outputs, output_type):
         return outputs
 
-    @property
-    def tokenizer(self) -> Optional[AnyTokenizer]:
-        return self.processor.tokenizer
-
-    @tokenizer.setter
-    def tokenizer(self, tokenizer: Optional[AnyTokenizer]) -> None:
-        self.processor.tokenizer = tokenizer
-
     def get_supported_tasks(self) -> tuple[SupportedTask, ...]:
         return self.engine_core.get_supported_tasks()
 
@@ -313,12 +314,6 @@ class LLMEngine:
 
         return processed_outputs.request_outputs
 
-    def get_vllm_config(self):
-        return self.vllm_config
-
-    def get_model_config(self):
-        return self.model_config
-
     def start_profile(self):
         self.engine_core.profile(True)
 
@@ -345,6 +340,14 @@ class LLMEngine:
         assert self.log_stats, "Stat logging disabled"
         return get_metrics_snapshot()
 
+    @property
+    def tokenizer(self) -> Optional[AnyTokenizer]:
+        return self.processor.tokenizer
+
+    @tokenizer.setter
+    def tokenizer(self, tokenizer: Optional[AnyTokenizer]) -> None:
+        self.processor.tokenizer = tokenizer
+
     def get_tokenizer(self) -> AnyTokenizer:
         if self.tokenizer is None:
             raise ValueError(

vllm/v1/engine/processor.py

@@ -37,6 +37,7 @@ class Processor:
     def __init__(
         self,
         vllm_config: VllmConfig,
+        tokenizer: Optional[AnyTokenizer],
         mm_registry: MultiModalRegistry = MULTIMODAL_REGISTRY,
     ) -> None:
         self.vllm_config = vllm_config
@@ -52,6 +53,7 @@ class Processor:
 
         self.input_preprocessor = InputPreprocessor(
             self.model_config,
+            tokenizer,
             mm_registry,
             mm_processor_cache=self.mm_processor_cache,
         )