vllm.engine.async_llm_engine

ENGINE_ITERATION_TIMEOUT_S `module-attribute` ¶

ENGINE_ITERATION_TIMEOUT_S = VLLM_ENGINE_ITERATION_TIMEOUT_S

STOP_ITERATION `module-attribute` ¶

STOP_ITERATION = Exception()

logger `module-attribute` ¶

logger = init_logger(__name__)

AsyncEngineDeadError ¶

Bases: RuntimeError

Source code in vllm/engine/async_llm_engine.py

class AsyncEngineDeadError(RuntimeError):
    pass

AsyncLLMEngine ¶

Bases: EngineClient

An asynchronous wrapper for LLMEngine.

This class is used to wrap the LLMEngine class to make it asynchronous. It uses asyncio to create a background loop that keeps processing incoming requests. The LLMEngine is kicked by the generate method when there are requests in the waiting queue. The generate method yields the outputs from the LLMEngine to the caller.

Parameters:

Name	Type	Description	Default
`log_requests`	`bool`	Whether to log the requests.	`True`
`start_engine_loop`	`bool`	If True, the background task to run the engine will be automatically started in the generate call.	`True`
`*args`		Arguments for `LLMEngine`.	`()`
`**kwargs`		Arguments for `LLMEngine`.	`{}`

Source code in vllm/engine/async_llm_engine.py

class AsyncLLMEngine(EngineClient):
    """An asynchronous wrapper for [`LLMEngine`][vllm.LLMEngine].

    This class is used to wrap the [`LLMEngine`][vllm.LLMEngine] class to
    make it asynchronous. It uses asyncio to create a background loop that keeps
    processing incoming requests. The [`LLMEngine`][vllm.LLMEngine] is kicked
    by the generate method when there are requests in the waiting queue. The
    generate method yields the outputs from the [`LLMEngine`][vllm.LLMEngine]
    to the caller.

    Args:
        log_requests: Whether to log the requests.
        start_engine_loop: If True, the background task to run the engine
            will be automatically started in the generate call.
        *args: Arguments for [`LLMEngine`][vllm.LLMEngine].
        **kwargs: Arguments for [`LLMEngine`][vllm.LLMEngine].
    """

    _engine_class: Type[_AsyncLLMEngine] = _AsyncLLMEngine

    def __init__(self,
                 *args,
                 log_requests: bool = True,
                 start_engine_loop: bool = True,
                 **kwargs) -> None:
        if envs.VLLM_USE_V1:
            raise ValueError(
                "Using V0 AsyncLLMEngine, but envs.VLLM_USE_V1=True. "
                "This should not happen. As a workaround, try using "
                "AsyncLLMEngine.from_vllm_config(...) or explicitly set "
                "VLLM_USE_V1=0 or 1 and report this issue on Github.")

        self.log_requests = log_requests
        self.engine = self._engine_class(*args, **kwargs)

        # This ensures quick processing of request outputs
        # so the append to asyncio queues is not delayed,
        # especially for multi-step.
        self.use_process_request_outputs_callback = (
            self.engine.model_config.use_async_output_proc)

        if self.use_process_request_outputs_callback:
            self.engine.process_request_outputs_callback = \
                weak_bind(self.process_request_outputs)

        self.background_loop: Optional[asyncio.Future] = None
        # We need to keep a reference to unshielded
        # task as well to prevent it from being garbage
        # collected
        self._background_loop_unshielded: Optional[asyncio.Task] = None
        self.start_engine_loop = start_engine_loop
        self._errored_with: Optional[BaseException] = None

        # Lazy initialized fields
        self._request_tracker: RequestTracker

    def __del__(self):
        if rt := getattr(self, "request_tracker", None):
            # Wake up engine loop so that it will exit cleanly
            rt.new_requests_event.set()

    @classmethod
    def _get_executor_cls(cls,
                          engine_config: VllmConfig) -> Type[ExecutorBase]:
        return LLMEngine._get_executor_cls(engine_config)

    @classmethod
    def from_vllm_config(
        cls,
        vllm_config: VllmConfig,
        start_engine_loop: bool = True,
        usage_context: UsageContext = UsageContext.ENGINE_CONTEXT,
        stat_loggers: Optional[dict[str, StatLoggerBase]] = None,
        disable_log_requests: bool = False,
        disable_log_stats: bool = False,
    ) -> "AsyncLLMEngine":
        """Create an AsyncLLMEngine from the EngineArgs."""

        return cls(
            vllm_config=vllm_config,
            executor_class=cls._get_executor_cls(vllm_config),
            start_engine_loop=start_engine_loop,
            log_requests=not disable_log_requests,
            log_stats=not disable_log_stats,
            usage_context=usage_context,
            stat_loggers=stat_loggers,
        )

    @classmethod
    def from_engine_args(
        cls,
        engine_args: AsyncEngineArgs,
        start_engine_loop: bool = True,
        usage_context: UsageContext = UsageContext.ENGINE_CONTEXT,
        stat_loggers: Optional[Dict[str, StatLoggerBase]] = None,
    ) -> "AsyncLLMEngine":
        """Creates an async LLM engine from the engine arguments."""

        vllm_config = engine_args.create_engine_config(usage_context)

        async_engine_cls = cls
        if envs.VLLM_USE_V1:
            from vllm.v1.engine.async_llm import AsyncLLM as V1AsyncLLMEngine
            async_engine_cls = V1AsyncLLMEngine

        return async_engine_cls.from_vllm_config(
            vllm_config=vllm_config,
            start_engine_loop=start_engine_loop,
            usage_context=usage_context,
            stat_loggers=stat_loggers,
            disable_log_stats=engine_args.disable_log_stats,
            disable_log_requests=engine_args.disable_log_requests,
        )

    @property
    def is_running(self) -> bool:
        return (self.background_loop is not None
                and self._background_loop_unshielded is not None
                and not self._background_loop_unshielded.done())

    @property
    def is_stopped(self) -> bool:
        return self.errored or (self.background_loop is not None and
                                self._background_loop_unshielded is not None
                                and self._background_loop_unshielded.done())

    @property
    def errored(self) -> bool:
        return self._errored_with is not None

    @property
    def dead_error(self) -> BaseException:
        return AsyncEngineDeadError(
            "Background loop is not running. If it was running, "
            "inspect the output to find the stacktrace of the "
            "error that caused the background loop to stop "
            "(AsyncEngineDeadError).")

    def set_errored(self, exc: Exception) -> None:
        self._errored_with = exc

    def _error_callback(self, exc: Exception) -> None:
        self.set_errored(exc)
        self._request_tracker.propagate_exception(exc)

    async def get_input_preprocessor(self) -> InputPreprocessor:
        return self.engine.input_preprocessor

    async def get_tokenizer(
        self,
        lora_request: Optional[LoRARequest] = None,
    ) -> AnyTokenizer:
        return await self.engine.get_tokenizer_async(lora_request)

    def start_background_loop(self) -> None:
        """Start the background loop."""
        if self.errored:
            raise AsyncEngineDeadError(
                "Background loop has errored already.") from self._errored_with
        if self.is_running:
            raise RuntimeError("Background loop is already running.")
        # Initialize the RequestTracker here so it uses the right event loop.
        self._request_tracker = RequestTracker()

        self._background_loop_unshielded = asyncio.get_event_loop(
        ).create_task(self.run_engine_loop(weakref.ref(self)))
        self._background_loop_unshielded.add_done_callback(
            partial(_log_task_completion, error_callback=self._error_callback))
        self.background_loop = asyncio.shield(self._background_loop_unshielded)

    def shutdown_background_loop(self) -> None:
        """
        Shut down the background loop.

        This method needs to be called during cleanup to remove
        references to `self` and properly GC the resources held
        by the async LLM engine (e.g., the executors as well as
        their resources).
        """
        if self._background_loop_unshielded is not None:
            self._background_loop_unshielded.cancel()
            self._background_loop_unshielded = None
        self.background_loop = None

    async def engine_step(self, virtual_engine: int) -> bool:
        """Kick the engine to process the waiting requests.

        Returns True if there are in-progress requests."""

        new_requests, aborted_requests = (
            self._request_tracker.get_new_and_aborted_requests())

        for new_request in new_requests:
            # Add the request into the vLLM engine's waiting queue.
            try:
                await self.engine.add_request_async(**new_request)
            except ValueError as e:
                # TODO: use a vLLM specific error for failed validation
                self._request_tracker.process_exception(
                    new_request["request_id"],
                    e,
                    verbose=self.log_requests,
                )

        if aborted_requests:
            await self._engine_abort(aborted_requests)

        request_outputs = await self.engine.step_async(virtual_engine)

        # Put the outputs into the corresponding streams.
        # If used as a callback, then already invoked inside
        # LLMEngine's _process_model_outputs
        if not self.use_process_request_outputs_callback:
            all_finished = self.process_request_outputs(request_outputs)
        else:
            # For callback case, we only need to detect when all
            # requests are finished
            all_finished = all(request_output.finished
                               for request_output in request_outputs)

        return not all_finished

    def process_request_outputs(self, request_outputs) -> bool:
        # Put the outputs into the corresponding streams.
        all_finished = True
        for request_output in request_outputs:
            self._request_tracker.process_request_output(
                request_output, verbose=self.log_requests)
            all_finished = all_finished and request_output.finished

        return all_finished

    async def _engine_abort(self, request_ids: Iterable[str]):
        self.engine.abort_request(request_ids)

    @staticmethod
    async def run_engine_loop(engine_ref: ReferenceType):
        """We use a weakref to the engine so that the running loop
        doesn't prevent the engine being garbage collected."""
        engine: Optional[AsyncLLMEngine] = engine_ref()
        if not engine:
            return

        pipeline_parallel_size = \
                engine.engine.parallel_config.pipeline_parallel_size
        has_requests_in_progress = [False] * pipeline_parallel_size
        while True:
            if not any(has_requests_in_progress):
                logger.debug("Waiting for new requests...")
                # Stop the execute model loop in parallel workers until there
                # are more requests to process. This avoids waiting
                # indefinitely in torch.distributed ops which may otherwise
                # timeout, and unblocks the RPC thread in the workers so that
                # they can process any other queued control plane messages,
                # such as add/remove lora adapters.
                await engine.engine.stop_remote_worker_execution_loop_async()
                request_tracker = engine._request_tracker
                # Allow engine to be garbage collected while
                # waiting for new requests
                del engine
                await asyncio.sleep(0)
                if engine_ref() is None:
                    return
                await request_tracker.wait_for_new_requests()
                engine = engine_ref()
                if not engine:
                    return
                logger.debug("Got new requests!")
                requests_in_progress = [
                    asyncio.create_task(engine.engine_step(ve))
                    for ve in range(pipeline_parallel_size)
                ]
                has_requests_in_progress = [True] * pipeline_parallel_size

            # Abort if iteration takes too long due to unrecoverable errors
            # (eg. NCCL timeouts).
            try:
                async with asyncio_timeout(ENGINE_ITERATION_TIMEOUT_S):
                    done, _ = await asyncio.wait(
                        requests_in_progress,
                        return_when=asyncio.FIRST_COMPLETED)
                    for _ in range(pipeline_parallel_size):
                        await asyncio.sleep(0)
                for task in done:
                    result = task.result()
                    virtual_engine = requests_in_progress.index(task)
                    has_unfinished_requests = (
                        engine.engine.
                        has_unfinished_requests_for_virtual_engine(
                            virtual_engine))
                    if result or has_unfinished_requests:
                        requests_in_progress[virtual_engine] = (
                            asyncio.create_task(
                                engine.engine_step(virtual_engine)))
                        has_requests_in_progress[virtual_engine] = True
                    else:
                        has_requests_in_progress[virtual_engine] = False
            except asyncio.TimeoutError as exc:
                logger.error(
                    "Engine iteration timed out. This should never happen!")
                engine.set_errored(exc)
                raise
            await asyncio.sleep(0)

    async def add_request(
        self,
        request_id: str,
        prompt: PromptType,
        params: Union[SamplingParams, PoolingParams],
        arrival_time: Optional[float] = None,
        lora_request: Optional[LoRARequest] = None,
        trace_headers: Optional[Mapping[str, str]] = None,
        prompt_adapter_request: Optional[PromptAdapterRequest] = None,
        priority: int = 0,
        data_parallel_rank: Optional[int] = None,
    ) -> AsyncGenerator[Union[RequestOutput, PoolingRequestOutput], None]:
        if not self.is_running:
            if self.start_engine_loop:
                self.start_background_loop()
            else:
                raise AsyncEngineDeadError(
                    "Background loop is not running. If it was running, "
                    "inspect the output to find the stacktrace of the "
                    "error that caused the background loop to stop "
                    "(AsyncEngineDeadError).")

        if (priority != 0
                and not self.engine.scheduler_config.policy == "priority"):
            raise ValueError(f"Got priority {priority} but "
                             "Priority scheduling is not enabled.")

        stream = self._request_tracker.add_request(
            request_id,
            verbose=self.log_requests,
            prompt=prompt,
            params=params,
            arrival_time=arrival_time or time.time(),
            lora_request=lora_request,
            trace_headers=trace_headers,
            prompt_adapter_request=prompt_adapter_request,
            priority=priority,
            data_parallel_rank=data_parallel_rank,
        )

        return stream.generator()

    async def generate(
        self,
        prompt: PromptType,
        sampling_params: SamplingParams,
        request_id: str,
        lora_request: Optional[LoRARequest] = None,
        trace_headers: Optional[Mapping[str, str]] = None,
        prompt_adapter_request: Optional[PromptAdapterRequest] = None,
        priority: int = 0,
        data_parallel_rank: Optional[int] = None,
    ) -> AsyncGenerator[RequestOutput, None]:
        """Generate outputs for a request.

        Generate outputs for a request. This method is a coroutine. It adds the
        request into the waiting queue of the LLMEngine and streams the outputs
        from the LLMEngine to the caller.

        Args:
            prompt: The prompt to the LLM. See
                [`PromptType`][vllm.inputs.PromptType] for more details about
                the format of each input.
            sampling_params: The sampling parameters of the request.
            request_id: The unique id of the request.
            lora_request: LoRA request to use for generation, if any.
            trace_headers: OpenTelemetry trace headers.
            prompt_adapter_request: Prompt Adapter request to use
                                            for generation, if any.
            priority: The priority of the request.
                Only applicable with priority scheduling.
            data_parallel_rank: The (global) data parallel rank that must
                handle this request. Only applicable if DP is enabled.
        Yields:
            The output `RequestOutput` objects from the LLMEngine
            for the request.

        Details:
            - If the engine is not running, start the background loop,
              which iteratively invokes
              [`engine_step`][vllm.engine.async_llm_engine.AsyncLLMEngine.engine_step]
              to process the waiting requests.
            - Add the request to the engine's `RequestTracker`.
              On the next background loop, this request will be sent to
              the underlying engine.
              Also, a corresponding `AsyncStream` will be created.
            - Wait for the request outputs from `AsyncStream` and yield them.

        Example:
            >>> # Please refer to entrypoints/api_server.py for
            >>> # the complete example.
            >>>
            >>> # initialize the engine and the example input
            >>> # note that engine_args here is AsyncEngineArgs instance
            >>> engine = AsyncLLMEngine.from_engine_args(engine_args)
            >>> example_input = {
            >>>     "prompt": "What is LLM?",
            >>>     "stream": False, # assume the non-streaming case
            >>>     "temperature": 0.0,
            >>>     "request_id": 0,
            >>> }
            >>>
            >>> # start the generation
            >>> results_generator = engine.generate(
            >>>    example_input["prompt"],
            >>>    SamplingParams(temperature=example_input["temperature"]),
            >>>    example_input["request_id"])
            >>>
            >>> # get the results
            >>> final_output = None
            >>> async for request_output in results_generator:
            >>>     if await request.is_disconnected():
            >>>         # Abort the request if the client disconnects.
            >>>         await engine.abort(request_id)
            >>>         # Return or raise an error
            >>>         ...
            >>>     final_output = request_output
            >>>
            >>> # Process and return the final output
            >>> ...
        """
        try:
            async for output in await self.add_request(
                    request_id,
                    prompt,
                    sampling_params,
                    lora_request=lora_request,
                    trace_headers=trace_headers,
                    prompt_adapter_request=prompt_adapter_request,
                    priority=priority,
                    data_parallel_rank=data_parallel_rank,
            ):
                yield LLMEngine.validate_output(output, RequestOutput)
        except asyncio.CancelledError:
            await self.abort(request_id)
            raise

    async def encode(
        self,
        prompt: PromptType,
        pooling_params: PoolingParams,
        request_id: str,
        lora_request: Optional[LoRARequest] = None,
        trace_headers: Optional[Mapping[str, str]] = None,
        priority: int = 0,
    ) -> AsyncGenerator[PoolingRequestOutput, None]:
        """Generate outputs for a request from a pooling model.

        Generate outputs for a request. This method is a coroutine. It adds the
        request into the waiting queue of the LLMEngine and streams the outputs
        from the LLMEngine to the caller.

        Args:
            prompt: The prompt to the LLM. See
                [`PromptType`][vllm.inputs.PromptType] for more details about
                the format of each input.
            pooling_params: The pooling parameters of the request.
            request_id: The unique id of the request.
            lora_request: LoRA request to use for generation, if any.
            trace_headers: OpenTelemetry trace headers.
            priority: The priority of the request.
                Only applicable with priority scheduling.

        Yields:
            The output `PoolingRequestOutput` objects from the LLMEngine
            for the request.

        Details:
            - If the engine is not running, start the background loop,
                which iteratively invokes
                [`vllm.engine.async_llm_engine.AsyncLLMEngine.engine_step`][]
                to process the waiting requests.
            - Add the request to the engine's `RequestTracker`.
                On the next background loop, this request will be sent to
                the underlying engine.
                Also, a corresponding `AsyncStream` will be created.
            - Wait for the request outputs from `AsyncStream` and yield them.

        Example:
        ```
        # Please refer to entrypoints/api_server.py for
        # the complete example.

        # initialize the engine and the example input
        # note that engine_args here is AsyncEngineArgs instance
        engine = AsyncLLMEngine.from_engine_args(engine_args)
        example_input = {
            "input": "What is LLM?",
            "request_id": 0,
        }

        # start the generation
        results_generator = engine.encode(
        example_input["input"],
        PoolingParams(),
        example_input["request_id"])

        # get the results
        final_output = None
        async for request_output in results_generator:
            if await request.is_disconnected():
                # Abort the request if the client disconnects.
                await engine.abort(request_id)
                # Return or raise an error
                ...
            final_output = request_output

        # Process and return the final output
        ...
        ```
        """
        try:
            async for output in await self.add_request(
                    request_id,
                    prompt,
                    pooling_params,
                    lora_request=lora_request,
                    trace_headers=trace_headers,
                    priority=priority,
            ):
                yield LLMEngine.validate_output(output, PoolingRequestOutput)
        except asyncio.CancelledError:
            await self.abort(request_id)
            raise

    async def abort(self, request_id: str) -> None:
        """Abort a request.

        Abort a submitted request. If the request is finished or not found,
        this method will be a no-op.

        Args:
            request_id: The unique id of the request.
        """
        if not self.is_running:
            raise AsyncEngineDeadError(
                "Background loop is not running. If it was running, "
                "inspect the output to find the stacktrace of the "
                "error that caused the background loop to stop "
                "(AsyncEngineDeadError).")

        return self._abort(request_id)

    def _abort(self, request_id: str) -> None:
        """Abort a request.

        Abort a submitted request. If the request is finished or not found,
        this method will be a no-op.

        Args:
            request_id: The unique id of the request.
        """
        self._request_tracker.abort_request(request_id,
                                            exception=asyncio.CancelledError,
                                            verbose=self.log_requests)

    async def get_vllm_config(self) -> VllmConfig:
        """Get the vllm configuration of the vLLM engine."""
        return self.engine.get_vllm_config()

    async def get_model_config(self) -> ModelConfig:
        """Get the model configuration of the vLLM engine."""
        return self.engine.get_model_config()

    async def get_parallel_config(self) -> ParallelConfig:
        """Get the parallel configuration of the vLLM engine."""
        return self.engine.get_parallel_config()

    async def get_decoding_config(self) -> DecodingConfig:
        """Get the decoding configuration of the vLLM engine."""
        return self.engine.get_decoding_config()

    async def get_scheduler_config(self) -> SchedulerConfig:
        """Get the scheduling configuration of the vLLM engine."""
        return self.engine.get_scheduler_config()

    async def get_lora_config(self) -> LoRAConfig:
        """Get the lora configuration of the vLLM engine."""
        return self.engine.get_lora_config()

    async def do_log_stats(
            self,
            scheduler_outputs: Optional[SchedulerOutputs] = None,
            model_output: Optional[List[SamplerOutput]] = None) -> None:
        self.engine.do_log_stats()

    async def check_health(self) -> None:
        """Raises an error if engine is unhealthy."""
        t = time.perf_counter()
        logger.debug("Starting health check...")
        if self.is_stopped:
            raise AsyncEngineDeadError("Background loop is stopped.")

        await self.engine.check_health_async()
        logger.debug("Health check took %fs", time.perf_counter() - t)

    async def is_tracing_enabled(self) -> bool:
        return self.engine.is_tracing_enabled()

    def add_logger(self, logger_name: str, logger: StatLoggerBase) -> None:
        self.engine.add_logger(logger_name=logger_name, logger=logger)

    def remove_logger(self, logger_name: str) -> None:
        self.engine.remove_logger(logger_name=logger_name)

    async def start_profile(self) -> None:
        self.engine.start_profile()

    async def stop_profile(self) -> None:
        self.engine.stop_profile()

    async def reset_mm_cache(self) -> None:
        self.engine.reset_mm_cache()

    async def reset_prefix_cache(self,
                                 device: Optional[Device] = None) -> None:
        self.engine.reset_prefix_cache(device)

    async def sleep(self, level: int = 1) -> None:
        self.engine.sleep(level)

    async def wake_up(self, tags: Optional[list[str]] = None) -> None:
        self.engine.wake_up(tags)

    async def is_sleeping(self) -> bool:
        return self.engine.is_sleeping()

    async def add_lora(self, lora_request: LoRARequest) -> None:
        self.engine.add_lora(lora_request)

    async def collective_rpc(self,
                             method: str,
                             timeout: Optional[float] = None,
                             args: tuple = (),
                             kwargs: Optional[dict] = None):
        """
        Perform a collective RPC call to the given path.
        """
        return await self.engine.collective_rpc_async(method, timeout, args,
                                                      kwargs)

_background_loop_unshielded `instance-attribute` ¶

_background_loop_unshielded: Optional[Task] = None

_engine_class `class-attribute` `instance-attribute` ¶

_engine_class: Type[_AsyncLLMEngine] = _AsyncLLMEngine

_errored_with `instance-attribute` ¶

_errored_with: Optional[BaseException] = None

_request_tracker `instance-attribute` ¶

_request_tracker: RequestTracker

background_loop `instance-attribute` ¶

background_loop: Optional[Future] = None

dead_error `property` ¶

dead_error: BaseException

engine `instance-attribute` ¶

engine = _engine_class(*args, **kwargs)

errored `property` ¶

errored: bool

is_running `property` ¶

is_running: bool

is_stopped `property` ¶

is_stopped: bool

log_requests `instance-attribute` ¶

log_requests = log_requests

start_engine_loop `instance-attribute` ¶

start_engine_loop = start_engine_loop

use_process_request_outputs_callback `instance-attribute` ¶

use_process_request_outputs_callback = use_async_output_proc

del ¶

__del__()

Source code in vllm/engine/async_llm_engine.py

def __del__(self):
    if rt := getattr(self, "request_tracker", None):
        # Wake up engine loop so that it will exit cleanly
        rt.new_requests_event.set()

init ¶

__init__(
    *args,
    log_requests: bool = True,
    start_engine_loop: bool = True,
    **kwargs,
) -> None

Source code in vllm/engine/async_llm_engine.py

def __init__(self,
             *args,
             log_requests: bool = True,
             start_engine_loop: bool = True,
             **kwargs) -> None:
    if envs.VLLM_USE_V1:
        raise ValueError(
            "Using V0 AsyncLLMEngine, but envs.VLLM_USE_V1=True. "
            "This should not happen. As a workaround, try using "
            "AsyncLLMEngine.from_vllm_config(...) or explicitly set "
            "VLLM_USE_V1=0 or 1 and report this issue on Github.")

    self.log_requests = log_requests
    self.engine = self._engine_class(*args, **kwargs)

    # This ensures quick processing of request outputs
    # so the append to asyncio queues is not delayed,
    # especially for multi-step.
    self.use_process_request_outputs_callback = (
        self.engine.model_config.use_async_output_proc)

    if self.use_process_request_outputs_callback:
        self.engine.process_request_outputs_callback = \
            weak_bind(self.process_request_outputs)

    self.background_loop: Optional[asyncio.Future] = None
    # We need to keep a reference to unshielded
    # task as well to prevent it from being garbage
    # collected
    self._background_loop_unshielded: Optional[asyncio.Task] = None
    self.start_engine_loop = start_engine_loop
    self._errored_with: Optional[BaseException] = None

    # Lazy initialized fields
    self._request_tracker: RequestTracker

_abort ¶

_abort(request_id: str) -> None

Abort a request.

Abort a submitted request. If the request is finished or not found, this method will be a no-op.

Parameters:

Name	Type	Description	Default
`request_id`	`str`	The unique id of the request.	required

Source code in vllm/engine/async_llm_engine.py

def _abort(self, request_id: str) -> None:
    """Abort a request.

    Abort a submitted request. If the request is finished or not found,
    this method will be a no-op.

    Args:
        request_id: The unique id of the request.
    """
    self._request_tracker.abort_request(request_id,
                                        exception=asyncio.CancelledError,
                                        verbose=self.log_requests)

_engine_abort `async` ¶

_engine_abort(request_ids: Iterable[str])

Source code in vllm/engine/async_llm_engine.py

async def _engine_abort(self, request_ids: Iterable[str]):
    self.engine.abort_request(request_ids)

_error_callback ¶

_error_callback(exc: Exception) -> None

Source code in vllm/engine/async_llm_engine.py

def _error_callback(self, exc: Exception) -> None:
    self.set_errored(exc)
    self._request_tracker.propagate_exception(exc)

_get_executor_cls `classmethod` ¶

_get_executor_cls(
    engine_config: VllmConfig,
) -> Type[ExecutorBase]

Source code in vllm/engine/async_llm_engine.py

@classmethod
def _get_executor_cls(cls,
                      engine_config: VllmConfig) -> Type[ExecutorBase]:
    return LLMEngine._get_executor_cls(engine_config)

abort `async` ¶

abort(request_id: str) -> None

Abort a request.

Abort a submitted request. If the request is finished or not found, this method will be a no-op.

Parameters:

Name	Type	Description	Default
`request_id`	`str`	The unique id of the request.	required

Source code in vllm/engine/async_llm_engine.py

async def abort(self, request_id: str) -> None:
    """Abort a request.

    Abort a submitted request. If the request is finished or not found,
    this method will be a no-op.

    Args:
        request_id: The unique id of the request.
    """
    if not self.is_running:
        raise AsyncEngineDeadError(
            "Background loop is not running. If it was running, "
            "inspect the output to find the stacktrace of the "
            "error that caused the background loop to stop "
            "(AsyncEngineDeadError).")

    return self._abort(request_id)

add_logger ¶

add_logger(
    logger_name: str, logger: StatLoggerBase
) -> None

Source code in vllm/engine/async_llm_engine.py

def add_logger(self, logger_name: str, logger: StatLoggerBase) -> None:
    self.engine.add_logger(logger_name=logger_name, logger=logger)

add_lora `async` ¶

add_lora(lora_request: LoRARequest) -> None

Source code in vllm/engine/async_llm_engine.py

async def add_lora(self, lora_request: LoRARequest) -> None:
    self.engine.add_lora(lora_request)

add_request `async` ¶

add_request(
    request_id: str,
    prompt: PromptType,
    params: Union[SamplingParams, PoolingParams],
    arrival_time: Optional[float] = None,
    lora_request: Optional[LoRARequest] = None,
    trace_headers: Optional[Mapping[str, str]] = None,
    prompt_adapter_request: Optional[
        PromptAdapterRequest
    ] = None,
    priority: int = 0,
    data_parallel_rank: Optional[int] = None,
) -> AsyncGenerator[
    Union[RequestOutput, PoolingRequestOutput], None
]

Source code in vllm/engine/async_llm_engine.py

async def add_request(
    self,
    request_id: str,
    prompt: PromptType,
    params: Union[SamplingParams, PoolingParams],
    arrival_time: Optional[float] = None,
    lora_request: Optional[LoRARequest] = None,
    trace_headers: Optional[Mapping[str, str]] = None,
    prompt_adapter_request: Optional[PromptAdapterRequest] = None,
    priority: int = 0,
    data_parallel_rank: Optional[int] = None,
) -> AsyncGenerator[Union[RequestOutput, PoolingRequestOutput], None]:
    if not self.is_running:
        if self.start_engine_loop:
            self.start_background_loop()
        else:
            raise AsyncEngineDeadError(
                "Background loop is not running. If it was running, "
                "inspect the output to find the stacktrace of the "
                "error that caused the background loop to stop "
                "(AsyncEngineDeadError).")

    if (priority != 0
            and not self.engine.scheduler_config.policy == "priority"):
        raise ValueError(f"Got priority {priority} but "
                         "Priority scheduling is not enabled.")

    stream = self._request_tracker.add_request(
        request_id,
        verbose=self.log_requests,
        prompt=prompt,
        params=params,
        arrival_time=arrival_time or time.time(),
        lora_request=lora_request,
        trace_headers=trace_headers,
        prompt_adapter_request=prompt_adapter_request,
        priority=priority,
        data_parallel_rank=data_parallel_rank,
    )

    return stream.generator()

check_health `async` ¶

check_health() -> None

Raises an error if engine is unhealthy.

Source code in vllm/engine/async_llm_engine.py

async def check_health(self) -> None:
    """Raises an error if engine is unhealthy."""
    t = time.perf_counter()
    logger.debug("Starting health check...")
    if self.is_stopped:
        raise AsyncEngineDeadError("Background loop is stopped.")

    await self.engine.check_health_async()
    logger.debug("Health check took %fs", time.perf_counter() - t)

collective_rpc `async` ¶

collective_rpc(
    method: str,
    timeout: Optional[float] = None,
    args: tuple = (),
    kwargs: Optional[dict] = None,
)

Perform a collective RPC call to the given path.

Source code in vllm/engine/async_llm_engine.py

async def collective_rpc(self,
                         method: str,
                         timeout: Optional[float] = None,
                         args: tuple = (),
                         kwargs: Optional[dict] = None):
    """
    Perform a collective RPC call to the given path.
    """
    return await self.engine.collective_rpc_async(method, timeout, args,
                                                  kwargs)

do_log_stats `async` ¶

do_log_stats(
    scheduler_outputs: Optional[SchedulerOutputs] = None,
    model_output: Optional[List[SamplerOutput]] = None,
) -> None

Source code in vllm/engine/async_llm_engine.py

async def do_log_stats(
        self,
        scheduler_outputs: Optional[SchedulerOutputs] = None,
        model_output: Optional[List[SamplerOutput]] = None) -> None:
    self.engine.do_log_stats()

encode `async` ¶

encode(
    prompt: PromptType,
    pooling_params: PoolingParams,
    request_id: str,
    lora_request: Optional[LoRARequest] = None,
    trace_headers: Optional[Mapping[str, str]] = None,
    priority: int = 0,
) -> AsyncGenerator[PoolingRequestOutput, None]

Generate outputs for a request from a pooling model.

Generate outputs for a request. This method is a coroutine. It adds the request into the waiting queue of the LLMEngine and streams the outputs from the LLMEngine to the caller.

Parameters:

Name	Type	Description	Default
`prompt`	`PromptType`	The prompt to the LLM. See `PromptType` for more details about the format of each input.	required
`pooling_params`	`PoolingParams`	The pooling parameters of the request.	required
`request_id`	`str`	The unique id of the request.	required
`lora_request`	`Optional[LoRARequest]`	LoRA request to use for generation, if any.	`None`
`trace_headers`	`Optional[Mapping[str, str]]`	OpenTelemetry trace headers.	`None`
`priority`	`int`	The priority of the request. Only applicable with priority scheduling.	`0`

Yields:

Type	Description
`AsyncGenerator[PoolingRequestOutput, None]`	The output `PoolingRequestOutput` objects from the LLMEngine
`AsyncGenerator[PoolingRequestOutput, None]`	for the request.

Details

If the engine is not running, start the background loop, which iteratively invokes vllm.engine.async_llm_engine.AsyncLLMEngine.engine_step to process the waiting requests.
Add the request to the engine's RequestTracker. On the next background loop, this request will be sent to the underlying engine. Also, a corresponding AsyncStream will be created.
Wait for the request outputs from AsyncStream and yield them.

Example:

# Please refer to entrypoints/api_server.py for
# the complete example.

# initialize the engine and the example input
# note that engine_args here is AsyncEngineArgs instance
engine = AsyncLLMEngine.from_engine_args(engine_args)
example_input = {
    "input": "What is LLM?",
    "request_id": 0,
}

# start the generation
results_generator = engine.encode(
example_input["input"],
PoolingParams(),
example_input["request_id"])

# get the results
final_output = None
async for request_output in results_generator:
    if await request.is_disconnected():
        # Abort the request if the client disconnects.
        await engine.abort(request_id)
        # Return or raise an error
        ...
    final_output = request_output

# Process and return the final output
...

Source code in vllm/engine/async_llm_engine.py

async def encode(
    self,
    prompt: PromptType,
    pooling_params: PoolingParams,
    request_id: str,
    lora_request: Optional[LoRARequest] = None,
    trace_headers: Optional[Mapping[str, str]] = None,
    priority: int = 0,
) -> AsyncGenerator[PoolingRequestOutput, None]:
    """Generate outputs for a request from a pooling model.

    Generate outputs for a request. This method is a coroutine. It adds the
    request into the waiting queue of the LLMEngine and streams the outputs
    from the LLMEngine to the caller.

    Args:
        prompt: The prompt to the LLM. See
            [`PromptType`][vllm.inputs.PromptType] for more details about
            the format of each input.
        pooling_params: The pooling parameters of the request.
        request_id: The unique id of the request.
        lora_request: LoRA request to use for generation, if any.
        trace_headers: OpenTelemetry trace headers.
        priority: The priority of the request.
            Only applicable with priority scheduling.

    Yields:
        The output `PoolingRequestOutput` objects from the LLMEngine
        for the request.

    Details:
        - If the engine is not running, start the background loop,
            which iteratively invokes
            [`vllm.engine.async_llm_engine.AsyncLLMEngine.engine_step`][]
            to process the waiting requests.
        - Add the request to the engine's `RequestTracker`.
            On the next background loop, this request will be sent to
            the underlying engine.
            Also, a corresponding `AsyncStream` will be created.
        - Wait for the request outputs from `AsyncStream` and yield them.

    Example:
    ```
    # Please refer to entrypoints/api_server.py for
    # the complete example.

    # initialize the engine and the example input
    # note that engine_args here is AsyncEngineArgs instance
    engine = AsyncLLMEngine.from_engine_args(engine_args)
    example_input = {
        "input": "What is LLM?",
        "request_id": 0,
    }

    # start the generation
    results_generator = engine.encode(
    example_input["input"],
    PoolingParams(),
    example_input["request_id"])

    # get the results
    final_output = None
    async for request_output in results_generator:
        if await request.is_disconnected():
            # Abort the request if the client disconnects.
            await engine.abort(request_id)
            # Return or raise an error
            ...
        final_output = request_output

    # Process and return the final output
    ...
    ```
    """
    try:
        async for output in await self.add_request(
                request_id,
                prompt,
                pooling_params,
                lora_request=lora_request,
                trace_headers=trace_headers,
                priority=priority,
        ):
            yield LLMEngine.validate_output(output, PoolingRequestOutput)
    except asyncio.CancelledError:
        await self.abort(request_id)
        raise

engine_step `async` ¶

engine_step(virtual_engine: int) -> bool

Kick the engine to process the waiting requests.

Returns True if there are in-progress requests.

Source code in vllm/engine/async_llm_engine.py

async def engine_step(self, virtual_engine: int) -> bool:
    """Kick the engine to process the waiting requests.

    Returns True if there are in-progress requests."""

    new_requests, aborted_requests = (
        self._request_tracker.get_new_and_aborted_requests())

    for new_request in new_requests:
        # Add the request into the vLLM engine's waiting queue.
        try:
            await self.engine.add_request_async(**new_request)
        except ValueError as e:
            # TODO: use a vLLM specific error for failed validation
            self._request_tracker.process_exception(
                new_request["request_id"],
                e,
                verbose=self.log_requests,
            )

    if aborted_requests:
        await self._engine_abort(aborted_requests)

    request_outputs = await self.engine.step_async(virtual_engine)

    # Put the outputs into the corresponding streams.
    # If used as a callback, then already invoked inside
    # LLMEngine's _process_model_outputs
    if not self.use_process_request_outputs_callback:
        all_finished = self.process_request_outputs(request_outputs)
    else:
        # For callback case, we only need to detect when all
        # requests are finished
        all_finished = all(request_output.finished
                           for request_output in request_outputs)

    return not all_finished

from_engine_args `classmethod` ¶

from_engine_args(
    engine_args: AsyncEngineArgs,
    start_engine_loop: bool = True,
    usage_context: UsageContext = ENGINE_CONTEXT,
    stat_loggers: Optional[
        Dict[str, StatLoggerBase]
    ] = None,
) -> AsyncLLMEngine

Creates an async LLM engine from the engine arguments.

Source code in vllm/engine/async_llm_engine.py

@classmethod
def from_engine_args(
    cls,
    engine_args: AsyncEngineArgs,
    start_engine_loop: bool = True,
    usage_context: UsageContext = UsageContext.ENGINE_CONTEXT,
    stat_loggers: Optional[Dict[str, StatLoggerBase]] = None,
) -> "AsyncLLMEngine":
    """Creates an async LLM engine from the engine arguments."""

    vllm_config = engine_args.create_engine_config(usage_context)

    async_engine_cls = cls
    if envs.VLLM_USE_V1:
        from vllm.v1.engine.async_llm import AsyncLLM as V1AsyncLLMEngine
        async_engine_cls = V1AsyncLLMEngine

    return async_engine_cls.from_vllm_config(
        vllm_config=vllm_config,
        start_engine_loop=start_engine_loop,
        usage_context=usage_context,
        stat_loggers=stat_loggers,
        disable_log_stats=engine_args.disable_log_stats,
        disable_log_requests=engine_args.disable_log_requests,
    )

from_vllm_config `classmethod` ¶

from_vllm_config(
    vllm_config: VllmConfig,
    start_engine_loop: bool = True,
    usage_context: UsageContext = ENGINE_CONTEXT,
    stat_loggers: Optional[
        dict[str, StatLoggerBase]
    ] = None,
    disable_log_requests: bool = False,
    disable_log_stats: bool = False,
) -> AsyncLLMEngine

Create an AsyncLLMEngine from the EngineArgs.

Source code in vllm/engine/async_llm_engine.py

@classmethod
def from_vllm_config(
    cls,
    vllm_config: VllmConfig,
    start_engine_loop: bool = True,
    usage_context: UsageContext = UsageContext.ENGINE_CONTEXT,
    stat_loggers: Optional[dict[str, StatLoggerBase]] = None,
    disable_log_requests: bool = False,
    disable_log_stats: bool = False,
) -> "AsyncLLMEngine":
    """Create an AsyncLLMEngine from the EngineArgs."""

    return cls(
        vllm_config=vllm_config,
        executor_class=cls._get_executor_cls(vllm_config),
        start_engine_loop=start_engine_loop,
        log_requests=not disable_log_requests,
        log_stats=not disable_log_stats,
        usage_context=usage_context,
        stat_loggers=stat_loggers,
    )

generate `async` ¶

generate(
    prompt: PromptType,
    sampling_params: SamplingParams,
    request_id: str,
    lora_request: Optional[LoRARequest] = None,
    trace_headers: Optional[Mapping[str, str]] = None,
    prompt_adapter_request: Optional[
        PromptAdapterRequest
    ] = None,
    priority: int = 0,
    data_parallel_rank: Optional[int] = None,
) -> AsyncGenerator[RequestOutput, None]

Generate outputs for a request.

Generate outputs for a request. This method is a coroutine. It adds the request into the waiting queue of the LLMEngine and streams the outputs from the LLMEngine to the caller.

Parameters:

Name	Type	Description	Default
`prompt`	`PromptType`	The prompt to the LLM. See `PromptType` for more details about the format of each input.	required
`sampling_params`	`SamplingParams`	The sampling parameters of the request.	required
`request_id`	`str`	The unique id of the request.	required
`lora_request`	`Optional[LoRARequest]`	LoRA request to use for generation, if any.	`None`
`trace_headers`	`Optional[Mapping[str, str]]`	OpenTelemetry trace headers.	`None`
`prompt_adapter_request`	`Optional[PromptAdapterRequest]`	Prompt Adapter request to use for generation, if any.	`None`
`priority`	`int`	The priority of the request. Only applicable with priority scheduling.	`0`
`data_parallel_rank`	`Optional[int]`	The (global) data parallel rank that must handle this request. Only applicable if DP is enabled.	`None`

Yields: The output RequestOutput objects from the LLMEngine for the request.

Details

If the engine is not running, start the background loop, which iteratively invokes engine_step to process the waiting requests.
Add the request to the engine's RequestTracker. On the next background loop, this request will be sent to the underlying engine. Also, a corresponding AsyncStream will be created.
Wait for the request outputs from AsyncStream and yield them.

Example

Please refer to entrypoints/api_server.py for¶

the complete example.¶

initialize the engine and the example input¶

note that engine_args here is AsyncEngineArgs instance¶

engine = AsyncLLMEngine.from_engine_args(engine_args) example_input = { "prompt": "What is LLM?", "stream": False, # assume the non-streaming case "temperature": 0.0, "request_id": 0, }

start the generation¶

results_generator = engine.generate( example_input["prompt"], SamplingParams(temperature=example_input["temperature"]), example_input["request_id"])

get the results¶

final_output = None async for request_output in results_generator: if await request.is_disconnected(): # Abort the request if the client disconnects. await engine.abort(request_id) # Return or raise an error ... final_output = request_output

Process and return the final output¶

...

Source code in vllm/engine/async_llm_engine.py

async def generate(
    self,
    prompt: PromptType,
    sampling_params: SamplingParams,
    request_id: str,
    lora_request: Optional[LoRARequest] = None,
    trace_headers: Optional[Mapping[str, str]] = None,
    prompt_adapter_request: Optional[PromptAdapterRequest] = None,
    priority: int = 0,
    data_parallel_rank: Optional[int] = None,
) -> AsyncGenerator[RequestOutput, None]:
    """Generate outputs for a request.

    Generate outputs for a request. This method is a coroutine. It adds the
    request into the waiting queue of the LLMEngine and streams the outputs
    from the LLMEngine to the caller.

    Args:
        prompt: The prompt to the LLM. See
            [`PromptType`][vllm.inputs.PromptType] for more details about
            the format of each input.
        sampling_params: The sampling parameters of the request.
        request_id: The unique id of the request.
        lora_request: LoRA request to use for generation, if any.
        trace_headers: OpenTelemetry trace headers.
        prompt_adapter_request: Prompt Adapter request to use
                                        for generation, if any.
        priority: The priority of the request.
            Only applicable with priority scheduling.
        data_parallel_rank: The (global) data parallel rank that must
            handle this request. Only applicable if DP is enabled.
    Yields:
        The output `RequestOutput` objects from the LLMEngine
        for the request.

    Details:
        - If the engine is not running, start the background loop,
          which iteratively invokes
          [`engine_step`][vllm.engine.async_llm_engine.AsyncLLMEngine.engine_step]
          to process the waiting requests.
        - Add the request to the engine's `RequestTracker`.
          On the next background loop, this request will be sent to
          the underlying engine.
          Also, a corresponding `AsyncStream` will be created.
        - Wait for the request outputs from `AsyncStream` and yield them.

    Example:
        >>> # Please refer to entrypoints/api_server.py for
        >>> # the complete example.
        >>>
        >>> # initialize the engine and the example input
        >>> # note that engine_args here is AsyncEngineArgs instance
        >>> engine = AsyncLLMEngine.from_engine_args(engine_args)
        >>> example_input = {
        >>>     "prompt": "What is LLM?",
        >>>     "stream": False, # assume the non-streaming case
        >>>     "temperature": 0.0,
        >>>     "request_id": 0,
        >>> }
        >>>
        >>> # start the generation
        >>> results_generator = engine.generate(
        >>>    example_input["prompt"],
        >>>    SamplingParams(temperature=example_input["temperature"]),
        >>>    example_input["request_id"])
        >>>
        >>> # get the results
        >>> final_output = None
        >>> async for request_output in results_generator:
        >>>     if await request.is_disconnected():
        >>>         # Abort the request if the client disconnects.
        >>>         await engine.abort(request_id)
        >>>         # Return or raise an error
        >>>         ...
        >>>     final_output = request_output
        >>>
        >>> # Process and return the final output
        >>> ...
    """
    try:
        async for output in await self.add_request(
                request_id,
                prompt,
                sampling_params,
                lora_request=lora_request,
                trace_headers=trace_headers,
                prompt_adapter_request=prompt_adapter_request,
                priority=priority,
                data_parallel_rank=data_parallel_rank,
        ):
            yield LLMEngine.validate_output(output, RequestOutput)
    except asyncio.CancelledError:
        await self.abort(request_id)
        raise

get_decoding_config `async` ¶

get_decoding_config() -> DecodingConfig

Get the decoding configuration of the vLLM engine.

Source code in vllm/engine/async_llm_engine.py

async def get_decoding_config(self) -> DecodingConfig:
    """Get the decoding configuration of the vLLM engine."""
    return self.engine.get_decoding_config()

get_input_preprocessor `async` ¶

get_input_preprocessor() -> InputPreprocessor

Source code in vllm/engine/async_llm_engine.py

async def get_input_preprocessor(self) -> InputPreprocessor:
    return self.engine.input_preprocessor

get_lora_config `async` ¶

get_lora_config() -> LoRAConfig

Get the lora configuration of the vLLM engine.

Source code in vllm/engine/async_llm_engine.py

async def get_lora_config(self) -> LoRAConfig:
    """Get the lora configuration of the vLLM engine."""
    return self.engine.get_lora_config()

get_model_config `async` ¶

get_model_config() -> ModelConfig

Get the model configuration of the vLLM engine.

Source code in vllm/engine/async_llm_engine.py

async def get_model_config(self) -> ModelConfig:
    """Get the model configuration of the vLLM engine."""
    return self.engine.get_model_config()

get_parallel_config `async` ¶

get_parallel_config() -> ParallelConfig

Get the parallel configuration of the vLLM engine.

Source code in vllm/engine/async_llm_engine.py

async def get_parallel_config(self) -> ParallelConfig:
    """Get the parallel configuration of the vLLM engine."""
    return self.engine.get_parallel_config()

get_scheduler_config `async` ¶

get_scheduler_config() -> SchedulerConfig

Get the scheduling configuration of the vLLM engine.

Source code in vllm/engine/async_llm_engine.py

async def get_scheduler_config(self) -> SchedulerConfig:
    """Get the scheduling configuration of the vLLM engine."""
    return self.engine.get_scheduler_config()

get_tokenizer `async` ¶

get_tokenizer(
    lora_request: Optional[LoRARequest] = None,
) -> AnyTokenizer

Source code in vllm/engine/async_llm_engine.py

async def get_tokenizer(
    self,
    lora_request: Optional[LoRARequest] = None,
) -> AnyTokenizer:
    return await self.engine.get_tokenizer_async(lora_request)

get_vllm_config `async` ¶

get_vllm_config() -> VllmConfig

Get the vllm configuration of the vLLM engine.

Source code in vllm/engine/async_llm_engine.py

async def get_vllm_config(self) -> VllmConfig:
    """Get the vllm configuration of the vLLM engine."""
    return self.engine.get_vllm_config()

is_sleeping `async` ¶

is_sleeping() -> bool

Source code in vllm/engine/async_llm_engine.py

async def is_sleeping(self) -> bool:
    return self.engine.is_sleeping()

is_tracing_enabled `async` ¶

is_tracing_enabled() -> bool

Source code in vllm/engine/async_llm_engine.py

async def is_tracing_enabled(self) -> bool:
    return self.engine.is_tracing_enabled()

process_request_outputs ¶

process_request_outputs(request_outputs) -> bool

Source code in vllm/engine/async_llm_engine.py

def process_request_outputs(self, request_outputs) -> bool:
    # Put the outputs into the corresponding streams.
    all_finished = True
    for request_output in request_outputs:
        self._request_tracker.process_request_output(
            request_output, verbose=self.log_requests)
        all_finished = all_finished and request_output.finished

    return all_finished

remove_logger ¶

remove_logger(logger_name: str) -> None

Source code in vllm/engine/async_llm_engine.py

def remove_logger(self, logger_name: str) -> None:
    self.engine.remove_logger(logger_name=logger_name)

reset_mm_cache `async` ¶

reset_mm_cache() -> None

Source code in vllm/engine/async_llm_engine.py

async def reset_mm_cache(self) -> None:
    self.engine.reset_mm_cache()

reset_prefix_cache `async` ¶

reset_prefix_cache(device: Optional[Device] = None) -> None

Source code in vllm/engine/async_llm_engine.py

async def reset_prefix_cache(self,
                             device: Optional[Device] = None) -> None:
    self.engine.reset_prefix_cache(device)

run_engine_loop `async` `staticmethod` ¶

run_engine_loop(engine_ref: ReferenceType)

We use a weakref to the engine so that the running loop doesn't prevent the engine being garbage collected.

Source code in vllm/engine/async_llm_engine.py

@staticmethod
async def run_engine_loop(engine_ref: ReferenceType):
    """We use a weakref to the engine so that the running loop
    doesn't prevent the engine being garbage collected."""
    engine: Optional[AsyncLLMEngine] = engine_ref()
    if not engine:
        return

    pipeline_parallel_size = \
            engine.engine.parallel_config.pipeline_parallel_size
    has_requests_in_progress = [False] * pipeline_parallel_size
    while True:
        if not any(has_requests_in_progress):
            logger.debug("Waiting for new requests...")
            # Stop the execute model loop in parallel workers until there
            # are more requests to process. This avoids waiting
            # indefinitely in torch.distributed ops which may otherwise
            # timeout, and unblocks the RPC thread in the workers so that
            # they can process any other queued control plane messages,
            # such as add/remove lora adapters.
            await engine.engine.stop_remote_worker_execution_loop_async()
            request_tracker = engine._request_tracker
            # Allow engine to be garbage collected while
            # waiting for new requests
            del engine
            await asyncio.sleep(0)
            if engine_ref() is None:
                return
            await request_tracker.wait_for_new_requests()
            engine = engine_ref()
            if not engine:
                return
            logger.debug("Got new requests!")
            requests_in_progress = [
                asyncio.create_task(engine.engine_step(ve))
                for ve in range(pipeline_parallel_size)
            ]
            has_requests_in_progress = [True] * pipeline_parallel_size

        # Abort if iteration takes too long due to unrecoverable errors
        # (eg. NCCL timeouts).
        try:
            async with asyncio_timeout(ENGINE_ITERATION_TIMEOUT_S):
                done, _ = await asyncio.wait(
                    requests_in_progress,
                    return_when=asyncio.FIRST_COMPLETED)
                for _ in range(pipeline_parallel_size):
                    await asyncio.sleep(0)
            for task in done:
                result = task.result()
                virtual_engine = requests_in_progress.index(task)
                has_unfinished_requests = (
                    engine.engine.
                    has_unfinished_requests_for_virtual_engine(
                        virtual_engine))
                if result or has_unfinished_requests:
                    requests_in_progress[virtual_engine] = (
                        asyncio.create_task(
                            engine.engine_step(virtual_engine)))
                    has_requests_in_progress[virtual_engine] = True
                else:
                    has_requests_in_progress[virtual_engine] = False
        except asyncio.TimeoutError as exc:
            logger.error(
                "Engine iteration timed out. This should never happen!")
            engine.set_errored(exc)
            raise
        await asyncio.sleep(0)

set_errored ¶

set_errored(exc: Exception) -> None

Source code in vllm/engine/async_llm_engine.py

def set_errored(self, exc: Exception) -> None:
    self._errored_with = exc

shutdown_background_loop ¶

shutdown_background_loop() -> None

Shut down the background loop.

This method needs to be called during cleanup to remove references to self and properly GC the resources held by the async LLM engine (e.g., the executors as well as their resources).

Source code in vllm/engine/async_llm_engine.py

def shutdown_background_loop(self) -> None:
    """
    Shut down the background loop.

    This method needs to be called during cleanup to remove
    references to `self` and properly GC the resources held
    by the async LLM engine (e.g., the executors as well as
    their resources).
    """
    if self._background_loop_unshielded is not None:
        self._background_loop_unshielded.cancel()
        self._background_loop_unshielded = None
    self.background_loop = None

sleep `async` ¶

sleep(level: int = 1) -> None

Source code in vllm/engine/async_llm_engine.py

async def sleep(self, level: int = 1) -> None:
    self.engine.sleep(level)

start_background_loop ¶

start_background_loop() -> None

Start the background loop.

Source code in vllm/engine/async_llm_engine.py

def start_background_loop(self) -> None:
    """Start the background loop."""
    if self.errored:
        raise AsyncEngineDeadError(
            "Background loop has errored already.") from self._errored_with
    if self.is_running:
        raise RuntimeError("Background loop is already running.")
    # Initialize the RequestTracker here so it uses the right event loop.
    self._request_tracker = RequestTracker()

    self._background_loop_unshielded = asyncio.get_event_loop(
    ).create_task(self.run_engine_loop(weakref.ref(self)))
    self._background_loop_unshielded.add_done_callback(
        partial(_log_task_completion, error_callback=self._error_callback))
    self.background_loop = asyncio.shield(self._background_loop_unshielded)

start_profile `async` ¶

start_profile() -> None

Source code in vllm/engine/async_llm_engine.py

async def start_profile(self) -> None:
    self.engine.start_profile()

stop_profile `async` ¶

stop_profile() -> None

Source code in vllm/engine/async_llm_engine.py

async def stop_profile(self) -> None:
    self.engine.stop_profile()

wake_up `async` ¶

wake_up(tags: Optional[list[str]] = None) -> None

Source code in vllm/engine/async_llm_engine.py

async def wake_up(self, tags: Optional[list[str]] = None) -> None:
    self.engine.wake_up(tags)

AsyncStream ¶

A stream of RequestOutputs or PoolingRequestOutputs for a request that can be iterated over asynchronously via an async generator.

Source code in vllm/engine/async_llm_engine.py

class AsyncStream:
    """A stream of RequestOutputs or PoolingRequestOutputs for a request
    that can be iterated over asynchronously via an async generator."""

    def __init__(self, request_id: str, cancel: Callable[[str], None]) -> None:
        self.request_id = request_id
        self._cancel = cancel
        self._queue: asyncio.Queue = asyncio.Queue()
        self._finished = False

    def put(self, item: Union[RequestOutput, PoolingRequestOutput,
                              Exception]) -> None:
        if not self._finished:
            self._queue.put_nowait(item)

    def finish(
        self,
        exception: Optional[Union[BaseException, Type[BaseException]]] = None,
    ) -> None:
        if not self._finished:
            self._finished = True
            self._queue.put_nowait(
                exception if self._is_raisable(exception) else STOP_ITERATION)

    @property
    def finished(self) -> bool:
        return self._finished

    async def generator(
        self
    ) -> AsyncGenerator[Union[RequestOutput, PoolingRequestOutput], None]:
        try:
            while True:
                result = await self._queue.get()
                if self._is_raisable(result):
                    if result == STOP_ITERATION:
                        return
                    raise result
                yield result
        except GeneratorExit:
            self._cancel(self.request_id)
            raise asyncio.CancelledError from None

    @staticmethod
    def _is_raisable(value: Any):
        return isinstance(value, BaseException) or \
                (isinstance(value, type) and \
                 issubclass(value, BaseException))

_cancel `instance-attribute` ¶

_cancel = cancel

_finished `instance-attribute` ¶

_finished = False

_queue `instance-attribute` ¶

_queue: Queue = Queue()

finished `property` ¶

finished: bool

request_id `instance-attribute` ¶

request_id = request_id

init ¶

__init__(
    request_id: str, cancel: Callable[[str], None]
) -> None

Source code in vllm/engine/async_llm_engine.py

def __init__(self, request_id: str, cancel: Callable[[str], None]) -> None:
    self.request_id = request_id
    self._cancel = cancel
    self._queue: asyncio.Queue = asyncio.Queue()
    self._finished = False

_is_raisable `staticmethod` ¶

_is_raisable(value: Any)

Source code in vllm/engine/async_llm_engine.py

@staticmethod
def _is_raisable(value: Any):
    return isinstance(value, BaseException) or \
            (isinstance(value, type) and \
             issubclass(value, BaseException))

finish ¶

finish(
    exception: Optional[
        Union[BaseException, Type[BaseException]]
    ] = None,
) -> None

Source code in vllm/engine/async_llm_engine.py

def finish(
    self,
    exception: Optional[Union[BaseException, Type[BaseException]]] = None,
) -> None:
    if not self._finished:
        self._finished = True
        self._queue.put_nowait(
            exception if self._is_raisable(exception) else STOP_ITERATION)

generator `async` ¶

generator() -> AsyncGenerator[
    Union[RequestOutput, PoolingRequestOutput], None
]

Source code in vllm/engine/async_llm_engine.py

async def generator(
    self
) -> AsyncGenerator[Union[RequestOutput, PoolingRequestOutput], None]:
    try:
        while True:
            result = await self._queue.get()
            if self._is_raisable(result):
                if result == STOP_ITERATION:
                    return
                raise result
            yield result
    except GeneratorExit:
        self._cancel(self.request_id)
        raise asyncio.CancelledError from None

put ¶

put(
    item: Union[
        RequestOutput, PoolingRequestOutput, Exception
    ],
) -> None

Source code in vllm/engine/async_llm_engine.py

def put(self, item: Union[RequestOutput, PoolingRequestOutput,
                          Exception]) -> None:
    if not self._finished:
        self._queue.put_nowait(item)

RequestTracker ¶

Synchronous abstraction for tracking requests.

Source code in vllm/engine/async_llm_engine.py

class RequestTracker:
    """Synchronous abstraction for tracking requests."""

    def __init__(self) -> None:
        self._request_streams: Dict[str, AsyncStream] = {}
        self._aborted_requests: asyncio.Queue[str] = asyncio.Queue()
        self._new_requests: asyncio.Queue[Tuple[AsyncStream,
                                                dict]] = asyncio.Queue()
        self.new_requests_event = asyncio.Event()

    def __contains__(self, item):
        return item in self._request_streams

    def __len__(self) -> int:
        return len(self._request_streams)

    def propagate_exception(self,
                            exc: Exception,
                            request_id: Optional[str] = None) -> None:
        """Propagate an exception to request streams
        (all if request_id is None)."""
        if request_id is not None:
            self.abort_request(request_id, exception=exc)
        else:
            # NB: tuple() used here because self.abort_request pops the stream
            # out of self._request_streams, so we can't iterate on it directly
            for rid in tuple(self._request_streams.keys()):
                self.abort_request(rid, exception=exc)

    def process_request_output(self,
                               request_output: Union[RequestOutput,
                                                     PoolingRequestOutput],
                               *,
                               verbose: bool = False) -> None:
        """Process a request output from the engine."""
        request_id = request_output.request_id
        finished = request_output.finished

        if finished:
            stream = self._request_streams.pop(request_id, None)
        else:
            stream = self._request_streams.get(request_id)
        # Guard against a KeyError which can occur if the request was aborted
        # while the output was generated
        if stream is not None:
            stream.put(request_output)
            if finished:
                stream.finish()

        if verbose and finished:
            logger.info("Finished request %s.", request_id)

    def process_exception(self,
                          request_id: str,
                          exception: BaseException,
                          *,
                          verbose: bool = False) -> None:
        """Propagate an exception from the engine."""
        if verbose:
            logger.info("Finished request %s.", request_id)
        self.abort_request(request_id, exception=exception)

    def add_request(self,
                    request_id: str,
                    *,
                    verbose: bool = False,
                    **engine_add_request_kwargs) -> AsyncStream:
        """Add a request to be sent to the engine on the next background
        loop iteration."""
        if request_id in self._request_streams:
            raise KeyError(f"Request {request_id} already exists.")

        abort_request = partial(self.abort_request, verbose=verbose)
        stream = AsyncStream(request_id, abort_request)
        self._new_requests.put_nowait((stream, {
            "request_id": request_id,
            **engine_add_request_kwargs
        }))

        self.new_requests_event.set()

        if verbose:
            logger.info("Added request %s.", request_id)

        return stream

    def abort_request(self,
                      request_id: str,
                      *,
                      exception: Optional[Union[BaseException,
                                                Type[BaseException]]] = None,
                      verbose: bool = False) -> None:
        """Abort a request during next background loop iteration."""
        if verbose:
            logger.info("Aborted request %s.", request_id)

        self._aborted_requests.put_nowait(request_id)

        stream = self._request_streams.pop(request_id, None)
        if stream is not None:
            stream.finish(exception=exception)

    def get_new_and_aborted_requests(self) -> Tuple[List[Dict], Set[str]]:
        """Get the new requests and finished requests to be
        sent to the engine."""
        new_requests: List[Dict] = []
        finished_requests: Set[str] = set()

        while not self._aborted_requests.empty():
            request_id = self._aborted_requests.get_nowait()
            finished_requests.add(request_id)

        while not self._new_requests.empty():
            stream, new_request = self._new_requests.get_nowait()
            request_id = stream.request_id
            if request_id in finished_requests:
                # The request has already been aborted.
                stream.finish(asyncio.CancelledError)
                finished_requests.discard(request_id)
            else:
                self._request_streams[request_id] = stream
                new_requests.append(new_request)

        return new_requests, finished_requests

    async def wait_for_new_requests(self):
        if not self.has_new_requests():
            await self.new_requests_event.wait()
        self.new_requests_event.clear()

    def has_new_requests(self):
        return not self._new_requests.empty()

_aborted_requests `instance-attribute` ¶

_aborted_requests: Queue[str] = Queue()

_new_requests `instance-attribute` ¶

_new_requests: Queue[Tuple[AsyncStream, dict]] = Queue()

_request_streams `instance-attribute` ¶

_request_streams: Dict[str, AsyncStream] = {}

new_requests_event `instance-attribute` ¶

new_requests_event = Event()

contains ¶

__contains__(item)

Source code in vllm/engine/async_llm_engine.py

def __contains__(self, item):
    return item in self._request_streams

init ¶

__init__() -> None

Source code in vllm/engine/async_llm_engine.py

def __init__(self) -> None:
    self._request_streams: Dict[str, AsyncStream] = {}
    self._aborted_requests: asyncio.Queue[str] = asyncio.Queue()
    self._new_requests: asyncio.Queue[Tuple[AsyncStream,
                                            dict]] = asyncio.Queue()
    self.new_requests_event = asyncio.Event()

len ¶

__len__() -> int

Source code in vllm/engine/async_llm_engine.py

def __len__(self) -> int:
    return len(self._request_streams)

abort_request ¶

abort_request(
    request_id: str,
    *,
    exception: Optional[
        Union[BaseException, Type[BaseException]]
    ] = None,
    verbose: bool = False,
) -> None

Abort a request during next background loop iteration.

Source code in vllm/engine/async_llm_engine.py

def abort_request(self,
                  request_id: str,
                  *,
                  exception: Optional[Union[BaseException,
                                            Type[BaseException]]] = None,
                  verbose: bool = False) -> None:
    """Abort a request during next background loop iteration."""
    if verbose:
        logger.info("Aborted request %s.", request_id)

    self._aborted_requests.put_nowait(request_id)

    stream = self._request_streams.pop(request_id, None)
    if stream is not None:
        stream.finish(exception=exception)

add_request ¶

add_request(
    request_id: str,
    *,
    verbose: bool = False,
    **engine_add_request_kwargs,
) -> AsyncStream

Add a request to be sent to the engine on the next background loop iteration.

Source code in vllm/engine/async_llm_engine.py

def add_request(self,
                request_id: str,
                *,
                verbose: bool = False,
                **engine_add_request_kwargs) -> AsyncStream:
    """Add a request to be sent to the engine on the next background
    loop iteration."""
    if request_id in self._request_streams:
        raise KeyError(f"Request {request_id} already exists.")

    abort_request = partial(self.abort_request, verbose=verbose)
    stream = AsyncStream(request_id, abort_request)
    self._new_requests.put_nowait((stream, {
        "request_id": request_id,
        **engine_add_request_kwargs
    }))

    self.new_requests_event.set()

    if verbose:
        logger.info("Added request %s.", request_id)

    return stream

get_new_and_aborted_requests ¶

get_new_and_aborted_requests() -> Tuple[
    List[Dict], Set[str]
]

Get the new requests and finished requests to be sent to the engine.

Source code in vllm/engine/async_llm_engine.py

def get_new_and_aborted_requests(self) -> Tuple[List[Dict], Set[str]]:
    """Get the new requests and finished requests to be
    sent to the engine."""
    new_requests: List[Dict] = []
    finished_requests: Set[str] = set()

    while not self._aborted_requests.empty():
        request_id = self._aborted_requests.get_nowait()
        finished_requests.add(request_id)

    while not self._new_requests.empty():
        stream, new_request = self._new_requests.get_nowait()
        request_id = stream.request_id
        if request_id in finished_requests:
            # The request has already been aborted.
            stream.finish(asyncio.CancelledError)
            finished_requests.discard(request_id)
        else:
            self._request_streams[request_id] = stream
            new_requests.append(new_request)

    return new_requests, finished_requests

has_new_requests ¶

has_new_requests()

Source code in vllm/engine/async_llm_engine.py

def has_new_requests(self):
    return not self._new_requests.empty()

process_exception ¶

process_exception(
    request_id: str,
    exception: BaseException,
    *,
    verbose: bool = False,
) -> None

Propagate an exception from the engine.

Source code in vllm/engine/async_llm_engine.py

def process_exception(self,
                      request_id: str,
                      exception: BaseException,
                      *,
                      verbose: bool = False) -> None:
    """Propagate an exception from the engine."""
    if verbose:
        logger.info("Finished request %s.", request_id)
    self.abort_request(request_id, exception=exception)

process_request_output ¶

process_request_output(
    request_output: Union[
        RequestOutput, PoolingRequestOutput
    ],
    *,
    verbose: bool = False,
) -> None

Process a request output from the engine.

Source code in vllm/engine/async_llm_engine.py

def process_request_output(self,
                           request_output: Union[RequestOutput,
                                                 PoolingRequestOutput],
                           *,
                           verbose: bool = False) -> None:
    """Process a request output from the engine."""
    request_id = request_output.request_id
    finished = request_output.finished

    if finished:
        stream = self._request_streams.pop(request_id, None)
    else:
        stream = self._request_streams.get(request_id)
    # Guard against a KeyError which can occur if the request was aborted
    # while the output was generated
    if stream is not None:
        stream.put(request_output)
        if finished:
            stream.finish()

    if verbose and finished:
        logger.info("Finished request %s.", request_id)

propagate_exception ¶

propagate_exception(
    exc: Exception, request_id: Optional[str] = None
) -> None

Propagate an exception to request streams (all if request_id is None).

Source code in vllm/engine/async_llm_engine.py

def propagate_exception(self,
                        exc: Exception,
                        request_id: Optional[str] = None) -> None:
    """Propagate an exception to request streams
    (all if request_id is None)."""
    if request_id is not None:
        self.abort_request(request_id, exception=exc)
    else:
        # NB: tuple() used here because self.abort_request pops the stream
        # out of self._request_streams, so we can't iterate on it directly
        for rid in tuple(self._request_streams.keys()):
            self.abort_request(rid, exception=exc)

wait_for_new_requests `async` ¶

wait_for_new_requests()

Source code in vllm/engine/async_llm_engine.py

async def wait_for_new_requests(self):
    if not self.has_new_requests():
        await self.new_requests_event.wait()
    self.new_requests_event.clear()

_AsyncLLMEngine ¶

Bases: LLMEngine

Extension of LLMEngine to add async methods.

Source code in vllm/engine/async_llm_engine.py

class _AsyncLLMEngine(LLMEngine):
    """Extension of LLMEngine to add async methods."""

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)

    async def step_async(
        self, virtual_engine: int
    ) -> List[Union[RequestOutput, PoolingRequestOutput]]:
        """Performs one decoding iteration and returns newly generated results.
        The workers are ran asynchronously if possible.

        This function performs one decoding iteration of the engine. It first
        schedules the sequences to be executed in the next iteration and the
        token blocks to be swapped in/out/copy. Then, it executes the model
        and updates the scheduler with the model outputs. Finally, it decodes
        the sequences and returns the newly generated results.
        """
        # these are cached outputs from previous iterations. None if on first
        # iteration
        cached_outputs = self.cached_scheduler_outputs[virtual_engine]
        seq_group_metadata_list = cached_outputs.seq_group_metadata_list
        scheduler_outputs = cached_outputs.scheduler_outputs
        allow_async_output_proc = cached_outputs.allow_async_output_proc

        ctx = self.scheduler_contexts[virtual_engine]

        # Clear outputs for each new scheduler iteration
        ctx.request_outputs.clear()

        # skip the scheduler if there are any remaining steps in the seq groups.
        # This ensures that the scheduler is only called again when the current
        # batch has completed.
        if not self._has_remaining_steps(seq_group_metadata_list):

            # Schedule iteration
            (seq_group_metadata_list, scheduler_outputs,
             allow_async_output_proc
             ) = self.scheduler[virtual_engine].schedule()

            ctx.seq_group_metadata_list = seq_group_metadata_list
            ctx.scheduler_outputs = scheduler_outputs

            if not scheduler_outputs.is_empty():
                # this will cause mamba_cache/minimax_cache failed
                # to release finished_requests_ids of the last steps
                finished_requests_ids = self.scheduler[
                    virtual_engine].get_and_reset_finished_requests_ids()

            # Maybe switch from async mode to sync mode
            if not allow_async_output_proc and len(ctx.output_queue) > 0:
                self._process_model_outputs(ctx=ctx)

            if (self.scheduler_config.is_multi_step
                    and scheduler_outputs.num_lookahead_slots > 0):
                # cache the scheduler outputs for the next iteration if we have
                # lookahead slots
                self._cache_scheduler_outputs_for_multi_step(
                    virtual_engine, seq_group_metadata_list, scheduler_outputs,
                    allow_async_output_proc)
        else:
            finished_requests_ids = list()

        assert seq_group_metadata_list is not None
        assert scheduler_outputs is not None

        if not scheduler_outputs.is_empty():

            # Check if we have a cached last_output from the previous iteration.
            # For supporting PP this is probably the best way to pass the
            # sampled_token_ids, as a separate broadcast over all the PP stages
            # will cause one virtual engine's microbatch to block the pipeline.
            last_sampled_token_ids = \
                self._get_last_sampled_token_ids(virtual_engine)

            execute_model_req = ExecuteModelRequest(
                seq_group_metadata_list=seq_group_metadata_list,
                blocks_to_swap_in=scheduler_outputs.blocks_to_swap_in,
                blocks_to_swap_out=scheduler_outputs.blocks_to_swap_out,
                blocks_to_copy=scheduler_outputs.blocks_to_copy,
                virtual_engine=virtual_engine,
                num_lookahead_slots=scheduler_outputs.num_lookahead_slots,
                running_queue_size=scheduler_outputs.running_queue_size,
                finished_requests_ids=finished_requests_ids,
                # We use ExecuteModelRequest to pass the last sampled_token_ids
                # to each of the non-last PP stages for in-place prepare_input.
                last_sampled_token_ids=last_sampled_token_ids)

            if allow_async_output_proc:
                execute_model_req.async_callback = self.async_callbacks[
                    virtual_engine]

            # Execute the model.
            outputs = await self.model_executor.execute_model_async(
                execute_model_req)

            # we need to do this here so that last step's sampled_token_ids can
            # be passed to the next iteration for PP.
            if self.scheduler_config.is_multi_step:
                self._update_cached_scheduler_output(virtual_engine, outputs)
        else:
            if len(ctx.output_queue) > 0:
                self._process_model_outputs(ctx=ctx)
            outputs = []

        # Finish the current step for all the sequence groups.
        if self.scheduler_config.is_multi_step:
            for seq_group in seq_group_metadata_list:
                seq_group.finish_step()

        if not self._has_remaining_steps(seq_group_metadata_list):
            # Clear the cache if we have finished all the steps
            if self.scheduler_config.is_multi_step:
                self.cached_scheduler_outputs[
                    virtual_engine] = SchedulerOutputState()

            # is_first_step_output is True only when the num_steps of all
            # the sequences are 1. When the num_steps > 1,
            # multi_step_model_runner does the first-step output append.
            is_first_step_output: bool = False if not seq_group_metadata_list \
                else seq_group_metadata_list[0].state.num_steps == 1

            ctx.append_output(outputs=outputs,
                              seq_group_metadata_list=seq_group_metadata_list,
                              scheduler_outputs=scheduler_outputs,
                              is_async=allow_async_output_proc,
                              is_last_step=True,
                              is_first_step_output=is_first_step_output)

            if outputs and allow_async_output_proc:
                assert len(
                    outputs
                ) == 1, "Async postprocessor expects only a single output set"
                self._advance_to_next_step(
                    outputs[0], seq_group_metadata_list,
                    scheduler_outputs.scheduled_seq_groups)

            if not allow_async_output_proc:
                self._process_model_outputs(ctx=ctx)

                # Log stats.
                self.do_log_stats(scheduler_outputs, outputs)

                # Tracing
                self.do_tracing(scheduler_outputs)

        else:
            # Multi-step case
            return ctx.request_outputs

        if not self.has_unfinished_requests():
            # Drain async postprocessor (if exists)
            if len(ctx.output_queue) > 0:
                self._process_model_outputs(ctx=ctx)
            assert len(ctx.output_queue) == 0

        return ctx.request_outputs

    async def stop_remote_worker_execution_loop_async(self) -> None:
        """Stop the remote worker execution loop."""
        await self.model_executor.stop_remote_worker_execution_loop_async()

    async def get_tokenizer_async(self,
                                  lora_request: Optional[LoRARequest] = None
                                  ) -> AnyTokenizer:
        return await (
            self.get_tokenizer_group().get_lora_tokenizer_async(lora_request))

    async def add_request_async(
        self,
        request_id: str,
        prompt: PromptType,
        params: Union[SamplingParams, PoolingParams],
        arrival_time: Optional[float] = None,
        lora_request: Optional[LoRARequest] = None,
        trace_headers: Optional[Mapping[str, str]] = None,
        prompt_adapter_request: Optional[PromptAdapterRequest] = None,
        priority: int = 0,
        data_parallel_rank: Optional[int] = None,
    ) -> None:
        """
        Async version of
        [`add_request`][vllm.engine.llm_engine.LLMEngine.add_request].
        """
        if lora_request is not None and not self.lora_config:
            raise ValueError(f"Got lora_request {lora_request} but LoRA is "
                             "not enabled!")
        if priority != 0 and not self.scheduler_config.policy == "priority":
            raise ValueError(f"Got priority {priority} but "
                             "Priority scheduling is not enabled.")
        if arrival_time is None:
            arrival_time = time.time()

        if data_parallel_rank is not None:
            raise ValueError("Targeting data_parallel_rank only supported "
                             "in v1 client.")

        if (isinstance(prompt, dict)
                and prompt.get("prompt_embeds", None) is not None
                and not prompt.get("prompt_token_ids", None)):
            # We use the -2 dimension (instead of 0) in case a batched input
            # of batch size 1 is passed in.
            prompt["prompt_token_ids"] = [0
                                          ] * prompt["prompt_embeds"].shape[-2]

        processed_inputs = await self.input_preprocessor.preprocess_async(
            prompt,
            lora_request=lora_request,
            prompt_adapter_request=prompt_adapter_request,
        )

        if isinstance(params, SamplingParams) and \
            params.guided_decoding is not None:
            # Guided decoding has an async implementation for building logits
            # processors in a separate threadpool.
            # We want to invoke that here instead of using the blocking
            # implementation in the LLMEngine
            params = await build_guided_decoding_logits_processor_async(
                sampling_params=params,
                tokenizer=await self.get_tokenizer_async(lora_request),
                default_guided_backend=self.decoding_config.backend,
                reasoning_backend=self.decoding_config.reasoning_backend,
                model_config=self.model_config)

        self._add_processed_request(
            request_id=request_id,
            processed_inputs=processed_inputs,
            params=params,
            arrival_time=arrival_time,
            lora_request=lora_request,
            prompt_adapter_request=prompt_adapter_request,
            trace_headers=trace_headers,
            priority=priority,
        )

    async def check_health_async(self) -> None:
        self.model_executor.check_health()

    async def collective_rpc_async(self,
                                   method: str,
                                   timeout: Optional[float] = None,
                                   args: tuple = (),
                                   kwargs: Optional[dict] = None):
        raise NotImplementedError

init ¶

__init__(*args, **kwargs)

Source code in vllm/engine/async_llm_engine.py

def __init__(self, *args, **kwargs):
    super().__init__(*args, **kwargs)

add_request_async `async` ¶

add_request_async(
    request_id: str,
    prompt: PromptType,
    params: Union[SamplingParams, PoolingParams],
    arrival_time: Optional[float] = None,
    lora_request: Optional[LoRARequest] = None,
    trace_headers: Optional[Mapping[str, str]] = None,
    prompt_adapter_request: Optional[
        PromptAdapterRequest
    ] = None,
    priority: int = 0,
    data_parallel_rank: Optional[int] = None,
) -> None

Async version of add_request.

Source code in vllm/engine/async_llm_engine.py

async def add_request_async(
    self,
    request_id: str,
    prompt: PromptType,
    params: Union[SamplingParams, PoolingParams],
    arrival_time: Optional[float] = None,
    lora_request: Optional[LoRARequest] = None,
    trace_headers: Optional[Mapping[str, str]] = None,
    prompt_adapter_request: Optional[PromptAdapterRequest] = None,
    priority: int = 0,
    data_parallel_rank: Optional[int] = None,
) -> None:
    """
    Async version of
    [`add_request`][vllm.engine.llm_engine.LLMEngine.add_request].
    """
    if lora_request is not None and not self.lora_config:
        raise ValueError(f"Got lora_request {lora_request} but LoRA is "
                         "not enabled!")
    if priority != 0 and not self.scheduler_config.policy == "priority":
        raise ValueError(f"Got priority {priority} but "
                         "Priority scheduling is not enabled.")
    if arrival_time is None:
        arrival_time = time.time()

    if data_parallel_rank is not None:
        raise ValueError("Targeting data_parallel_rank only supported "
                         "in v1 client.")

    if (isinstance(prompt, dict)
            and prompt.get("prompt_embeds", None) is not None
            and not prompt.get("prompt_token_ids", None)):
        # We use the -2 dimension (instead of 0) in case a batched input
        # of batch size 1 is passed in.
        prompt["prompt_token_ids"] = [0
                                      ] * prompt["prompt_embeds"].shape[-2]

    processed_inputs = await self.input_preprocessor.preprocess_async(
        prompt,
        lora_request=lora_request,
        prompt_adapter_request=prompt_adapter_request,
    )

    if isinstance(params, SamplingParams) and \
        params.guided_decoding is not None:
        # Guided decoding has an async implementation for building logits
        # processors in a separate threadpool.
        # We want to invoke that here instead of using the blocking
        # implementation in the LLMEngine
        params = await build_guided_decoding_logits_processor_async(
            sampling_params=params,
            tokenizer=await self.get_tokenizer_async(lora_request),
            default_guided_backend=self.decoding_config.backend,
            reasoning_backend=self.decoding_config.reasoning_backend,
            model_config=self.model_config)

    self._add_processed_request(
        request_id=request_id,
        processed_inputs=processed_inputs,
        params=params,
        arrival_time=arrival_time,
        lora_request=lora_request,
        prompt_adapter_request=prompt_adapter_request,
        trace_headers=trace_headers,
        priority=priority,
    )

check_health_async `async` ¶

check_health_async() -> None

Source code in vllm/engine/async_llm_engine.py

async def check_health_async(self) -> None:
    self.model_executor.check_health()

collective_rpc_async `async` ¶

collective_rpc_async(
    method: str,
    timeout: Optional[float] = None,
    args: tuple = (),
    kwargs: Optional[dict] = None,
)

Source code in vllm/engine/async_llm_engine.py

async def collective_rpc_async(self,
                               method: str,
                               timeout: Optional[float] = None,
                               args: tuple = (),
                               kwargs: Optional[dict] = None):
    raise NotImplementedError

get_tokenizer_async `async` ¶

get_tokenizer_async(
    lora_request: Optional[LoRARequest] = None,
) -> AnyTokenizer

Source code in vllm/engine/async_llm_engine.py

async def get_tokenizer_async(self,
                              lora_request: Optional[LoRARequest] = None
                              ) -> AnyTokenizer:
    return await (
        self.get_tokenizer_group().get_lora_tokenizer_async(lora_request))

step_async `async` ¶

step_async(
    virtual_engine: int,
) -> List[Union[RequestOutput, PoolingRequestOutput]]

Performs one decoding iteration and returns newly generated results. The workers are ran asynchronously if possible.

This function performs one decoding iteration of the engine. It first schedules the sequences to be executed in the next iteration and the token blocks to be swapped in/out/copy. Then, it executes the model and updates the scheduler with the model outputs. Finally, it decodes the sequences and returns the newly generated results.

Source code in vllm/engine/async_llm_engine.py

async def step_async(
    self, virtual_engine: int
) -> List[Union[RequestOutput, PoolingRequestOutput]]:
    """Performs one decoding iteration and returns newly generated results.
    The workers are ran asynchronously if possible.

    This function performs one decoding iteration of the engine. It first
    schedules the sequences to be executed in the next iteration and the
    token blocks to be swapped in/out/copy. Then, it executes the model
    and updates the scheduler with the model outputs. Finally, it decodes
    the sequences and returns the newly generated results.
    """
    # these are cached outputs from previous iterations. None if on first
    # iteration
    cached_outputs = self.cached_scheduler_outputs[virtual_engine]
    seq_group_metadata_list = cached_outputs.seq_group_metadata_list
    scheduler_outputs = cached_outputs.scheduler_outputs
    allow_async_output_proc = cached_outputs.allow_async_output_proc

    ctx = self.scheduler_contexts[virtual_engine]

    # Clear outputs for each new scheduler iteration
    ctx.request_outputs.clear()

    # skip the scheduler if there are any remaining steps in the seq groups.
    # This ensures that the scheduler is only called again when the current
    # batch has completed.
    if not self._has_remaining_steps(seq_group_metadata_list):

        # Schedule iteration
        (seq_group_metadata_list, scheduler_outputs,
         allow_async_output_proc
         ) = self.scheduler[virtual_engine].schedule()

        ctx.seq_group_metadata_list = seq_group_metadata_list
        ctx.scheduler_outputs = scheduler_outputs

        if not scheduler_outputs.is_empty():
            # this will cause mamba_cache/minimax_cache failed
            # to release finished_requests_ids of the last steps
            finished_requests_ids = self.scheduler[
                virtual_engine].get_and_reset_finished_requests_ids()

        # Maybe switch from async mode to sync mode
        if not allow_async_output_proc and len(ctx.output_queue) > 0:
            self._process_model_outputs(ctx=ctx)

        if (self.scheduler_config.is_multi_step
                and scheduler_outputs.num_lookahead_slots > 0):
            # cache the scheduler outputs for the next iteration if we have
            # lookahead slots
            self._cache_scheduler_outputs_for_multi_step(
                virtual_engine, seq_group_metadata_list, scheduler_outputs,
                allow_async_output_proc)
    else:
        finished_requests_ids = list()

    assert seq_group_metadata_list is not None
    assert scheduler_outputs is not None

    if not scheduler_outputs.is_empty():

        # Check if we have a cached last_output from the previous iteration.
        # For supporting PP this is probably the best way to pass the
        # sampled_token_ids, as a separate broadcast over all the PP stages
        # will cause one virtual engine's microbatch to block the pipeline.
        last_sampled_token_ids = \
            self._get_last_sampled_token_ids(virtual_engine)

        execute_model_req = ExecuteModelRequest(
            seq_group_metadata_list=seq_group_metadata_list,
            blocks_to_swap_in=scheduler_outputs.blocks_to_swap_in,
            blocks_to_swap_out=scheduler_outputs.blocks_to_swap_out,
            blocks_to_copy=scheduler_outputs.blocks_to_copy,
            virtual_engine=virtual_engine,
            num_lookahead_slots=scheduler_outputs.num_lookahead_slots,
            running_queue_size=scheduler_outputs.running_queue_size,
            finished_requests_ids=finished_requests_ids,
            # We use ExecuteModelRequest to pass the last sampled_token_ids
            # to each of the non-last PP stages for in-place prepare_input.
            last_sampled_token_ids=last_sampled_token_ids)

        if allow_async_output_proc:
            execute_model_req.async_callback = self.async_callbacks[
                virtual_engine]

        # Execute the model.
        outputs = await self.model_executor.execute_model_async(
            execute_model_req)

        # we need to do this here so that last step's sampled_token_ids can
        # be passed to the next iteration for PP.
        if self.scheduler_config.is_multi_step:
            self._update_cached_scheduler_output(virtual_engine, outputs)
    else:
        if len(ctx.output_queue) > 0:
            self._process_model_outputs(ctx=ctx)
        outputs = []

    # Finish the current step for all the sequence groups.
    if self.scheduler_config.is_multi_step:
        for seq_group in seq_group_metadata_list:
            seq_group.finish_step()

    if not self._has_remaining_steps(seq_group_metadata_list):
        # Clear the cache if we have finished all the steps
        if self.scheduler_config.is_multi_step:
            self.cached_scheduler_outputs[
                virtual_engine] = SchedulerOutputState()

        # is_first_step_output is True only when the num_steps of all
        # the sequences are 1. When the num_steps > 1,
        # multi_step_model_runner does the first-step output append.
        is_first_step_output: bool = False if not seq_group_metadata_list \
            else seq_group_metadata_list[0].state.num_steps == 1

        ctx.append_output(outputs=outputs,
                          seq_group_metadata_list=seq_group_metadata_list,
                          scheduler_outputs=scheduler_outputs,
                          is_async=allow_async_output_proc,
                          is_last_step=True,
                          is_first_step_output=is_first_step_output)

        if outputs and allow_async_output_proc:
            assert len(
                outputs
            ) == 1, "Async postprocessor expects only a single output set"
            self._advance_to_next_step(
                outputs[0], seq_group_metadata_list,
                scheduler_outputs.scheduled_seq_groups)

        if not allow_async_output_proc:
            self._process_model_outputs(ctx=ctx)

            # Log stats.
            self.do_log_stats(scheduler_outputs, outputs)

            # Tracing
            self.do_tracing(scheduler_outputs)

    else:
        # Multi-step case
        return ctx.request_outputs

    if not self.has_unfinished_requests():
        # Drain async postprocessor (if exists)
        if len(ctx.output_queue) > 0:
            self._process_model_outputs(ctx=ctx)
        assert len(ctx.output_queue) == 0

    return ctx.request_outputs

stop_remote_worker_execution_loop_async `async` ¶

stop_remote_worker_execution_loop_async() -> None

Stop the remote worker execution loop.

Source code in vllm/engine/async_llm_engine.py

async def stop_remote_worker_execution_loop_async(self) -> None:
    """Stop the remote worker execution loop."""
    await self.model_executor.stop_remote_worker_execution_loop_async()

_log_task_completion ¶

_log_task_completion(
    task: Task, error_callback: Callable[[Exception], None]
) -> None

This function is only intended for the engine.run_engine_loop() task.

In particular, that task runs a while True loop that can only exit if there is an exception.

Source code in vllm/engine/async_llm_engine.py

def _log_task_completion(task: asyncio.Task,
                         error_callback: Callable[[Exception], None]) -> None:
    """This function is only intended for the `engine.run_engine_loop()` task.

    In particular, that task runs a `while True` loop that can only exit if
    there is an exception.
    """

    exception = None
    try:
        return_value = task.result()
        raise AssertionError(
            f"The engine background task should never finish without an "
            f"exception. {return_value}")
    except asyncio.exceptions.CancelledError:
        # We assume that if the task is cancelled, we are gracefully shutting
        # down. This should only happen on program exit.
        logger.info("Engine is gracefully shutting down.")
    except Exception as e:
        exception = e
        logger.error("Engine background task failed", exc_info=e)
        error_callback(exception)
        raise AsyncEngineDeadError(
            "Task finished unexpectedly. This should never happen! "
            "Please open an issue on GitHub. See stack trace above for the "
            "actual cause.") from e

build_guided_decoding_logits_processor_async `async` ¶

build_guided_decoding_logits_processor_async(
    sampling_params: SamplingParams,
    tokenizer: AnyTokenizer,
    default_guided_backend: str,
    reasoning_backend: Optional[str],
    model_config: ModelConfig,
) -> SamplingParams

Constructs logits processors based on the guided_decoding, logits_bias, and allowed_token_ids fields in sampling_params. Deletes those fields and adds the constructed logits processors to the logits_processors field. Modifies sampling params in-place and returns the modified sampling params.

Source code in vllm/engine/async_llm_engine.py

async def build_guided_decoding_logits_processor_async(
        sampling_params: SamplingParams, tokenizer: AnyTokenizer,
        default_guided_backend: str, reasoning_backend: Optional[str],
        model_config: ModelConfig) -> SamplingParams:
    """Constructs logits processors based on the guided_decoding,
    logits_bias, and allowed_token_ids fields in sampling_params. Deletes
    those fields and adds the constructed logits processors to the
    logits_processors field. Modifies sampling params in-place and returns
    the modified sampling params."""
    if sampling_params.guided_decoding is None:
        return sampling_params

    # Defensively copy sampling params since guided decoding logits
    # processors can have different state for each request
    sampling_params = copy.copy(sampling_params)
    guided_decoding = sampling_params.guided_decoding

    logger.debug(
        "Building guided decoding logits processor. "
        "guided_decoding: %s%s", guided_decoding,
        f", reasoning_backend: {reasoning_backend}"
        if reasoning_backend is not None else "")

    guided_decoding.backend = guided_decoding.backend or default_guided_backend

    processor = await get_guided_decoding_logits_processor(
        guided_params=guided_decoding,
        tokenizer=tokenizer,
        reasoning_backend=reasoning_backend,
        model_config=model_config)

    if processor:
        if sampling_params.logits_processors is None:
            sampling_params.logits_processors = []
        sampling_params.logits_processors.append(processor)

    # Unset guided decoding params after constructing the lp from them
    sampling_params.guided_decoding = None

    return sampling_params

vllm.engine.async_llm_engine

ENGINE_ITERATION_TIMEOUT_S module-attribute ¶

STOP_ITERATION module-attribute ¶

logger module-attribute ¶

AsyncEngineDeadError ¶

AsyncLLMEngine ¶

_background_loop_unshielded instance-attribute ¶

_engine_class class-attribute instance-attribute ¶

_errored_with instance-attribute ¶

_request_tracker instance-attribute ¶

background_loop instance-attribute ¶

dead_error property ¶

engine instance-attribute ¶

errored property ¶

is_running property ¶

is_stopped property ¶

log_requests instance-attribute ¶

start_engine_loop instance-attribute ¶

use_process_request_outputs_callback instance-attribute ¶

__del__ ¶

__init__ ¶

_abort ¶

_engine_abort async ¶

_error_callback ¶

_get_executor_cls classmethod ¶

abort async ¶

add_logger ¶

add_lora async ¶

add_request async ¶

check_health async ¶

collective_rpc async ¶

do_log_stats async ¶

encode async ¶

engine_step async ¶

from_engine_args classmethod ¶

from_vllm_config classmethod ¶

generate async ¶

Please refer to entrypoints/api_server.py for¶

the complete example.¶

initialize the engine and the example input¶

note that engine_args here is AsyncEngineArgs instance¶

start the generation¶

get the results¶

Process and return the final output¶

get_decoding_config async ¶

get_input_preprocessor async ¶

get_lora_config async ¶

get_model_config async ¶

get_parallel_config async ¶

get_scheduler_config async ¶

get_tokenizer async ¶

get_vllm_config async ¶

is_sleeping async ¶

is_tracing_enabled async ¶

process_request_outputs ¶

remove_logger ¶

reset_mm_cache async ¶

reset_prefix_cache async ¶

run_engine_loop async staticmethod ¶

set_errored ¶

shutdown_background_loop ¶

sleep async ¶

start_background_loop ¶

start_profile async ¶

stop_profile async ¶

wake_up async ¶

AsyncStream ¶

_cancel instance-attribute ¶

_finished instance-attribute ¶

_queue instance-attribute ¶

finished property ¶

request_id instance-attribute ¶

__init__ ¶

_is_raisable staticmethod ¶

finish ¶

generator async ¶

put ¶

RequestTracker ¶

_aborted_requests instance-attribute ¶

_new_requests instance-attribute ¶

ENGINE_ITERATION_TIMEOUT_S `module-attribute` ¶

STOP_ITERATION `module-attribute` ¶

logger `module-attribute` ¶

_background_loop_unshielded `instance-attribute` ¶

_engine_class `class-attribute` `instance-attribute` ¶

_errored_with `instance-attribute` ¶

_request_tracker `instance-attribute` ¶

background_loop `instance-attribute` ¶

dead_error `property` ¶

engine `instance-attribute` ¶

errored `property` ¶

is_running `property` ¶

is_stopped `property` ¶

log_requests `instance-attribute` ¶

start_engine_loop `instance-attribute` ¶

use_process_request_outputs_callback `instance-attribute` ¶

del ¶

init ¶

_engine_abort `async` ¶

_get_executor_cls `classmethod` ¶

abort `async` ¶

add_lora `async` ¶

add_request `async` ¶

check_health `async` ¶

collective_rpc `async` ¶

do_log_stats `async` ¶

encode `async` ¶

engine_step `async` ¶

from_engine_args `classmethod` ¶

from_vllm_config `classmethod` ¶

generate `async` ¶

get_decoding_config `async` ¶

get_input_preprocessor `async` ¶

get_lora_config `async` ¶

get_model_config `async` ¶

get_parallel_config `async` ¶

get_scheduler_config `async` ¶

get_tokenizer `async` ¶

get_vllm_config `async` ¶

is_sleeping `async` ¶

is_tracing_enabled `async` ¶

reset_mm_cache `async` ¶

reset_prefix_cache `async` ¶

run_engine_loop `async` `staticmethod` ¶

sleep `async` ¶

start_profile `async` ¶

stop_profile `async` ¶

wake_up `async` ¶

_cancel `instance-attribute` ¶

_finished `instance-attribute` ¶

_queue `instance-attribute` ¶

finished `property` ¶

request_id `instance-attribute` ¶

init ¶

_is_raisable `staticmethod` ¶

generator `async` ¶

_aborted_requests `instance-attribute` ¶

_new_requests `instance-attribute` ¶

_request_streams `instance-attribute` ¶

new_requests_event `instance-attribute` ¶

contains ¶

init ¶

len ¶

wait_for_new_requests `async` ¶

init ¶

add_request_async `async` ¶

check_health_async `async` ¶

collective_rpc_async `async` ¶

get_tokenizer_async `async` ¶

step_async `async` ¶

stop_remote_worker_execution_loop_async `async` ¶

build_guided_decoding_logits_processor_async `async` ¶