vllm.config

BlockSize module-attribute

BlockSize = Literal[1, 8, 16, 32, 64, 128]

CacheDType module-attribute

CacheDType = Literal['auto', 'fp8', 'fp8_e4m3', 'fp8_e5m2']

ConfigT module-attribute

ConfigT = TypeVar('ConfigT', bound=ConfigType)

ConfigType module-attribute

ConfigType = type[DataclassInstance]

DetailedTraceModules module-attribute

DetailedTraceModules = Literal['model', 'worker', 'all']

Device module-attribute

Device = Literal[
    "auto", "cuda", "neuron", "cpu", "tpu", "xpu", "hpu"
]

DistributedExecutorBackend module-attribute

DistributedExecutorBackend = Literal[
    "ray", "mp", "uni", "external_launcher"
]

GuidedDecodingBackend module-attribute

GuidedDecodingBackend = Literal[
    GuidedDecodingBackendV0, GuidedDecodingBackendV1
]

GuidedDecodingBackendV0 module-attribute

GuidedDecodingBackendV0 = Literal[
    "auto",
    "outlines",
    "lm-format-enforcer",
    "xgrammar",
    "guidance",
]

GuidedDecodingBackendV1 module-attribute

GuidedDecodingBackendV1 = Literal[
    "auto", "xgrammar", "guidance"
]

HfOverrides module-attribute

HfOverrides = Union[
    dict[str, Any],
    Callable[[PretrainedConfig], PretrainedConfig],
]

KVConsumer module-attribute

KVConsumer = Literal['kv_consumer', 'kv_both']

KVProducer module-attribute

KVProducer = Literal['kv_producer', 'kv_both']

KVRole module-attribute

KVRole = Literal[KVProducer, KVConsumer]

LoRADType module-attribute

LoRADType = Literal['auto', 'float16', 'bfloat16']

ModelDType module-attribute

ModelDType = Literal[
    "auto",
    "half",
    "float16",
    "bfloat16",
    "float",
    "float32",
]

PreemptionMode module-attribute

PreemptionMode = Literal['swap', 'recompute']

PrefixCachingHashAlgo module-attribute

PrefixCachingHashAlgo = Literal['builtin', 'sha256']

RunnerType module-attribute

RunnerType = Literal[
    "generate", "pooling", "draft", "transcription"
]

SchedulerPolicy module-attribute

SchedulerPolicy = Literal['fcfs', 'priority']

SpeculativeAcceptanceMethod module-attribute

SpeculativeAcceptanceMethod = Literal[
    "rejection_sampler", "typical_acceptance_sampler"
]

SpeculativeMethod module-attribute

SpeculativeMethod = Literal[
    "ngram",
    "eagle",
    "eagle3",
    "medusa",
    "mlp_speculator",
    "draft_model",
    "deepseek_mtp",
]

T module-attribute

T = TypeVar('T')

TaskOption module-attribute

TaskOption = Literal[
    "auto",
    "generate",
    "embedding",
    "embed",
    "classify",
    "score",
    "reward",
    "transcription",
]

TokenizerMode module-attribute

TokenizerMode = Literal["auto", "slow", "mistral", "custom"]

_FLOAT16_NOT_SUPPORTED_MODELS module-attribute

_FLOAT16_NOT_SUPPORTED_MODELS = {
    "gemma2": "Numerical instability. Please use bfloat16 or float32 instead.",
    "gemma3": "Numerical instability. Please use bfloat16 or float32 instead.",
    "plamo2": "Numerical instability. Please use bfloat16 or float32 instead.",
    "glm4": "Numerical instability. Please use bfloat16 or float32 instead.",
}

_RUNNER_TASKS module-attribute

_RUNNER_TASKS: dict[RunnerType, list[_ResolvedTask]] = {
    "generate": ["generate"],
    "pooling": ["embed", "classify", "score", "reward"],
    "draft": ["draft"],
    "transcription": ["transcription"],
}

_ResolvedTask module-attribute

_ResolvedTask = Literal[
    "generate",
    "embed",
    "classify",
    "score",
    "reward",
    "draft",
    "transcription",
]

_STR_DTYPE_TO_TORCH_DTYPE module-attribute

_STR_DTYPE_TO_TORCH_DTYPE = {
    "half": float16,
    "float16": float16,
    "float": float32,
    "float32": float32,
    "bfloat16": bfloat16,
}

_TASK_RUNNER module-attribute

_TASK_RUNNER: dict[_ResolvedTask, RunnerType] = {
    task: _OuAlGfor(runner, tasks) in items()
    for task in tasks
}

_current_prefix module-attribute

_current_prefix: Optional[str] = None

_current_vllm_config module-attribute

_current_vllm_config: Optional[VllmConfig] = None

logger module-attribute

logger = init_logger(__name__)

CacheConfig

Configuration for the KV cache.

Source code in vllm/config.py

@config
@dataclass
class CacheConfig:
    """Configuration for the KV cache."""

    block_size: SkipValidation[BlockSize] = None  # type: ignore
    """Size of a contiguous cache block in number of tokens. This is ignored on
    neuron devices and set to `--max-model-len`. On CUDA devices, only block
    sizes up to 32 are supported. On HPU devices, block size defaults to 128.

    This config has no static default. If left unspecified by the user, it will
    be set in `Platform.check_and_update_configs()` based on the current
    platform."""
    gpu_memory_utilization: float = 0.9
    """The fraction of GPU memory to be used for the model executor, which can
    range from 0 to 1. For example, a value of 0.5 would imply 50% GPU memory
    utilization. If unspecified, will use the default value of 0.9. This is a
    per-instance limit, and only applies to the current vLLM instance. It does
    not matter if you have another vLLM instance running on the same GPU. For
    example, if you have two vLLM instances running on the same GPU, you can
    set the GPU memory utilization to 0.5 for each instance."""
    swap_space: float = 4
    """Size of the CPU swap space per GPU (in GiB)."""
    cache_dtype: CacheDType = "auto"
    """Data type for kv cache storage. If "auto", will use model data type.
    CUDA 11.8+ supports fp8 (=fp8_e4m3) and fp8_e5m2. ROCm (AMD GPU) supports
    fp8 (=fp8_e4m3)."""
    is_attention_free: bool = False
    """Whether the model is attention-free. This is primarily set in
    `ModelConfig` and that value should be manually duplicated here."""
    num_gpu_blocks_override: Optional[int] = None
    """Number of GPU blocks to use. This overrides the profiled `num_gpu_blocks`
    if specified. Does nothing if `None`. Used for testing preemption."""
    sliding_window: Optional[int] = None
    """Sliding window size for the KV cache. This is primarily set in
    `ModelConfig` and that value should be manually duplicated here."""
    enable_prefix_caching: Optional[bool] = None
    """Whether to enable prefix caching. Disabled by default for V0. Enabled by
    default for V1."""
    prefix_caching_hash_algo: PrefixCachingHashAlgo = "builtin"
    """Set the hash algorithm for prefix caching:\n
    - "builtin" is Python's built-in hash.\n
    - "sha256" is collision resistant but with certain overheads."""
    cpu_offload_gb: float = 0
    """The space in GiB to offload to CPU, per GPU. Default is 0, which means
    no offloading. Intuitively, this argument can be seen as a virtual way to
    increase the GPU memory size. For example, if you have one 24 GB GPU and
    set this to 10, virtually you can think of it as a 34 GB GPU. Then you can
    load a 13B model with BF16 weight, which requires at least 26GB GPU memory.
    Note that this requires fast CPU-GPU interconnect, as part of the model is
    loaded from CPU memory to GPU memory on the fly in each model forward pass.
    """
    calculate_kv_scales: bool = False
    """This enables dynamic calculation of `k_scale` and `v_scale` when
    kv_cache_dtype is fp8. If `False`, the scales will be loaded from the model
    checkpoint if available. Otherwise, the scales will default to 1.0."""
    cpu_kvcache_space_bytes: Optional[int] = None
    """(CPU backend only) CPU key-value cache space."""

    # Will be set after profiling.
    num_gpu_blocks: Optional[int] = field(default=None, init=False)
    """The number of blocks to allocate for GPU memory."""
    num_cpu_blocks: Optional[int] = field(default=None, init=False)
    """The number of blocks to allocate for CPU memory."""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        factors: list[Any] = []
        factors.append(self.cache_dtype)
        # `cpu_offload_gb` does not use `torch.compile` yet.
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self) -> None:
        self.swap_space_bytes = self.swap_space * GiB_bytes

        self._verify_cache_dtype()
        self._verify_prefix_caching()

    def metrics_info(self):
        # convert cache_config to dict(key: str, value: str) for prometheus
        # metrics info
        return {key: str(value) for key, value in self.__dict__.items()}

    @model_validator(mode='after')
    def _verify_args(self) -> Self:
        if self.cpu_offload_gb < 0:
            raise ValueError("CPU offload space must be non-negative"
                             f", but got {self.cpu_offload_gb}")

        if self.gpu_memory_utilization > 1.0:
            raise ValueError(
                "GPU memory utilization must be less than 1.0. Got "
                f"{self.gpu_memory_utilization}.")

        return self

    def _verify_cache_dtype(self) -> None:
        if self.cache_dtype == "auto":
            pass
        elif self.cache_dtype in get_args(CacheDType):
            logger.info(
                "Using fp8 data type to store kv cache. It reduces the GPU "
                "memory footprint and boosts the performance. "
                "Meanwhile, it may cause accuracy drop without a proper "
                "scaling factor")
        else:
            raise ValueError(f"Unknown kv cache dtype: {self.cache_dtype}")

    def _verify_prefix_caching(self) -> None:
        if not self.enable_prefix_caching:
            return

        if self.sliding_window is not None and not envs.VLLM_USE_V1:
            raise NotImplementedError(
                "Prefix caching is not supported with sliding window. "
                "Run with --disable-sliding-window to use prefix caching.")

        if (self.enable_prefix_caching and self.prefix_caching_hash_algo
                not in get_args(PrefixCachingHashAlgo)):
            raise ValueError(
                "Unknown prefix caching hash algorithm: "
                f"{self.prefix_caching_hash_algo}. Must be one of "
                f"{get_args(PrefixCachingHashAlgo)}.")

    def verify_with_parallel_config(
        self,
        parallel_config: "ParallelConfig",
    ) -> None:
        total_cpu_memory = get_cpu_memory()
        # FIXME(woosuk): Here, it is assumed that the GPUs in a tensor parallel
        # group are in the same node. However, the GPUs may span multiple nodes.
        num_gpus_per_node = parallel_config.tensor_parallel_size
        cpu_memory_usage = self.swap_space_bytes * num_gpus_per_node

        msg = (f"{cpu_memory_usage / GiB_bytes:.2f} GiB out of the "
               f"{total_cpu_memory / GiB_bytes:.2f} GiB total CPU memory "
               "is allocated for the swap space.")
        if cpu_memory_usage > 0.7 * total_cpu_memory:
            raise ValueError("Too large swap space. " + msg)
        elif cpu_memory_usage > 0.4 * total_cpu_memory:
            logger.warning("Possibly too large swap space. %s", msg)

block_size class-attribute instance-attribute

block_size: SkipValidation[BlockSize] = None

Size of a contiguous cache block in number of tokens. This is ignored on neuron devices and set to --max-model-len. On CUDA devices, only block sizes up to 32 are supported. On HPU devices, block size defaults to 128.

This config has no static default. If left unspecified by the user, it will be set in Platform.check_and_update_configs() based on the current platform.

cache_dtype class-attribute instance-attribute

cache_dtype: CacheDType = 'auto'

Data type for kv cache storage. If "auto", will use model data type. CUDA 11.8+ supports fp8 (=fp8_e4m3) and fp8_e5m2. ROCm (AMD GPU) supports fp8 (=fp8_e4m3).

calculate_kv_scales class-attribute instance-attribute

calculate_kv_scales: bool = False

This enables dynamic calculation of k_scale and v_scale when kv_cache_dtype is fp8. If False, the scales will be loaded from the model checkpoint if available. Otherwise, the scales will default to 1.0.

cpu_kvcache_space_bytes class-attribute instance-attribute

cpu_kvcache_space_bytes: Optional[int] = None

(CPU backend only) CPU key-value cache space.

cpu_offload_gb class-attribute instance-attribute

cpu_offload_gb: float = 0

The space in GiB to offload to CPU, per GPU. Default is 0, which means no offloading. Intuitively, this argument can be seen as a virtual way to increase the GPU memory size. For example, if you have one 24 GB GPU and set this to 10, virtually you can think of it as a 34 GB GPU. Then you can load a 13B model with BF16 weight, which requires at least 26GB GPU memory. Note that this requires fast CPU-GPU interconnect, as part of the model is loaded from CPU memory to GPU memory on the fly in each model forward pass.

enable_prefix_caching class-attribute instance-attribute

enable_prefix_caching: Optional[bool] = None

Whether to enable prefix caching. Disabled by default for V0. Enabled by default for V1.

gpu_memory_utilization class-attribute instance-attribute

gpu_memory_utilization: float = 0.9

The fraction of GPU memory to be used for the model executor, which can range from 0 to 1. For example, a value of 0.5 would imply 50% GPU memory utilization. If unspecified, will use the default value of 0.9. This is a per-instance limit, and only applies to the current vLLM instance. It does not matter if you have another vLLM instance running on the same GPU. For example, if you have two vLLM instances running on the same GPU, you can set the GPU memory utilization to 0.5 for each instance.

is_attention_free class-attribute instance-attribute

is_attention_free: bool = False

Whether the model is attention-free. This is primarily set in ModelConfig and that value should be manually duplicated here.

num_cpu_blocks class-attribute instance-attribute

num_cpu_blocks: Optional[int] = field(
    default=None, init=False
)

The number of blocks to allocate for CPU memory.

num_gpu_blocks class-attribute instance-attribute

num_gpu_blocks: Optional[int] = field(
    default=None, init=False
)

The number of blocks to allocate for GPU memory.

num_gpu_blocks_override class-attribute instance-attribute

num_gpu_blocks_override: Optional[int] = None

Number of GPU blocks to use. This overrides the profiled num_gpu_blocks if specified. Does nothing if None. Used for testing preemption.

prefix_caching_hash_algo class-attribute instance-attribute

prefix_caching_hash_algo: PrefixCachingHashAlgo = 'builtin'

Set the hash algorithm for prefix caching:

• "builtin" is Python's built-in hash.

• "sha256" is collision resistant but with certain overheads.

sliding_window class-attribute instance-attribute

sliding_window: Optional[int] = None

Sliding window size for the KV cache. This is primarily set in ModelConfig and that value should be manually duplicated here.

swap_space class-attribute instance-attribute

swap_space: float = 4

Size of the CPU swap space per GPU (in GiB).

__post_init__

__post_init__() -> None

Source code in vllm/config.py

def __post_init__(self) -> None:
    self.swap_space_bytes = self.swap_space * GiB_bytes

    self._verify_cache_dtype()
    self._verify_prefix_caching()

_verify_args

_verify_args() -> Self

Source code in vllm/config.py

@model_validator(mode='after')
def _verify_args(self) -> Self:
    if self.cpu_offload_gb < 0:
        raise ValueError("CPU offload space must be non-negative"
                         f", but got {self.cpu_offload_gb}")

    if self.gpu_memory_utilization > 1.0:
        raise ValueError(
            "GPU memory utilization must be less than 1.0. Got "
            f"{self.gpu_memory_utilization}.")

    return self

_verify_cache_dtype

_verify_cache_dtype() -> None

Source code in vllm/config.py

def _verify_cache_dtype(self) -> None:
    if self.cache_dtype == "auto":
        pass
    elif self.cache_dtype in get_args(CacheDType):
        logger.info(
            "Using fp8 data type to store kv cache. It reduces the GPU "
            "memory footprint and boosts the performance. "
            "Meanwhile, it may cause accuracy drop without a proper "
            "scaling factor")
    else:
        raise ValueError(f"Unknown kv cache dtype: {self.cache_dtype}")

_verify_prefix_caching

_verify_prefix_caching() -> None

Source code in vllm/config.py

def _verify_prefix_caching(self) -> None:
    if not self.enable_prefix_caching:
        return

    if self.sliding_window is not None and not envs.VLLM_USE_V1:
        raise NotImplementedError(
            "Prefix caching is not supported with sliding window. "
            "Run with --disable-sliding-window to use prefix caching.")

    if (self.enable_prefix_caching and self.prefix_caching_hash_algo
            not in get_args(PrefixCachingHashAlgo)):
        raise ValueError(
            "Unknown prefix caching hash algorithm: "
            f"{self.prefix_caching_hash_algo}. Must be one of "
            f"{get_args(PrefixCachingHashAlgo)}.")

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    factors: list[Any] = []
    factors.append(self.cache_dtype)
    # `cpu_offload_gb` does not use `torch.compile` yet.
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

metrics_info

metrics_info()

Source code in vllm/config.py

def metrics_info(self):
    # convert cache_config to dict(key: str, value: str) for prometheus
    # metrics info
    return {key: str(value) for key, value in self.__dict__.items()}

verify_with_parallel_config

verify_with_parallel_config(
    parallel_config: ParallelConfig,
) -> None

Source code in vllm/config.py

def verify_with_parallel_config(
    self,
    parallel_config: "ParallelConfig",
) -> None:
    total_cpu_memory = get_cpu_memory()
    # FIXME(woosuk): Here, it is assumed that the GPUs in a tensor parallel
    # group are in the same node. However, the GPUs may span multiple nodes.
    num_gpus_per_node = parallel_config.tensor_parallel_size
    cpu_memory_usage = self.swap_space_bytes * num_gpus_per_node

    msg = (f"{cpu_memory_usage / GiB_bytes:.2f} GiB out of the "
           f"{total_cpu_memory / GiB_bytes:.2f} GiB total CPU memory "
           "is allocated for the swap space.")
    if cpu_memory_usage > 0.7 * total_cpu_memory:
        raise ValueError("Too large swap space. " + msg)
    elif cpu_memory_usage > 0.4 * total_cpu_memory:
        logger.warning("Possibly too large swap space. %s", msg)

CompilationConfig

Configuration for compilation. It has three parts:

• Top-level Compilation control:
  • level
  • debug_dump_path
  • cache_dir
  • backend
  • custom_ops
  • splitting_ops
• CudaGraph capture:
  • use_cudagraph
  • cudagraph_capture_sizes
  • cudagraph_num_of_warmups
  • cudagraph_copy_inputs
  • full_cuda_graph
• Inductor compilation:
  • use_inductor
  • compile_sizes
  • inductor_compile_config
  • inductor_passes
  • custom inductor passes

Why we have different sizes for cudagraph and inductor: - cudagraph: a cudagraph captured for a specific size can only be used for the same size. We need to capture all the sizes we want to use. - inductor: a graph compiled by inductor for a general shape can be used for different sizes. Inductor can also compile for specific sizes, where it can have more information to optimize the graph with fully static shapes. However, we find the general shape compilation is sufficient for most cases. It might be beneficial to compile for certain small batchsizes, where inductor is good at optimizing.

Source code in vllm/config.py

@config
@dataclass
class CompilationConfig:
    """Configuration for compilation. It has three parts:

    - Top-level Compilation control:
        - [`level`][vllm.config.CompilationConfig.level]
        - [`debug_dump_path`][vllm.config.CompilationConfig.debug_dump_path]
        - [`cache_dir`][vllm.config.CompilationConfig.cache_dir]
        - [`backend`][vllm.config.CompilationConfig.backend]
        - [`custom_ops`][vllm.config.CompilationConfig.custom_ops]
        - [`splitting_ops`][vllm.config.CompilationConfig.splitting_ops]
    - CudaGraph capture:
        - [`use_cudagraph`][vllm.config.CompilationConfig.use_cudagraph]
        - [`cudagraph_capture_sizes`]
        [vllm.config.CompilationConfig.cudagraph_capture_sizes]
        - [`cudagraph_num_of_warmups`]
        [vllm.config.CompilationConfig.cudagraph_num_of_warmups]
        - [`cudagraph_copy_inputs`]
        [vllm.config.CompilationConfig.cudagraph_copy_inputs]
        - [`full_cuda_graph`][vllm.config.CompilationConfig.full_cuda_graph]
    - Inductor compilation:
        - [`use_inductor`][vllm.config.CompilationConfig.use_inductor]
        - [`compile_sizes`][vllm.config.CompilationConfig.compile_sizes]
        - [`inductor_compile_config`]
        [vllm.config.CompilationConfig.inductor_compile_config]
        - [`inductor_passes`][vllm.config.CompilationConfig.inductor_passes]
        - custom inductor passes

    Why we have different sizes for cudagraph and inductor:
    - cudagraph: a cudagraph captured for a specific size can only be used
        for the same size. We need to capture all the sizes we want to use.
    - inductor: a graph compiled by inductor for a general shape can be used
        for different sizes. Inductor can also compile for specific sizes,
        where it can have more information to optimize the graph with fully
        static shapes. However, we find the general shape compilation is
        sufficient for most cases. It might be beneficial to compile for
        certain small batchsizes, where inductor is good at optimizing.
    """
    # Top-level Compilation control
    level: int = 0
    """The level of compilation:

    - 0: no compilation.
    - 1: dynamo as is.
    - 2: dynamo once.
    - 3: piecewise compilation."""
    debug_dump_path: str = ""
    """The path to dump the debug information."""
    cache_dir: str = ""
    """The directory to store the compiled graph, to accelerate Inductor
    compilation. By default, it will use model-related information to generate
    a cache directory."""
    backend: str = ""
    """The backend for compilation. It needs to be a string:

    - "" (empty string): use the default backend.
    - "eager"/"openxla"/...: use the specified backend registered in PyTorch.
    - "full.module.name": a qualified name which can be used to import the

    backend function.
    We use string to avoid serialization issues when using compilation in a
    distributed setting. When the compilation level is 1 or 2, the backend is
    used for the compilation directly (it sees the whole graph). When the
    compilation level is 3, the backend is used for the piecewise compilation
    (it sees a part of the graph)."""
    custom_ops: list[str] = field(default_factory=list)
    """Fine-grained control over which custom ops to enable/disable. Use 'all'
    to enable all, 'none' to disable all. Also specify a list of custom op
    names to enable (prefixed with a '+'), or disable (prefixed with a '-').
    Examples:

    - 'all,-op1' to enable all except op1
    - 'none,+op1,+op2' to enable only op1 and op2

    By default, all custom ops are enabled when running without Inductor and
    disabled when running with Inductor (compile_level >= Inductor)."""
    splitting_ops: list[str] = field(default_factory=list)
    """A list of ops to split the full graph into subgraphs, used in piecewise
    compilation."""

    # Inductor capture
    use_inductor: bool = True
    """Whether to use inductor compilation:

    - False: inductor compilation is not used. graph runs in eager.
    - True: inductor compilation is used. one graph for symbolic shape
        is compiled. In addition, compile for compile_sizes,
        using configurations in inductor_compile_config."""
    compile_sizes: Optional[list[Union[int, str]]] = None
    """Sizes to compile for inductor. In addition
    to integers, it also supports "cudagraph_capture_sizes" to
    specify the sizes for cudagraph capture."""
    inductor_compile_config: dict = field(default_factory=dict)
    """Additional configurations for inductor.
    - None: use default configurations."""
    inductor_passes: dict[str, str] = field(default_factory=dict)
    """Additional passes for inductor. It is a dictionary
    from pass name to pass function qualified name. We use function
    name because the config uses JSON format. If we pass the config
    from Python, functions can also be passed directly via Python object
    constructor, e.g. `CompilationConfig(inductor_passes={"a": func})`."""

    # CudaGraph compilation
    use_cudagraph: bool = field(default_factory=lambda: envs.VLLM_USE_V1)
    """Whether to use cudagraph inside compilation.
    - False: cudagraph inside compilation is not used.
    - True: cudagraph inside compilation is used. It requires
        that all input buffers have fixed addresses, and all
        splitting ops write their outputs to input buffers.
    In the vLLM V1 Engine, this flag only applies for
    CompilationLevel.PIECEWISE (aka -O3).
    Note that this is orthogonal to the cudagraph capture logic
    outside of compilation.
    TODO: move outside cudagraph logic into compilation.
    torch.compile will handle cudagraph capture logic in the future."""
    cudagraph_num_of_warmups: int = 0
    """Number of warmup runs for cudagraph.
    It means the first several runs will be treated as warmup runs.
    Only after that, the execution will be recorded, and the recorded
    cudagraph will be used for subsequent runs."""
    cudagraph_capture_sizes: Optional[list[int]] = None
    """Sizes to capture cudagraph.
    - None (default): capture sizes are inferred from vllm config.
    - list[int]: capture sizes are specified as given."""
    cudagraph_copy_inputs: bool = False
    """Whether to copy input tensors for
    cudagraph. If the caller can guarantee that the same input buffers
    are always used, it can set this to False. Otherwise, it should
    set this to True, and the compiler will copy the input to an
    internally managed buffer. Default is False."""
    full_cuda_graph: bool = False
    """whether to use a full cuda graph for the entire forward pass rather than
    splitting certain operations such as attention into subgraphs. Thus this
    flag cannot be used together with splitting_ops. This may provide
    performance benefits for smaller models."""

    pass_config: PassConfig = field(default_factory=PassConfig)
    """Custom inductor passes, see PassConfig for more details"""

    max_capture_size: int = field(default=None, init=False)  # type: ignore
    """not configurable, computed after init"""
    local_cache_dir: str = field(default=None, init=False)  # type: ignore
    """local cache dir for each rank"""
    bs_to_padded_graph_size: list[int] = field(
        default=None,  # type: ignore
        init=False)
    """optimization:
    Intuitively, bs_to_padded_graph_size should be dict[int, int].
    since we know all keys are in a range [0, max_capture_size],
    we can optimize it to list[int] for better lookup performance."""

    # keep track of enabled and disabled custom ops
    enabled_custom_ops: Counter[str] = field(default_factory=Counter,
                                             init=False)
    """custom ops that are enabled"""
    disabled_custom_ops: Counter[str] = field(default_factory=Counter,
                                              init=False)
    """custom ops that are disabled"""
    traced_files: set[str] = field(default_factory=set, init=False)
    """files that are traced for compilation"""
    compilation_time: float = field(default=0.0, init=False)
    """time taken for compilation"""

    static_forward_context: dict[str, Any] = field(default_factory=dict,
                                                   init=False)
    """Per-model forward context
    Map from layer name to layer objects that need to be accessed outside
    model code, e.g., Attention, FusedMOE when dp_size>1."""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        factors: list[Any] = []
        factors.append(self.level)
        factors.append(self.backend)
        factors.append(self.custom_ops)
        factors.append(self.splitting_ops)
        factors.append(self.use_inductor)
        factors.append(self.inductor_compile_config)
        factors.append(self.inductor_passes)
        factors.append(self.pass_config.uuid())
        return hashlib.sha256(str(factors).encode()).hexdigest()

    def __repr__(self) -> str:
        exclude = {
            "static_forward_context": True,
            "enabled_custom_ops": True,
            "disabled_custom_ops": True,
            "compilation_time": True,
            "bs_to_padded_graph_size": True,
            "pass_config": True,
            "traced_files": True,
            "inductor_compile_config": {
                "post_grad_custom_post_pass": True,
            },
        }
        # The cast to string is necessary because Pydantic is mocked in docs
        # builds and sphinx-argparse doesn't know the return type of decode()
        return str(
            TypeAdapter(CompilationConfig).dump_json(
                self,
                exclude=exclude,  # type: ignore[arg-type]
                exclude_unset=True).decode())

    __str__ = __repr__

    @classmethod
    def from_cli(cls, cli_value: str) -> "CompilationConfig":
        """Parse the CLI value for the compilation config."""
        if cli_value in ["0", "1", "2", "3"]:
            return cls(level=int(cli_value))
        return TypeAdapter(CompilationConfig).validate_json(cli_value)

    def __post_init__(self) -> None:
        count_none = self.custom_ops.count("none")
        count_all = self.custom_ops.count("all")
        assert count_none + count_all <= 1, "Can only specify 'none' or 'all'"

        # TODO(zou3519/luka): There are 2 issues with auto-functionalization V2:
        # 1. A bug in PyTorch, fixed in 2.7:
        #    https://github.com/pytorch/pytorch/issues/147924
        # 2. Custom passes (fusion) rely on auto-functionalization V1 and don't
        #    work with V2. Addressing this will take extra engineering effort
        #    and it is not yet a priority. RFC here:
        #    https://github.com/vllm-project/vllm/issues/14703

        if is_torch_equal_or_newer("2.6"):
            KEY = 'enable_auto_functionalized_v2'
            if KEY not in self.inductor_compile_config:
                self.inductor_compile_config[KEY] = False

        for k, v in self.inductor_passes.items():
            if not isinstance(v, str):
                assert callable(v), (
                    f"pass {k} should be callable or a qualified name")
                self.inductor_compile_config[k] = v if isinstance(
                    v, InductorPass) else CallableInductorPass(v)
                continue

            # resolve function from qualified name
            names = v.split(".")
            module = ".".join(names[:-1])
            func_name = names[-1]
            func = __import__(module).__dict__[func_name]
            self.inductor_compile_config[k] = func if isinstance(
                func, InductorPass) else CallableInductorPass(func)

        if isinstance(self.pass_config, dict):
            self.pass_config = PassConfig(**self.pass_config)

    def init_backend(self, vllm_config: "VllmConfig") -> Union[str, Callable]:
        if self.level == CompilationLevel.NO_COMPILATION:
            raise ValueError("No compilation level is set.")

        from torch._dynamo.backends.registry import list_backends
        torch_backends = list_backends(exclude_tags=tuple())
        if self.level in [
                CompilationLevel.DYNAMO_AS_IS, CompilationLevel.DYNAMO_ONCE
        ]:
            if self.backend == "":
                return "eager"
            if self.backend in torch_backends:
                return self.backend
            return resolve_obj_by_qualname(self.backend)

        # TODO: pass user-specified backend to piecewise compilation
        # merge with the config use_inductor
        assert self.level == CompilationLevel.PIECEWISE

        from vllm.compilation.backends import VllmBackend
        return VllmBackend(vllm_config)

    def init_with_cudagraph_sizes(self,
                                  cudagraph_capture_sizes: list[int]) -> None:
        """To complete the initialization of config,
        we need to know the cudagraph sizes."""

        if self.cudagraph_capture_sizes is None:
            self.cudagraph_capture_sizes = cudagraph_capture_sizes
        else:
            # de-duplicate the sizes provided by the config
            dedup_sizes = list(set(self.cudagraph_capture_sizes))
            if len(dedup_sizes) < len(self.cudagraph_capture_sizes):
                logger.info(("cudagraph sizes specified by model runner"
                             " %s is overridden by config %s"),
                            cudagraph_capture_sizes, dedup_sizes)
            self.cudagraph_capture_sizes = dedup_sizes

        computed_compile_sizes = []
        if self.compile_sizes is not None:
            # de-duplicate the sizes provided by the config
            self.compile_sizes = list(set(self.compile_sizes))
            for x in self.compile_sizes:
                if isinstance(x, str):
                    assert x == "cudagraph_capture_sizes", \
                    "Unrecognized size type in compile_sizes, " \
                    f"expect 'cudagraph_capture_sizes', got {x}"
                    computed_compile_sizes.extend(self.cudagraph_capture_sizes)
                else:
                    assert isinstance(x, int)
                    computed_compile_sizes.append(x)
        self.compile_sizes = computed_compile_sizes  # type: ignore

        # sort to make sure cudagraph capture sizes are in descending order
        self.cudagraph_capture_sizes.sort(reverse=True)
        self.max_capture_size = self.cudagraph_capture_sizes[
            0] if self.cudagraph_capture_sizes else 0

        # pre-compute the mapping from batch size to padded graph size
        self.bs_to_padded_graph_size = [
            0 for i in range(self.max_capture_size + 1)
        ]
        for end, start in zip(self.cudagraph_capture_sizes,
                              self.cudagraph_capture_sizes[1:] + [0]):
            for bs in range(start, end):
                if bs == start:
                    self.bs_to_padded_graph_size[bs] = start
                else:
                    self.bs_to_padded_graph_size[bs] = end
        self.bs_to_padded_graph_size[
            self.max_capture_size] = self.max_capture_size

    def set_splitting_ops_for_v1(self):
        # NOTE: this function needs to be called
        if self.splitting_ops and self.full_cuda_graph:
            raise ValueError("full_cuda_graph cannot be used together with "
                             "splitting_ops, as Full CUDA graph will override "
                             f"the splitting_ops: {self.splitting_ops}")

        if not self.splitting_ops:
            self.splitting_ops = [] if self.full_cuda_graph else [
                "vllm.unified_attention",
                "vllm.unified_attention_with_output",
            ]

__str__ class-attribute instance-attribute

__str__ = __repr__

backend class-attribute instance-attribute

backend: str = ''

The backend for compilation. It needs to be a string:

• "" (empty string): use the default backend.
• "eager"/"openxla"/...: use the specified backend registered in PyTorch.
• "full.module.name": a qualified name which can be used to import the

backend function. We use string to avoid serialization issues when using compilation in a distributed setting. When the compilation level is 1 or 2, the backend is used for the compilation directly (it sees the whole graph). When the compilation level is 3, the backend is used for the piecewise compilation (it sees a part of the graph).

bs_to_padded_graph_size class-attribute instance-attribute

bs_to_padded_graph_size: list[int] = field(
    default=None, init=False
)

optimization: Intuitively, bs_to_padded_graph_size should be dict[int, int]. since we know all keys are in a range [0, max_capture_size], we can optimize it to list[int] for better lookup performance.

cache_dir class-attribute instance-attribute

cache_dir: str = ''

The directory to store the compiled graph, to accelerate Inductor compilation. By default, it will use model-related information to generate a cache directory.

compilation_time class-attribute instance-attribute

compilation_time: float = field(default=0.0, init=False)

time taken for compilation

compile_sizes class-attribute instance-attribute

compile_sizes: Optional[list[Union[int, str]]] = None

Sizes to compile for inductor. In addition to integers, it also supports "cudagraph_capture_sizes" to specify the sizes for cudagraph capture.

cudagraph_capture_sizes class-attribute instance-attribute

cudagraph_capture_sizes: Optional[list[int]] = None

Sizes to capture cudagraph. - None (default): capture sizes are inferred from vllm config. - list[int]: capture sizes are specified as given.

cudagraph_copy_inputs class-attribute instance-attribute

cudagraph_copy_inputs: bool = False

Whether to copy input tensors for cudagraph. If the caller can guarantee that the same input buffers are always used, it can set this to False. Otherwise, it should set this to True, and the compiler will copy the input to an internally managed buffer. Default is False.

cudagraph_num_of_warmups class-attribute instance-attribute

cudagraph_num_of_warmups: int = 0

Number of warmup runs for cudagraph. It means the first several runs will be treated as warmup runs. Only after that, the execution will be recorded, and the recorded cudagraph will be used for subsequent runs.

custom_ops class-attribute instance-attribute

custom_ops: list[str] = field(default_factory=list)

Fine-grained control over which custom ops to enable/disable. Use 'all' to enable all, 'none' to disable all. Also specify a list of custom op names to enable (prefixed with a '+'), or disable (prefixed with a '-'). Examples:

• 'all,-op1' to enable all except op1
• 'none,+op1,+op2' to enable only op1 and op2

By default, all custom ops are enabled when running without Inductor and disabled when running with Inductor (compile_level >= Inductor).

debug_dump_path class-attribute instance-attribute

debug_dump_path: str = ''

The path to dump the debug information.

disabled_custom_ops class-attribute instance-attribute

disabled_custom_ops: Counter[str] = field(
    default_factory=Counter, init=False
)

custom ops that are disabled

enabled_custom_ops class-attribute instance-attribute

enabled_custom_ops: Counter[str] = field(
    default_factory=Counter, init=False
)

custom ops that are enabled

full_cuda_graph class-attribute instance-attribute

full_cuda_graph: bool = False

whether to use a full cuda graph for the entire forward pass rather than splitting certain operations such as attention into subgraphs. Thus this flag cannot be used together with splitting_ops. This may provide performance benefits for smaller models.

inductor_compile_config class-attribute instance-attribute

inductor_compile_config: dict = field(default_factory=dict)

Additional configurations for inductor. - None: use default configurations.

inductor_passes class-attribute instance-attribute

inductor_passes: dict[str, str] = field(
    default_factory=dict
)

Additional passes for inductor. It is a dictionary from pass name to pass function qualified name. We use function name because the config uses JSON format. If we pass the config from Python, functions can also be passed directly via Python object constructor, e.g. CompilationConfig(inductor_passes={"a": func}).

level class-attribute instance-attribute

level: int = 0

The level of compilation:

• 0: no compilation.
• 1: dynamo as is.
• 2: dynamo once.
• 3: piecewise compilation.

local_cache_dir class-attribute instance-attribute

local_cache_dir: str = field(default=None, init=False)

local cache dir for each rank

max_capture_size class-attribute instance-attribute

max_capture_size: int = field(default=None, init=False)

not configurable, computed after init

pass_config class-attribute instance-attribute

pass_config: PassConfig = field(default_factory=PassConfig)

Custom inductor passes, see PassConfig for more details

splitting_ops class-attribute instance-attribute

splitting_ops: list[str] = field(default_factory=list)

A list of ops to split the full graph into subgraphs, used in piecewise compilation.

static_forward_context class-attribute instance-attribute

static_forward_context: dict[str, Any] = field(
    default_factory=dict, init=False
)

Per-model forward context Map from layer name to layer objects that need to be accessed outside model code, e.g., Attention, FusedMOE when dp_size>1.

traced_files class-attribute instance-attribute

traced_files: set[str] = field(
    default_factory=set, init=False
)

files that are traced for compilation

use_cudagraph class-attribute instance-attribute

use_cudagraph: bool = field(
    default_factory=lambda: VLLM_USE_V1
)

Whether to use cudagraph inside compilation. - False: cudagraph inside compilation is not used. - True: cudagraph inside compilation is used. It requires that all input buffers have fixed addresses, and all splitting ops write their outputs to input buffers. In the vLLM V1 Engine, this flag only applies for CompilationLevel.PIECEWISE (aka -O3). Note that this is orthogonal to the cudagraph capture logic outside of compilation. TODO: move outside cudagraph logic into compilation. torch.compile will handle cudagraph capture logic in the future.

use_inductor class-attribute instance-attribute

use_inductor: bool = True

Whether to use inductor compilation:

• False: inductor compilation is not used. graph runs in eager.
• True: inductor compilation is used. one graph for symbolic shape is compiled. In addition, compile for compile_sizes, using configurations in inductor_compile_config.

__post_init__

__post_init__() -> None

Source code in vllm/config.py

def __post_init__(self) -> None:
    count_none = self.custom_ops.count("none")
    count_all = self.custom_ops.count("all")
    assert count_none + count_all <= 1, "Can only specify 'none' or 'all'"

    # TODO(zou3519/luka): There are 2 issues with auto-functionalization V2:
    # 1. A bug in PyTorch, fixed in 2.7:
    #    https://github.com/pytorch/pytorch/issues/147924
    # 2. Custom passes (fusion) rely on auto-functionalization V1 and don't
    #    work with V2. Addressing this will take extra engineering effort
    #    and it is not yet a priority. RFC here:
    #    https://github.com/vllm-project/vllm/issues/14703

    if is_torch_equal_or_newer("2.6"):
        KEY = 'enable_auto_functionalized_v2'
        if KEY not in self.inductor_compile_config:
            self.inductor_compile_config[KEY] = False

    for k, v in self.inductor_passes.items():
        if not isinstance(v, str):
            assert callable(v), (
                f"pass {k} should be callable or a qualified name")
            self.inductor_compile_config[k] = v if isinstance(
                v, InductorPass) else CallableInductorPass(v)
            continue

        # resolve function from qualified name
        names = v.split(".")
        module = ".".join(names[:-1])
        func_name = names[-1]
        func = __import__(module).__dict__[func_name]
        self.inductor_compile_config[k] = func if isinstance(
            func, InductorPass) else CallableInductorPass(func)

    if isinstance(self.pass_config, dict):
        self.pass_config = PassConfig(**self.pass_config)

__repr__

__repr__() -> str

Source code in vllm/config.py

def __repr__(self) -> str:
    exclude = {
        "static_forward_context": True,
        "enabled_custom_ops": True,
        "disabled_custom_ops": True,
        "compilation_time": True,
        "bs_to_padded_graph_size": True,
        "pass_config": True,
        "traced_files": True,
        "inductor_compile_config": {
            "post_grad_custom_post_pass": True,
        },
    }
    # The cast to string is necessary because Pydantic is mocked in docs
    # builds and sphinx-argparse doesn't know the return type of decode()
    return str(
        TypeAdapter(CompilationConfig).dump_json(
            self,
            exclude=exclude,  # type: ignore[arg-type]
            exclude_unset=True).decode())

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    factors: list[Any] = []
    factors.append(self.level)
    factors.append(self.backend)
    factors.append(self.custom_ops)
    factors.append(self.splitting_ops)
    factors.append(self.use_inductor)
    factors.append(self.inductor_compile_config)
    factors.append(self.inductor_passes)
    factors.append(self.pass_config.uuid())
    return hashlib.sha256(str(factors).encode()).hexdigest()

from_cli classmethod

from_cli(cli_value: str) -> CompilationConfig

Parse the CLI value for the compilation config.

Source code in vllm/config.py

@classmethod
def from_cli(cls, cli_value: str) -> "CompilationConfig":
    """Parse the CLI value for the compilation config."""
    if cli_value in ["0", "1", "2", "3"]:
        return cls(level=int(cli_value))
    return TypeAdapter(CompilationConfig).validate_json(cli_value)

init_backend

init_backend(
    vllm_config: VllmConfig,
) -> Union[str, Callable]

Source code in vllm/config.py

def init_backend(self, vllm_config: "VllmConfig") -> Union[str, Callable]:
    if self.level == CompilationLevel.NO_COMPILATION:
        raise ValueError("No compilation level is set.")

    from torch._dynamo.backends.registry import list_backends
    torch_backends = list_backends(exclude_tags=tuple())
    if self.level in [
            CompilationLevel.DYNAMO_AS_IS, CompilationLevel.DYNAMO_ONCE
    ]:
        if self.backend == "":
            return "eager"
        if self.backend in torch_backends:
            return self.backend
        return resolve_obj_by_qualname(self.backend)

    # TODO: pass user-specified backend to piecewise compilation
    # merge with the config use_inductor
    assert self.level == CompilationLevel.PIECEWISE

    from vllm.compilation.backends import VllmBackend
    return VllmBackend(vllm_config)

init_with_cudagraph_sizes

init_with_cudagraph_sizes(
    cudagraph_capture_sizes: list[int],
) -> None

To complete the initialization of config, we need to know the cudagraph sizes.

Source code in vllm/config.py

def init_with_cudagraph_sizes(self,
                              cudagraph_capture_sizes: list[int]) -> None:
    """To complete the initialization of config,
    we need to know the cudagraph sizes."""

    if self.cudagraph_capture_sizes is None:
        self.cudagraph_capture_sizes = cudagraph_capture_sizes
    else:
        # de-duplicate the sizes provided by the config
        dedup_sizes = list(set(self.cudagraph_capture_sizes))
        if len(dedup_sizes) < len(self.cudagraph_capture_sizes):
            logger.info(("cudagraph sizes specified by model runner"
                         " %s is overridden by config %s"),
                        cudagraph_capture_sizes, dedup_sizes)
        self.cudagraph_capture_sizes = dedup_sizes

    computed_compile_sizes = []
    if self.compile_sizes is not None:
        # de-duplicate the sizes provided by the config
        self.compile_sizes = list(set(self.compile_sizes))
        for x in self.compile_sizes:
            if isinstance(x, str):
                assert x == "cudagraph_capture_sizes", \
                "Unrecognized size type in compile_sizes, " \
                f"expect 'cudagraph_capture_sizes', got {x}"
                computed_compile_sizes.extend(self.cudagraph_capture_sizes)
            else:
                assert isinstance(x, int)
                computed_compile_sizes.append(x)
    self.compile_sizes = computed_compile_sizes  # type: ignore

    # sort to make sure cudagraph capture sizes are in descending order
    self.cudagraph_capture_sizes.sort(reverse=True)
    self.max_capture_size = self.cudagraph_capture_sizes[
        0] if self.cudagraph_capture_sizes else 0

    # pre-compute the mapping from batch size to padded graph size
    self.bs_to_padded_graph_size = [
        0 for i in range(self.max_capture_size + 1)
    ]
    for end, start in zip(self.cudagraph_capture_sizes,
                          self.cudagraph_capture_sizes[1:] + [0]):
        for bs in range(start, end):
            if bs == start:
                self.bs_to_padded_graph_size[bs] = start
            else:
                self.bs_to_padded_graph_size[bs] = end
    self.bs_to_padded_graph_size[
        self.max_capture_size] = self.max_capture_size

set_splitting_ops_for_v1

set_splitting_ops_for_v1()

Source code in vllm/config.py

def set_splitting_ops_for_v1(self):
    # NOTE: this function needs to be called
    if self.splitting_ops and self.full_cuda_graph:
        raise ValueError("full_cuda_graph cannot be used together with "
                         "splitting_ops, as Full CUDA graph will override "
                         f"the splitting_ops: {self.splitting_ops}")

    if not self.splitting_ops:
        self.splitting_ops = [] if self.full_cuda_graph else [
            "vllm.unified_attention",
            "vllm.unified_attention_with_output",
        ]

CompilationLevel

Source code in vllm/config.py

class CompilationLevel:
    # constants for the levels of the compilation process
    NO_COMPILATION = 0
    DYNAMO_AS_IS = 1
    DYNAMO_ONCE = 2
    PIECEWISE = 3

DYNAMO_AS_IS class-attribute instance-attribute

DYNAMO_AS_IS = 1

DYNAMO_ONCE class-attribute instance-attribute

DYNAMO_ONCE = 2

NO_COMPILATION class-attribute instance-attribute

NO_COMPILATION = 0

PIECEWISE class-attribute instance-attribute

PIECEWISE = 3

DecodingConfig

Dataclass which contains the decoding strategy of the engine.

Source code in vllm/config.py

@config
@dataclass
class DecodingConfig:
    """Dataclass which contains the decoding strategy of the engine."""

    @property
    @deprecated(
        "`guided_decoding_backend` is deprecated and has been renamed to "
        "`backend`. This will be removed in v0.10.0. Please use the "
        "`backend` argument instead.")
    def guided_decoding_backend(self) -> GuidedDecodingBackend:
        return self.backend

    @guided_decoding_backend.setter
    def guided_decoding_backend(self, value: GuidedDecodingBackend):
        self.backend = value

    backend: GuidedDecodingBackend = "auto" if envs.VLLM_USE_V1 else "xgrammar"
    """Which engine will be used for guided decoding (JSON schema / regex etc)
    by default. With "auto", we will make opinionated choices based on request
    contents and what the backend libraries currently support, so the behavior
    is subject to change in each release."""

    disable_fallback: bool = False
    """If `True`, vLLM will not fallback to a different backend on error."""

    disable_any_whitespace: bool = False
    """If `True`, the model will not generate any whitespace during guided
    decoding. This is only supported for xgrammar and guidance backends."""

    disable_additional_properties: bool = False
    """If `True`, the `guidance` backend will not use `additionalProperties`
    in the JSON schema. This is only supported for the `guidance` backend and
    is used to better align its behaviour with `outlines` and `xgrammar`."""

    reasoning_backend: str = ""
    """Select the reasoning parser depending on the model that you're using.
    This is used to parse the reasoning content into OpenAI API format."""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # this config will not affect the computation graph.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self):
        if ":" in self.backend:
            self._extract_backend_options()

        if envs.VLLM_USE_V1:
            valid_guided_backends = get_args(GuidedDecodingBackendV1)
        else:
            valid_guided_backends = get_args(GuidedDecodingBackendV0)
        if self.backend not in valid_guided_backends:
            raise ValueError(f"Invalid backend '{self.backend}',"
                             f" must be one of {valid_guided_backends}")
        if (self.disable_any_whitespace
                and self.backend not in ("xgrammar", "guidance")):
            raise ValueError("disable_any_whitespace is only supported for "
                             "xgrammar and guidance backends.")
        if (self.disable_additional_properties and self.backend != "guidance"):
            raise ValueError("disable_additional_properties is only supported "
                             "for the guidance backend.")

    @deprecated(
        "Passing guided decoding backend options inside backend in the format "
        "'backend:...' is deprecated. This will be removed in v0.10.0. Please "
        "use the dedicated arguments '--disable-fallback', "
        "'--disable-any-whitespace' and '--disable-additional-properties' "
        "instead.")
    def _extract_backend_options(self):
        """Extract backend options from the backend string."""
        backend, options = self.backend.split(":")
        self.backend = cast(GuidedDecodingBackend, backend)
        options_set = set(options.strip().split(","))
        if "no-fallback" in options_set:
            self.disable_fallback = True
        if "disable-any-whitespace" in options_set:
            self.disable_any_whitespace = True
        if "no-additional-properties" in options_set:
            self.disable_additional_properties = True

backend class-attribute instance-attribute

backend: GuidedDecodingBackend = (
    "auto" if VLLM_USE_V1 else "xgrammar"
)

Which engine will be used for guided decoding (JSON schema / regex etc) by default. With "auto", we will make opinionated choices based on request contents and what the backend libraries currently support, so the behavior is subject to change in each release.

disable_additional_properties class-attribute instance-attribute

disable_additional_properties: bool = False

If True, the guidance backend will not use additionalProperties in the JSON schema. This is only supported for the guidance backend and is used to better align its behaviour with outlines and xgrammar.

disable_any_whitespace class-attribute instance-attribute

disable_any_whitespace: bool = False

If True, the model will not generate any whitespace during guided decoding. This is only supported for xgrammar and guidance backends.

disable_fallback class-attribute instance-attribute

disable_fallback: bool = False

If True, vLLM will not fallback to a different backend on error.

guided_decoding_backend property writable

guided_decoding_backend: GuidedDecodingBackend

reasoning_backend class-attribute instance-attribute

reasoning_backend: str = ''

Select the reasoning parser depending on the model that you're using. This is used to parse the reasoning content into OpenAI API format.

__post_init__

__post_init__()

Source code in vllm/config.py

def __post_init__(self):
    if ":" in self.backend:
        self._extract_backend_options()

    if envs.VLLM_USE_V1:
        valid_guided_backends = get_args(GuidedDecodingBackendV1)
    else:
        valid_guided_backends = get_args(GuidedDecodingBackendV0)
    if self.backend not in valid_guided_backends:
        raise ValueError(f"Invalid backend '{self.backend}',"
                         f" must be one of {valid_guided_backends}")
    if (self.disable_any_whitespace
            and self.backend not in ("xgrammar", "guidance")):
        raise ValueError("disable_any_whitespace is only supported for "
                         "xgrammar and guidance backends.")
    if (self.disable_additional_properties and self.backend != "guidance"):
        raise ValueError("disable_additional_properties is only supported "
                         "for the guidance backend.")

_extract_backend_options

_extract_backend_options()

Extract backend options from the backend string.

Source code in vllm/config.py

@deprecated(
    "Passing guided decoding backend options inside backend in the format "
    "'backend:...' is deprecated. This will be removed in v0.10.0. Please "
    "use the dedicated arguments '--disable-fallback', "
    "'--disable-any-whitespace' and '--disable-additional-properties' "
    "instead.")
def _extract_backend_options(self):
    """Extract backend options from the backend string."""
    backend, options = self.backend.split(":")
    self.backend = cast(GuidedDecodingBackend, backend)
    options_set = set(options.strip().split(","))
    if "no-fallback" in options_set:
        self.disable_fallback = True
    if "disable-any-whitespace" in options_set:
        self.disable_any_whitespace = True
    if "no-additional-properties" in options_set:
        self.disable_additional_properties = True

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # this config will not affect the computation graph.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

DeviceConfig

Configuration for the device to use for vLLM execution.

Source code in vllm/config.py

@config
@dataclass(config=ConfigDict(arbitrary_types_allowed=True))
class DeviceConfig:
    """Configuration for the device to use for vLLM execution."""

    device: SkipValidation[Optional[Union[Device, torch.device]]] = "auto"
    """Device type for vLLM execution.
    This parameter is deprecated and will be
    removed in a future release.
    It will now be set automatically based
    on the current platform."""
    device_type: str = field(init=False)
    """Device type from the current platform. This is set in
    `__post_init__`."""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # the device/platform information will be summarized
        # by torch/vllm automatically.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self):
        if self.device == "auto":
            # Automated device type detection
            from vllm.platforms import current_platform
            self.device_type = current_platform.device_type
            if not self.device_type:
                raise RuntimeError(
                    "Failed to infer device type, please set "
                    "the environment variable `VLLM_LOGGING_LEVEL=DEBUG` "
                    "to turn on verbose logging to help debug the issue.")
        else:
            # Device type is assigned explicitly
            if isinstance(self.device, str):
                self.device_type = self.device
            elif isinstance(self.device, torch.device):
                self.device_type = self.device.type

        # Some device types require processing inputs on CPU
        if self.device_type in ["neuron"]:
            self.device = torch.device("cpu")
        elif self.device_type in ["tpu"]:
            self.device = None
        else:
            # Set device with device type
            self.device = torch.device(self.device_type)

device class-attribute instance-attribute

device: SkipValidation[Optional[Union[Device, device]]] = (
    "auto"
)

Device type for vLLM execution. This parameter is deprecated and will be removed in a future release. It will now be set automatically based on the current platform.

device_type class-attribute instance-attribute

device_type: str = field(init=False)

Device type from the current platform. This is set in __post_init__.

__post_init__

__post_init__()

Source code in vllm/config.py

def __post_init__(self):
    if self.device == "auto":
        # Automated device type detection
        from vllm.platforms import current_platform
        self.device_type = current_platform.device_type
        if not self.device_type:
            raise RuntimeError(
                "Failed to infer device type, please set "
                "the environment variable `VLLM_LOGGING_LEVEL=DEBUG` "
                "to turn on verbose logging to help debug the issue.")
    else:
        # Device type is assigned explicitly
        if isinstance(self.device, str):
            self.device_type = self.device
        elif isinstance(self.device, torch.device):
            self.device_type = self.device.type

    # Some device types require processing inputs on CPU
    if self.device_type in ["neuron"]:
        self.device = torch.device("cpu")
    elif self.device_type in ["tpu"]:
        self.device = None
    else:
        # Set device with device type
        self.device = torch.device(self.device_type)

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # the device/platform information will be summarized
    # by torch/vllm automatically.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

KVEventsConfig

Configuration for KV event publishing.

Source code in vllm/config.py

@config
@dataclass
class KVEventsConfig:
    """Configuration for KV event publishing."""

    enable_kv_cache_events: bool = False
    """If True, enable KV cache events for tracking block storage and removal.
    Events can be published externally by zmq using the event publisher config.
    """

    publisher: str = "null"
    """The publisher to use for publishing kv events. Can be "null", "zmq".
    """

    endpoint: str = "tcp://*:5557"
    """The zmq endpoint to use for publishing kv events.
    """

    replay_endpoint: Optional[str] = None
    """The zmq endpoint to use for replaying kv events.
    """

    buffer_steps: int = 10_000
    """The number of steps to cache for replay endpoint. Will only save
    events from the last N steps for the replay endpoint.
    """

    hwm: int = 100_000
    """The zmq high water mark for the event publisher. After queueing N events,
    events will start dropping if the consumer is not keeping up.
    """

    max_queue_size: int = 100_000
    """The maximum number of events to queue while waiting for publishing.
    """

    topic: str = ""
    """The topic to use for the event publisher. Consumers can subscribe to
    this topic to receive events.
    """

buffer_steps class-attribute instance-attribute

buffer_steps: int = 10000

The number of steps to cache for replay endpoint. Will only save events from the last N steps for the replay endpoint.

enable_kv_cache_events class-attribute instance-attribute

enable_kv_cache_events: bool = False

If True, enable KV cache events for tracking block storage and removal. Events can be published externally by zmq using the event publisher config.

endpoint class-attribute instance-attribute

endpoint: str = 'tcp://*:5557'

The zmq endpoint to use for publishing kv events.

hwm class-attribute instance-attribute

hwm: int = 100000

The zmq high water mark for the event publisher. After queueing N events, events will start dropping if the consumer is not keeping up.

max_queue_size class-attribute instance-attribute

max_queue_size: int = 100000

The maximum number of events to queue while waiting for publishing.

publisher class-attribute instance-attribute

publisher: str = 'null'

The publisher to use for publishing kv events. Can be "null", "zmq".

replay_endpoint class-attribute instance-attribute

replay_endpoint: Optional[str] = None

The zmq endpoint to use for replaying kv events.

topic class-attribute instance-attribute

topic: str = ''

The topic to use for the event publisher. Consumers can subscribe to this topic to receive events.

KVTransferConfig

Configuration for distributed KV cache transfer.

Source code in vllm/config.py

@config
@dataclass
class KVTransferConfig:
    """Configuration for distributed KV cache transfer."""

    kv_connector: Optional[str] = None
    """The KV connector for vLLM to transmit KV caches between vLLM instances.
    """

    engine_id: Optional[str] = None
    """The engine id for KV transfers."""

    kv_buffer_device: Optional[str] = "cuda"
    """The device used by kv connector to buffer the KV cache.
    Currently only support 'cuda'."""

    kv_buffer_size: float = 1e9
    """The buffer size for TorchDistributedConnector. Measured in number of
    bytes. Recommended value: 1e9 (about 1GB)."""

    kv_role: Optional[KVRole] = None
    """Whether this vLLM instance produces, consumes KV cache, or both. Choices
    are 'kv_producer', 'kv_consumer', and 'kv_both'."""

    kv_rank: Optional[int] = None
    """The rank of this vLLM instance in the KV cache transfer. Typical value:
    0 for prefill instance, 1 for decode instance.
    Currently only 1P1D is supported."""

    kv_parallel_size: int = 1
    """The number of parallel instances for KV cache transfer. For
    PyNcclConnector, this should be 2."""

    kv_ip: str = "127.0.0.1"
    """The KV connector ip, used to build distributed connection."""

    kv_port: int = 14579
    """The KV connector port, used to build distributed connection."""

    kv_connector_extra_config: dict[str, Any] = field(default_factory=dict)
    """any extra config that the connector may need."""

    kv_connector_module_path: Optional[str] = None
    """The Python module path to dynamically load the KV connector from.
    Only supported in V1."""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # this config will not affect the computation graph.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self) -> None:
        if self.engine_id is None:
            self.engine_id = str(uuid.uuid4())

        if self.kv_role is not None and self.kv_role not in get_args(KVRole):
            raise ValueError(f"Unsupported kv_role: {self.kv_role}. "
                             f"Supported roles are {get_args(KVRole)}")

        if self.kv_connector is not None and self.kv_role is None:
            raise ValueError("Please specify kv_disagg_role when kv_connector "
                             f"is set, supported roles are {get_args(KVRole)}")

    @property
    def is_kv_transfer_instance(self) -> bool:
        return self.kv_connector is not None and \
            self.kv_role in get_args(KVRole)

    @property
    def is_kv_producer(self) -> bool:
        return self.kv_connector is not None and \
            self.kv_role in get_args(KVProducer)

    @property
    def is_kv_consumer(self) -> bool:
        return self.kv_connector is not None and \
            self.kv_role in get_args(KVConsumer)

    def get_from_extra_config(self, key, default) -> Any:
        return self.kv_connector_extra_config.get(key, default)

engine_id class-attribute instance-attribute

engine_id: Optional[str] = None

The engine id for KV transfers.

is_kv_consumer property

is_kv_consumer: bool

is_kv_producer property

is_kv_producer: bool

is_kv_transfer_instance property

is_kv_transfer_instance: bool

kv_buffer_device class-attribute instance-attribute

kv_buffer_device: Optional[str] = 'cuda'

The device used by kv connector to buffer the KV cache. Currently only support 'cuda'.

kv_buffer_size class-attribute instance-attribute

kv_buffer_size: float = 1000000000.0

The buffer size for TorchDistributedConnector. Measured in number of bytes. Recommended value: 1e9 (about 1GB).

kv_connector class-attribute instance-attribute

kv_connector: Optional[str] = None

The KV connector for vLLM to transmit KV caches between vLLM instances.

kv_connector_extra_config class-attribute instance-attribute

kv_connector_extra_config: dict[str, Any] = field(
    default_factory=dict
)

any extra config that the connector may need.

kv_connector_module_path class-attribute instance-attribute

kv_connector_module_path: Optional[str] = None

The Python module path to dynamically load the KV connector from. Only supported in V1.

kv_ip class-attribute instance-attribute

kv_ip: str = '127.0.0.1'

The KV connector ip, used to build distributed connection.

kv_parallel_size class-attribute instance-attribute

kv_parallel_size: int = 1

The number of parallel instances for KV cache transfer. For PyNcclConnector, this should be 2.

kv_port class-attribute instance-attribute

kv_port: int = 14579

The KV connector port, used to build distributed connection.

kv_rank class-attribute instance-attribute

kv_rank: Optional[int] = None

The rank of this vLLM instance in the KV cache transfer. Typical value: 0 for prefill instance, 1 for decode instance. Currently only 1P1D is supported.

kv_role class-attribute instance-attribute

kv_role: Optional[KVRole] = None

Whether this vLLM instance produces, consumes KV cache, or both. Choices are 'kv_producer', 'kv_consumer', and 'kv_both'.

__post_init__

__post_init__() -> None

Source code in vllm/config.py

def __post_init__(self) -> None:
    if self.engine_id is None:
        self.engine_id = str(uuid.uuid4())

    if self.kv_role is not None and self.kv_role not in get_args(KVRole):
        raise ValueError(f"Unsupported kv_role: {self.kv_role}. "
                         f"Supported roles are {get_args(KVRole)}")

    if self.kv_connector is not None and self.kv_role is None:
        raise ValueError("Please specify kv_disagg_role when kv_connector "
                         f"is set, supported roles are {get_args(KVRole)}")

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # this config will not affect the computation graph.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

get_from_extra_config

get_from_extra_config(key, default) -> Any

Source code in vllm/config.py

def get_from_extra_config(self, key, default) -> Any:
    return self.kv_connector_extra_config.get(key, default)

LoRAConfig

Configuration for LoRA.

Source code in vllm/config.py

@config
@dataclass(config=ConfigDict(arbitrary_types_allowed=True))
class LoRAConfig:
    """Configuration for LoRA."""

    max_lora_rank: int = 16
    """Max LoRA rank."""
    max_loras: int = 1
    """Max number of LoRAs in a single batch."""
    fully_sharded_loras: bool = False
    """By default, only half of the LoRA computation is sharded with tensor
    parallelism. Enabling this will use the fully sharded layers. At high
    sequence length, max rank or tensor parallel size, this is likely faster.
    """
    max_cpu_loras: Optional[int] = None
    """Maximum number of LoRAs to store in CPU memory. Must be >= than
    `max_loras`."""
    lora_dtype: Union[torch.dtype, LoRADType] = "auto"
    """Data type for LoRA. If auto, will default to base model dtype."""
    lora_extra_vocab_size: int = 256
    """Maximum size of extra vocabulary that can be present in a LoRA adapter
    (added to the base model vocabulary)."""
    lora_vocab_padding_size: ClassVar[int] = current_platform\
        .get_lora_vocab_padding_size()
    long_lora_scaling_factors: Optional[tuple[float, ...]] = None
    """Specify multiple scaling factors (which can be different from base model
    scaling factor - see eg. Long LoRA) to allow for multiple LoRA adapters
    trained with those scaling factors to be used at the same time. If not
    specified, only adapters trained with the base model scaling factor are
    allowed."""
    bias_enabled: bool = False
    """Enable bias for LoRA adapters."""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        factors: list[Any] = []
        factors.append(self.max_lora_rank)
        factors.append(self.max_loras)
        factors.append(self.fully_sharded_loras)
        factors.append(self.lora_dtype)
        factors.append(self.lora_extra_vocab_size)
        factors.append(self.lora_vocab_padding_size)
        factors.append(self.long_lora_scaling_factors)
        factors.append(self.bias_enabled)
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self):
        # Setting the maximum rank to 512 should be able to satisfy the vast
        # majority of applications.
        possible_max_ranks = (8, 16, 32, 64, 128, 256, 320, 512)
        possible_lora_extra_vocab_size = (256, 512)
        if self.max_lora_rank not in possible_max_ranks:
            raise ValueError(
                f"max_lora_rank ({self.max_lora_rank}) must be one of "
                f"{possible_max_ranks}.")
        if self.lora_extra_vocab_size not in possible_lora_extra_vocab_size:
            raise ValueError(
                f"lora_extra_vocab_size ({self.lora_extra_vocab_size}) "
                f"must be one of {possible_lora_extra_vocab_size}.")
        if self.max_loras < 1:
            raise ValueError(f"max_loras ({self.max_loras}) must be >= 1.")
        if self.max_cpu_loras is None:
            self.max_cpu_loras = self.max_loras
        elif self.max_cpu_loras < self.max_loras:
            raise ValueError(
                f"max_cpu_loras ({self.max_cpu_loras}) must be >= "
                f"max_loras ({self.max_loras})")

    def verify_with_cache_config(self, cache_config: CacheConfig):
        if cache_config.cpu_offload_gb > 0 and not envs.VLLM_USE_V1:
            raise ValueError(
                "V0 LoRA does not support CPU offload, please use V1.")

    def verify_with_model_config(self, model_config: ModelConfig):
        if self.lora_dtype in (None, "auto"):
            self.lora_dtype = model_config.dtype
        elif isinstance(self.lora_dtype, str):
            self.lora_dtype = getattr(torch, self.lora_dtype)

    def verify_lora_support(self):
        if self.long_lora_scaling_factors is not None and envs.VLLM_USE_V1:
            raise ValueError(
                "V1 LoRA does not support long LoRA, please use V0.")

bias_enabled class-attribute instance-attribute

bias_enabled: bool = False

Enable bias for LoRA adapters.

fully_sharded_loras class-attribute instance-attribute

fully_sharded_loras: bool = False

By default, only half of the LoRA computation is sharded with tensor parallelism. Enabling this will use the fully sharded layers. At high sequence length, max rank or tensor parallel size, this is likely faster.

long_lora_scaling_factors class-attribute instance-attribute

long_lora_scaling_factors: Optional[tuple[float, ...]] = (
    None
)

Specify multiple scaling factors (which can be different from base model scaling factor - see eg. Long LoRA) to allow for multiple LoRA adapters trained with those scaling factors to be used at the same time. If not specified, only adapters trained with the base model scaling factor are allowed.

lora_dtype class-attribute instance-attribute

lora_dtype: Union[dtype, LoRADType] = 'auto'

Data type for LoRA. If auto, will default to base model dtype.

lora_extra_vocab_size class-attribute instance-attribute

lora_extra_vocab_size: int = 256

Maximum size of extra vocabulary that can be present in a LoRA adapter (added to the base model vocabulary).

lora_vocab_padding_size class-attribute

lora_vocab_padding_size: int = get_lora_vocab_padding_size()

max_cpu_loras class-attribute instance-attribute

max_cpu_loras: Optional[int] = None

Maximum number of LoRAs to store in CPU memory. Must be >= than max_loras.

max_lora_rank class-attribute instance-attribute

max_lora_rank: int = 16

Max LoRA rank.

max_loras class-attribute instance-attribute

max_loras: int = 1

Max number of LoRAs in a single batch.

__post_init__

__post_init__()

Source code in vllm/config.py

def __post_init__(self):
    # Setting the maximum rank to 512 should be able to satisfy the vast
    # majority of applications.
    possible_max_ranks = (8, 16, 32, 64, 128, 256, 320, 512)
    possible_lora_extra_vocab_size = (256, 512)
    if self.max_lora_rank not in possible_max_ranks:
        raise ValueError(
            f"max_lora_rank ({self.max_lora_rank}) must be one of "
            f"{possible_max_ranks}.")
    if self.lora_extra_vocab_size not in possible_lora_extra_vocab_size:
        raise ValueError(
            f"lora_extra_vocab_size ({self.lora_extra_vocab_size}) "
            f"must be one of {possible_lora_extra_vocab_size}.")
    if self.max_loras < 1:
        raise ValueError(f"max_loras ({self.max_loras}) must be >= 1.")
    if self.max_cpu_loras is None:
        self.max_cpu_loras = self.max_loras
    elif self.max_cpu_loras < self.max_loras:
        raise ValueError(
            f"max_cpu_loras ({self.max_cpu_loras}) must be >= "
            f"max_loras ({self.max_loras})")

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    factors: list[Any] = []
    factors.append(self.max_lora_rank)
    factors.append(self.max_loras)
    factors.append(self.fully_sharded_loras)
    factors.append(self.lora_dtype)
    factors.append(self.lora_extra_vocab_size)
    factors.append(self.lora_vocab_padding_size)
    factors.append(self.long_lora_scaling_factors)
    factors.append(self.bias_enabled)
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

verify_lora_support

verify_lora_support()

Source code in vllm/config.py

def verify_lora_support(self):
    if self.long_lora_scaling_factors is not None and envs.VLLM_USE_V1:
        raise ValueError(
            "V1 LoRA does not support long LoRA, please use V0.")

verify_with_cache_config

verify_with_cache_config(cache_config: CacheConfig)

Source code in vllm/config.py

def verify_with_cache_config(self, cache_config: CacheConfig):
    if cache_config.cpu_offload_gb > 0 and not envs.VLLM_USE_V1:
        raise ValueError(
            "V0 LoRA does not support CPU offload, please use V1.")

verify_with_model_config

verify_with_model_config(model_config: ModelConfig)

Source code in vllm/config.py

def verify_with_model_config(self, model_config: ModelConfig):
    if self.lora_dtype in (None, "auto"):
        self.lora_dtype = model_config.dtype
    elif isinstance(self.lora_dtype, str):
        self.lora_dtype = getattr(torch, self.lora_dtype)

LoadConfig

Configuration for loading the model weights.

Source code in vllm/config.py

@config
@dataclass
class LoadConfig:
    """Configuration for loading the model weights."""

    load_format: Union[str, LoadFormat,
                       "BaseModelLoader"] = LoadFormat.AUTO.value
    """The format of the model weights to load:\n
    - "auto" will try to load the weights in the safetensors format and fall
    back to the pytorch bin format if safetensors format is not available.\n
    - "pt" will load the weights in the pytorch bin format.\n
    - "safetensors" will load the weights in the safetensors format.\n
    - "npcache" will load the weights in pytorch format and store a numpy cache
    to speed up the loading.\n
    - "dummy" will initialize the weights with random values, which is mainly
    for profiling.\n
    - "tensorizer" will use CoreWeave's tensorizer library for fast weight
    loading. See the Tensorize vLLM Model script in the Examples section for
    more information.\n
    - "runai_streamer" will load the Safetensors weights using Run:ai Model
    Streamer.\n
    - "bitsandbytes" will load the weights using bitsandbytes quantization.\n
    - "sharded_state" will load weights from pre-sharded checkpoint files,
    supporting efficient loading of tensor-parallel models.\n
    - "gguf" will load weights from GGUF format files (details specified in
    https://github.com/ggml-org/ggml/blob/master/docs/gguf.md).\n
    - "mistral" will load weights from consolidated safetensors files used by
    Mistral models."""
    download_dir: Optional[str] = None
    """Directory to download and load the weights, default to the default
    cache directory of Hugging Face."""
    model_loader_extra_config: Union[dict, TensorizerConfig] = field(
        default_factory=dict)
    """Extra config for model loader. This will be passed to the model loader
    corresponding to the chosen load_format."""
    ignore_patterns: Optional[Union[list[str], str]] = None
    """The list of patterns to ignore when loading the model. Default to
    "original/**/*" to avoid repeated loading of llama's checkpoints."""
    use_tqdm_on_load: bool = True
    """Whether to enable tqdm for showing progress bar when loading model
    weights."""
    pt_load_map_location: Union[str, dict[str, str]] = "cpu"
    """
    pt_load_map_location: the map location for loading pytorch checkpoint, to
    support loading checkpoints can only be loaded on certain devices like
    "cuda", this is equivalent to {"": "cuda"}. Another supported format is
    mapping from different devices like from GPU 1 to GPU 0:
    {"cuda:1": "cuda:0"}. Note that when passed from command line, the strings
    in dictionary needs to be double quoted for json parsing. For more details,
    see original doc for `map_location` in https://pytorch.org/docs/stable/generated/torch.load.html
    """

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # this config will not affect the computation graph.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self):
        if isinstance(self.load_format, str):
            load_format = self.load_format.lower()
            self.load_format = LoadFormat(load_format)

        if self.ignore_patterns is not None and len(self.ignore_patterns) > 0:
            logger.info(
                "Ignoring the following patterns when downloading weights: %s",
                self.ignore_patterns)
        else:
            self.ignore_patterns = ["original/**/*"]

download_dir class-attribute instance-attribute

download_dir: Optional[str] = None

Directory to download and load the weights, default to the default cache directory of Hugging Face.

ignore_patterns class-attribute instance-attribute

ignore_patterns: Optional[Union[list[str], str]] = None

The list of patterns to ignore when loading the model. Default to "original/*/" to avoid repeated loading of llama's checkpoints.

load_format class-attribute instance-attribute

load_format: Union[str, LoadFormat, BaseModelLoader] = (
    AUTO.value
)

The format of the model weights to load:

• "auto" will try to load the weights in the safetensors format and fall back to the pytorch bin format if safetensors format is not available.

• "pt" will load the weights in the pytorch bin format.

• "safetensors" will load the weights in the safetensors format.

• "npcache" will load the weights in pytorch format and store a numpy cache to speed up the loading.

• "dummy" will initialize the weights with random values, which is mainly for profiling.

• "tensorizer" will use CoreWeave's tensorizer library for fast weight loading. See the Tensorize vLLM Model script in the Examples section for more information.

• "runai_streamer" will load the Safetensors weights using Run:ai Model Streamer.

• "bitsandbytes" will load the weights using bitsandbytes quantization.

• "sharded_state" will load weights from pre-sharded checkpoint files, supporting efficient loading of tensor-parallel models.

• "gguf" will load weights from GGUF format files (details specified in https://github.com/ggml-org/ggml/blob/master/docs/gguf.md).

• "mistral" will load weights from consolidated safetensors files used by Mistral models.

model_loader_extra_config class-attribute instance-attribute

model_loader_extra_config: Union[dict, TensorizerConfig] = (
    field(default_factory=dict)
)

Extra config for model loader. This will be passed to the model loader corresponding to the chosen load_format.

pt_load_map_location class-attribute instance-attribute

pt_load_map_location: Union[str, dict[str, str]] = 'cpu'

pt_load_map_location: the map location for loading pytorch checkpoint, to support loading checkpoints can only be loaded on certain devices like "cuda", this is equivalent to {"": "cuda"}. Another supported format is mapping from different devices like from GPU 1 to GPU 0: {"cuda:1": "cuda:0"}. Note that when passed from command line, the strings in dictionary needs to be double quoted for json parsing. For more details, see original doc for map_location in https://pytorch.org/docs/stable/generated/torch.load.html

use_tqdm_on_load class-attribute instance-attribute

use_tqdm_on_load: bool = True

Whether to enable tqdm for showing progress bar when loading model weights.

__post_init__

__post_init__()

Source code in vllm/config.py

def __post_init__(self):
    if isinstance(self.load_format, str):
        load_format = self.load_format.lower()
        self.load_format = LoadFormat(load_format)

    if self.ignore_patterns is not None and len(self.ignore_patterns) > 0:
        logger.info(
            "Ignoring the following patterns when downloading weights: %s",
            self.ignore_patterns)
    else:
        self.ignore_patterns = ["original/**/*"]

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # this config will not affect the computation graph.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

LoadFormat

Bases: str, Enum

Source code in vllm/config.py

class LoadFormat(str, enum.Enum):
    AUTO = "auto"
    PT = "pt"
    SAFETENSORS = "safetensors"
    NPCACHE = "npcache"
    DUMMY = "dummy"
    TENSORIZER = "tensorizer"
    SHARDED_STATE = "sharded_state"
    GGUF = "gguf"
    BITSANDBYTES = "bitsandbytes"
    MISTRAL = "mistral"
    RUNAI_STREAMER = "runai_streamer"
    RUNAI_STREAMER_SHARDED = "runai_streamer_sharded"
    FASTSAFETENSORS = "fastsafetensors"

AUTO class-attribute instance-attribute

AUTO = 'auto'

BITSANDBYTES class-attribute instance-attribute

BITSANDBYTES = 'bitsandbytes'

DUMMY class-attribute instance-attribute

DUMMY = 'dummy'

FASTSAFETENSORS class-attribute instance-attribute

FASTSAFETENSORS = 'fastsafetensors'

GGUF class-attribute instance-attribute

GGUF = 'gguf'

MISTRAL class-attribute instance-attribute

MISTRAL = 'mistral'

NPCACHE class-attribute instance-attribute

NPCACHE = 'npcache'

PT class-attribute instance-attribute

PT = 'pt'

RUNAI_STREAMER class-attribute instance-attribute

RUNAI_STREAMER = 'runai_streamer'

RUNAI_STREAMER_SHARDED class-attribute instance-attribute

RUNAI_STREAMER_SHARDED = 'runai_streamer_sharded'

SAFETENSORS class-attribute instance-attribute

SAFETENSORS = 'safetensors'

SHARDED_STATE class-attribute instance-attribute

SHARDED_STATE = 'sharded_state'

TENSORIZER class-attribute instance-attribute

TENSORIZER = 'tensorizer'

ModelConfig

Configuration for the model.

Source code in vllm/config.py

@config
@dataclass(config=ConfigDict(arbitrary_types_allowed=True))
class ModelConfig:
    """Configuration for the model."""

    model: str = "facebook/opt-125m"
    """Name or path of the Hugging Face model to use. It is also used as the
    content for `model_name` tag in metrics output when `served_model_name` is
    not specified."""
    task: Literal[TaskOption, Literal["draft"]] = "auto"
    """The task to use the model for. Each vLLM instance only supports one
    task, even if the same model can be used for multiple tasks. When the model
    only supports one task, "auto" can be used to select it; otherwise, you
    must specify explicitly which task to use."""
    tokenizer: SkipValidation[str] = None  # type: ignore
    """Name or path of the Hugging Face tokenizer to use. If unspecified, model
    name or path will be used."""
    tokenizer_mode: TokenizerMode = "auto"
    """Tokenizer mode:\n
    - "auto" will use the fast tokenizer if available.\n
    - "slow" will always use the slow tokenizer.\n
    - "mistral" will always use the tokenizer from `mistral_common`.\n
    - "custom" will use --tokenizer to select the preregistered tokenizer."""
    trust_remote_code: bool = False
    """Trust remote code (e.g., from HuggingFace) when downloading the model
    and tokenizer."""
    dtype: Union[ModelDType, torch.dtype] = "auto"
    """Data type for model weights and activations:\n
    - "auto" will use FP16 precision for FP32 and FP16 models, and BF16
    precision for BF16 models.\n
    - "half" for FP16. Recommended for AWQ quantization.\n
    - "float16" is the same as "half".\n
    - "bfloat16" for a balance between precision and range.\n
    - "float" is shorthand for FP32 precision.\n
    - "float32" for FP32 precision."""
    seed: Optional[int] = None
    """Random seed for reproducibility. Initialized to None in V0, but
    initialized to 0 in V1."""
    hf_config_path: Optional[str] = None
    """Name or path of the Hugging Face config to use. If unspecified, model
    name or path will be used."""
    allowed_local_media_path: str = ""
    """Allowing API requests to read local images or videos from directories
    specified by the server file system. This is a security risk. Should only
    be enabled in trusted environments."""
    revision: Optional[str] = None
    """The specific model version to use. It can be a branch name, a tag name,
    or a commit id. If unspecified, will use the default version."""
    code_revision: Optional[str] = None
    """The specific revision to use for the model code on the Hugging Face Hub.
    It can be a branch name, a tag name, or a commit id. If unspecified, will
    use the default version."""
    rope_scaling: dict[str, Any] = field(default_factory=dict)
    """RoPE scaling configuration. For example,
    `{"rope_type":"dynamic","factor":2.0}`."""
    rope_theta: Optional[float] = None
    """RoPE theta. Use with `rope_scaling`. In some cases, changing the RoPE
    theta improves the performance of the scaled model."""
    tokenizer_revision: Optional[str] = None
    """The specific revision to use for the tokenizer on the Hugging Face Hub.
    It can be a branch name, a tag name, or a commit id. If unspecified, will
    use the default version."""
    max_model_len: SkipValidation[int] = None  # type: ignore
    """Model context length (prompt and output). If unspecified, will be
    automatically derived from the model config.

    When passing via `--max-model-len`, supports k/m/g/K/M/G in human-readable
    format. Examples:\n
    - 1k -> 1000\n
    - 1K -> 1024\n
    - 25.6k -> 25,600"""
    spec_target_max_model_len: Optional[int] = None
    """Specify the maximum length for spec decoding draft models."""
    quantization: SkipValidation[Optional[QuantizationMethods]] = None
    """Method used to quantize the weights. If `None`, we first check the
    `quantization_config` attribute in the model config file. If that is
    `None`, we assume the model weights are not quantized and use `dtype` to
    determine the data type of the weights."""
    enforce_eager: bool = False
    """Whether to always use eager-mode PyTorch. If True, we will disable CUDA
    graph and always execute the model in eager mode. If False, we will use
    CUDA graph and eager execution in hybrid for maximal performance and
    flexibility."""
    max_seq_len_to_capture: int = 8192
    """Maximum sequence len covered by CUDA graphs. When a sequence has context
    length larger than this, we fall back to eager mode. Additionally for
    encoder-decoder models, if the sequence length of the encoder input is
    larger than this, we fall back to the eager mode."""
    max_logprobs: int = 20
    """Maximum number of log probabilities to return when `logprobs` is
    specified in `SamplingParams`. The default value comes the default for the
    OpenAI Chat Completions API."""
    disable_sliding_window: bool = False
    """Whether to disable sliding window. If True, we will disable the sliding
    window functionality of the model, capping to sliding window size. If the
    model does not support sliding window, this argument is ignored."""
    disable_cascade_attn: bool = False
    """Disable cascade attention for V1. While cascade attention does not
    change the mathematical correctness, disabling it could be useful for
    preventing potential numerical issues. Note that even if this is set to
    False, cascade attention will be only used when the heuristic tells that
    it's beneficial."""
    skip_tokenizer_init: bool = False
    """Skip initialization of tokenizer and detokenizer. Expects valid
    `prompt_token_ids` and `None` for prompt from the input. The generated
    output will contain token ids."""
    enable_prompt_embeds: bool = False
    """If `True`, enables passing text embeddings as inputs via the
    `prompt_embeds` key. Note that enabling this will double the time required
    for graph compilation."""
    served_model_name: Optional[Union[str, list[str]]] = None
    """The model name(s) used in the API. If multiple names are provided, the
    server will respond to any of the provided names. The model name in the
    model field of a response will be the first name in this list. If not
    specified, the model name will be the same as the `--model` argument. Noted
    that this name(s) will also be used in `model_name` tag content of
    prometheus metrics, if multiple names provided, metrics tag will take the
    first one."""
    limit_mm_per_prompt: dict[str, int] = field(default_factory=dict)
    """Maximum number of data items per modality per prompt. Only applicable
    for multimodal models."""
    use_async_output_proc: bool = True
    """Whether to use async output processor."""
    config_format: Union[str, ConfigFormat] = ConfigFormat.AUTO.value
    """The format of the model config to load:\n
    - "auto" will try to load the config in hf format if available else it
    will try to load in mistral format.\n
    - "hf" will load the config in hf format.\n
    - "mistral" will load the config in mistral format."""
    hf_token: Optional[Union[bool, str]] = None
    """The token to use as HTTP bearer authorization for remote files . If
    `True`, will use the token generated when running `huggingface-cli login`
    (stored in `~/.huggingface`)."""
    hf_overrides: HfOverrides = field(default_factory=dict)
    """If a dictionary, contains arguments to be forwarded to the Hugging Face
    config. If a callable, it is called to update the HuggingFace config."""
    mm_processor_kwargs: Optional[dict[str, Any]] = None
    """Arguments to be forwarded to the model's processor for multi-modal data,
    e.g., image processor. Overrides for the multi-modal processor obtained
    from `AutoProcessor.from_pretrained`. The available overrides depend on the
    model that is being run. For example, for Phi-3-Vision: `{"num_crops": 4}`.
    """
    disable_mm_preprocessor_cache: bool = False
    """If `True`, disable caching of the multi-modal preprocessor/mapper (not
    recommended)."""
    override_neuron_config: dict[str, Any] = field(default_factory=dict)
    """Initialize non-default neuron config or override default neuron config
    that are specific to Neuron devices, this argument will be used to
    configure the neuron config that can not be gathered from the vllm
    arguments. e.g. `{"cast_logits_dtype": "bfloat16"}`."""
    pooler_config: Optional["PoolerConfig"] = field(init=False)
    """Pooler config which controls the behaviour of output pooling in pooling
    models."""
    override_pooler_config: Optional[Union[dict, "PoolerConfig"]] = None
    """Initialize non-default pooling config or override default pooling config
    for the pooling model. e.g. `{"pooling_type": "mean", "normalize": false}`.
    """
    logits_processor_pattern: Optional[str] = None
    """Optional regex pattern specifying valid logits processor qualified names
    that can be passed with the `logits_processors` extra completion argument.
    Defaults to `None`, which allows no processors."""
    generation_config: str = "auto"
    """The folder path to the generation config. Defaults to `"auto"`, the
    generation config will be loaded from model path. If set to `"vllm"`, no
    generation config is loaded, vLLM defaults will be used. If set to a folder
    path, the generation config will be loaded from the specified folder path.
    If `max_new_tokens` is specified in generation config, then it sets a
    server-wide limit on the number of output tokens for all requests."""
    override_generation_config: dict[str, Any] = field(default_factory=dict)
    """Overrides or sets generation config. e.g. `{"temperature": 0.5}`. If
    used with `--generation-config auto`, the override parameters will be
    merged with the default config from the model. If used with
    `--generation-config vllm`, only the override parameters are used."""
    enable_sleep_mode: bool = False
    """Enable sleep mode for the engine (only cuda platform is supported)."""
    model_impl: Union[str, ModelImpl] = ModelImpl.AUTO.value
    """Which implementation of the model to use:\n
    - "auto" will try to use the vLLM implementation, if it exists, and fall
    back to the Transformers implementation if no vLLM implementation is
    available.\n
    - "vllm" will use the vLLM model implementation.\n
    - "transformers" will use the Transformers model implementation."""
    override_attention_dtype: Optional[str] = None
    """Override dtype for attention"""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        factors: list[Any] = []
        factors.append(self.model)
        factors.append(self.dtype)
        factors.append(self.quantization)
        factors.append(self.revision)
        factors.append(self.code_revision)
        factors.append(self.max_model_len)
        factors.append(self.max_logprobs)
        factors.append(self.disable_sliding_window)
        factors.append(self.trust_remote_code)
        factors.append(self.generation_config)
        factors.append(self.model_impl)
        factors.append(self.override_generation_config)
        factors.append(self.rope_scaling)
        factors.append(self.rope_theta)
        # hf_config can control how the model looks!
        factors.append(self.hf_config.to_json_string())
        str_factors = str(factors)
        assert_hashable(str_factors)
        return hashlib.sha256(str(factors).encode()).hexdigest()

    def __post_init__(self) -> None:
        # Set the default seed to 0 in V1.
        # NOTE(woosuk): In V0, we set the default seed to None because the
        # driver worker shares the same process as the user process, and thus
        # setting a seed affects the user process as well.
        # In V1, we use separate processes for workers (unless
        # VLLM_ENABLE_V1_MULTIPROCESSING=0), so setting a seed here
        # doesn't affect the user process. However, without a consistent seed,
        # different tensor parallel workers would sample different tokens,
        # leading to inconsistent results.
        if envs.VLLM_USE_V1 and self.seed is None:
            self.seed = 0
            if not envs.VLLM_ENABLE_V1_MULTIPROCESSING:
                logger.warning(
                    "The global random seed is set to %d. Since "
                    "VLLM_ENABLE_V1_MULTIPROCESSING is set to False, this may "
                    "affect the random state of the Python process that "
                    "launched vLLM.", self.seed)

        self.model = maybe_model_redirect(self.model)
        # The tokenizer is consistent with the model by default.
        if self.tokenizer is None:
            self.tokenizer = self.model
        if self.tokenizer_revision is None:
            self.tokenizer_revision = self.revision
        self.tokenizer = maybe_model_redirect(self.tokenizer)

        if isinstance(self.hf_config_path, str):
            self.hf_config_path = maybe_model_redirect(self.hf_config_path)

        if callable(self.hf_overrides):
            hf_overrides_kw = {}
            hf_overrides_fn = self.hf_overrides
        else:
            hf_overrides_kw = self.hf_overrides
            hf_overrides_fn = None

        if self.rope_scaling:
            hf_override: dict[str, Any] = {"rope_scaling": self.rope_scaling}
            hf_overrides_kw.update(hf_override)
            hf_overrides_str = json.dumps(hf_overrides_kw)
            msg = (
                "`--rope-scaling` will be removed in a future release. "
                f"'Please instead use `--hf-overrides '{hf_overrides_str}'`")
            warnings.warn(DeprecationWarning(msg), stacklevel=2)
        if self.rope_theta is not None:
            hf_override = {"rope_theta": self.rope_theta}
            hf_overrides_kw.update(hf_override)
            hf_overrides_str = json.dumps(hf_overrides_kw)
            msg = (
                "`--rope-theta` will be removed in a future release. "
                f"'Please instead use `--hf-overrides '{hf_overrides_str}'`")
            warnings.warn(DeprecationWarning(msg), stacklevel=2)

        self.maybe_pull_model_tokenizer_for_s3(self.model, self.tokenizer)

        if (backend := envs.VLLM_ATTENTION_BACKEND
            ) and backend == "FLASHINFER" and find_spec("flashinfer") is None:
            raise ValueError(
                "VLLM_ATTENTION_BACKEND is set to FLASHINFER, but flashinfer "
                "module was not found. See "
                "https://github.com/vllm-project/vllm/blob/main/docker/Dockerfile "  # noqa: E501
                "for instructions on how to install it.")

        from vllm.platforms import current_platform

        if (self.override_attention_dtype is not None
                and not current_platform.is_rocm()):
            warnings.warn(
                "override-attention-dtype is set but not using ROCm platform",
                stacklevel=2)

        if (self.enable_sleep_mode
                and not current_platform.is_sleep_mode_available()):
            raise ValueError(
                "Sleep mode is not supported on current platform.")

        if isinstance(self.config_format, str):
            self.config_format = ConfigFormat(self.config_format)

        hf_config = get_config(self.hf_config_path or self.model,
                               self.trust_remote_code, self.revision,
                               self.code_revision, self.config_format)

        if hf_overrides_kw:
            logger.info("Overriding HF config with %s", hf_overrides_kw)
            hf_config.update(hf_overrides_kw)
        if hf_overrides_fn:
            logger.info("Overriding HF config with %s", hf_overrides_fn)
            hf_config = hf_overrides_fn(hf_config)

        self.hf_config = hf_config

        self.hf_text_config = get_hf_text_config(self.hf_config)
        self.attention_chunk_size = getattr(self.hf_text_config,
                                            "attention_chunk_size", None)
        self.encoder_config = self._get_encoder_config()
        self.hf_image_processor_config = get_hf_image_processor_config(
            self.model, hf_token=self.hf_token, revision=self.revision)

        supported_tasks, task = self._resolve_task(self.task)
        self.supported_tasks = supported_tasks
        self.task = task
        if self.task in ("draft", "generate"):
            self.truncation_side = "left"
        else:
            self.truncation_side = "right"

        self.pooler_config = self._init_pooler_config()

        self.dtype = _get_and_verify_dtype(
            self.model,
            self.hf_config,
            self.dtype,
            is_pooling_model=self.runner_type == "pooling",
            revision=self.revision,
        )

        # Workaround for Gemma 2 which uses interleaved sliding window
        # attention, but it's not specified in its config. TODO: remove this
        # when Gemma 2 is fixed in Transformers.
        if self.hf_text_config.model_type == "gemma2":
            self.hf_text_config.sliding_window_pattern = 2

        sliding_window = getattr(self.hf_text_config, "sliding_window", None)
        sliding_window_pattern = getattr(self.hf_text_config,
                                         "sliding_window_pattern", None)
        has_interleaved_attention = sliding_window_pattern is not None or (
            isinstance(sliding_window, list))

        if not self.disable_sliding_window and has_interleaved_attention:
            if (backend :=
                    envs.VLLM_ATTENTION_BACKEND) in ("XFORMERS", "FLASHINFER"):
                sliding_window_len_min = get_min_sliding_window(
                    self.hf_text_config.sliding_window)

                logger.warning_once(
                    "%s has interleaved attention, which is currently not supported by the %s backend. Disabling sliding window and capping the max length to the sliding window size (%d).",  # noqa: E501
                    self.hf_text_config.model_type,
                    backend,
                    sliding_window_len_min,
                )
                self.disable_sliding_window = True
            else:
                # for a model with interleaved attention,
                # the scheduler and the model treat it as full attention
                # (i.e., not dropping any tokens outside the window).
                # only the attention layer itself is aware of the sliding
                # window, and use the window size to compute the attention.
                self.hf_text_config.interleaved_sliding_window = sliding_window

                if hasattr(self.hf_text_config, "sliding_window"):
                    delattr(self.hf_text_config, "sliding_window")

                sliding_window = None

        self.original_max_model_len = self.max_model_len
        self.max_model_len = self.get_and_verify_max_len(self.max_model_len)
        self.served_model_name = get_served_model_name(self.model,
                                                       self.served_model_name)
        self.multimodal_config = self._init_multimodal_config()
        if not self.skip_tokenizer_init:
            self._verify_tokenizer_mode()

        self.is_attention_free = self._init_attention_free()
        self.is_hybrid = self._init_is_hybrid()
        self.has_noops = self._init_has_noops()
        self.has_inner_state = self._init_has_inner_state()

        if (not current_platform.is_neuron() and self.override_neuron_config):
            raise ValueError(
                "`override_neuron_config` is only supported on Neuron.")

        self._verify_quantization()
        self._verify_cuda_graph()
        self._verify_bnb_config()

    @field_validator("quantization", mode="before")
    @classmethod
    def validate_quantization_before(cls, value: Any) -> Any:
        if isinstance(value, str):
            return value.lower()
        return value

    @model_validator(mode="after")
    def validate_model_config_after(self: "ModelConfig") -> "ModelConfig":
        if not isinstance(self.tokenizer, str):
            raise ValueError("tokenizer must be a string after __post_init__.")
        if not isinstance(self.max_model_len, int):
            raise ValueError(
                "max_model_len must be an integer after __post_init__.")
        return self

    @property
    def registry(self):
        return ModelRegistry

    @property
    def architectures(self) -> list[str]:
        return getattr(self.hf_config, "architectures", [])

    def maybe_pull_model_tokenizer_for_s3(self, model: str,
                                          tokenizer: str) -> None:
        """Pull model/tokenizer from S3 to temporary directory when needed.

        Args:
            model: Model name or path
            tokenizer: Tokenizer name or path
        """
        if not (is_s3(model) or is_s3(tokenizer)):
            return

        if is_s3(model):
            s3_model = S3Model()
            s3_model.pull_files(model,
                                allow_pattern=["*.model", "*.py", "*.json"])
            self.model_weights = model
            self.model = s3_model.dir

            # If tokenizer is same as model, download to same directory
            if model == tokenizer:
                s3_model.pull_files(
                    model, ignore_pattern=["*.pt", "*.safetensors", "*.bin"])
                self.tokenizer = s3_model.dir
                return

        # Only download tokenizer if needed and not already handled
        if is_s3(tokenizer):
            s3_tokenizer = S3Model()
            s3_tokenizer.pull_files(
                model, ignore_pattern=["*.pt", "*.safetensors", "*.bin"])
            self.tokenizer = s3_tokenizer.dir

    def _init_multimodal_config(self) -> Optional["MultiModalConfig"]:
        if self.registry.is_multimodal_model(self.architectures):
            return MultiModalConfig(
                limit_per_prompt=self.limit_mm_per_prompt,
                mm_processor_kwargs=self.mm_processor_kwargs,
                disable_mm_preprocessor_cache=self.
                disable_mm_preprocessor_cache)

        if self.limit_mm_per_prompt:
            raise ValueError("`limit_mm_per_prompt` is only supported for "
                             "multimodal models.")
        if self.mm_processor_kwargs:
            raise ValueError("`mm_processor_kwargs` is only supported for "
                             "multimodal models.")
        if self.disable_mm_preprocessor_cache:
            raise ValueError("`disable_mm_preprocessor_cache` is only "
                             "supported for multimodal models.")

        return None

    def _get_encoder_config(self):
        return get_sentence_transformer_tokenizer_config(
            self.model, self.revision)

    def _init_pooler_config(self) -> Optional["PoolerConfig"]:
        if self.runner_type == "pooling":
            if isinstance(self.override_pooler_config, dict):
                self.override_pooler_config = PoolerConfig(
                    **self.override_pooler_config)

            pooler_config = self.override_pooler_config or PoolerConfig()

            base_config = get_pooling_config(self.model, self.revision)
            if base_config is not None:
                # Only set values that are not overridden by the user
                for k, v in base_config.items():
                    if getattr(pooler_config, k) is None:
                        setattr(pooler_config, k, v)

            if self.is_matryoshka:
                if pooler_config.normalize is None:
                    pooler_config.normalize = True
                elif not pooler_config.normalize:
                    raise ValueError(
                        "`normalize` must be enabled (set to True) "
                        "for models that are compatible with "
                        "Matryoshka Representation.")

            return pooler_config

        return None

    def _init_attention_free(self) -> bool:
        return self.registry.is_attention_free_model(self.architectures)

    def _init_is_hybrid(self) -> bool:
        return self.registry.is_hybrid_model(self.architectures)

    def _init_has_noops(self) -> bool:
        architectures = getattr(self.hf_config, "architectures", [])
        return self.registry.is_noops_model(architectures)

    def _init_has_inner_state(self) -> bool:
        return self.registry.model_has_inner_state(self.architectures)

    def _verify_tokenizer_mode(self) -> None:
        tokenizer_mode = cast(TokenizerMode, self.tokenizer_mode.lower())
        if tokenizer_mode not in get_args(TokenizerMode):
            raise ValueError(
                f"Unknown tokenizer mode: {self.tokenizer_mode}. Must be "
                f"one of {get_args(TokenizerMode)}.")
        self.tokenizer_mode = tokenizer_mode

    def _get_preferred_task(
        self,
        architectures: list[str],
        supported_tasks: set[_ResolvedTask],
    ) -> Optional[_ResolvedTask]:
        model_id = self.model
        if get_pooling_config(model_id, self.revision):
            return "embed"
        if self.registry.is_cross_encoder_model(architectures):
            return "score"
        if self.registry.is_transcription_model(architectures):
            return "transcription"

        suffix_to_preferred_task: list[tuple[str, _ResolvedTask]] = [
            # Other models follow this pattern
            ("ForCausalLM", "generate"),
            ("ForConditionalGeneration", "generate"),
            ("ForSequenceClassification", "classify"),
            ("ChatModel", "generate"),
            ("LMHeadModel", "generate"),
            ("EmbeddingModel", "embed"),
            ("RewardModel", "reward"),
        ]
        _, arch = self.registry.inspect_model_cls(architectures)

        for suffix, pref_task in suffix_to_preferred_task:
            if arch.endswith(suffix) and pref_task in supported_tasks:
                return pref_task

        return None

    def _resolve_task(
        self,
        task_option: Literal[TaskOption, Literal["draft"]],
    ) -> tuple[set[_ResolvedTask], _ResolvedTask]:
        if task_option == "draft":
            return {"draft"}, "draft"

        registry = self.registry
        architectures = self.architectures

        runner_support: dict[RunnerType, bool] = {
            # NOTE: Listed from highest to lowest priority,
            # in case the model supports multiple of them
            "transcription": registry.is_transcription_model(architectures),
            "generate": registry.is_text_generation_model(architectures),
            "pooling": registry.is_pooling_model(architectures),
        }
        supported_runner_types_lst: list[RunnerType] = [
            runner_type
            for runner_type, is_supported in runner_support.items()
            if is_supported
        ]

        supported_tasks_lst: list[_ResolvedTask] = [
            task for runner_type in supported_runner_types_lst
            for task in _RUNNER_TASKS[runner_type]
        ]
        supported_tasks = set(supported_tasks_lst)

        if task_option == "auto":
            selected_task = next(iter(supported_tasks_lst))

            if len(supported_tasks_lst) > 1:
                preferred_task = self._get_preferred_task(
                    architectures, supported_tasks)
                if preferred_task is not None:
                    selected_task = preferred_task

                logger.info(
                    "This model supports multiple tasks: %s. "
                    "Defaulting to '%s'.", supported_tasks, selected_task)
        else:
            # Aliases
            if task_option == "embedding":
                msg = ("The 'embedding' task has been renamed to "
                       "'embed', please use the new name. The old name "
                       "will be removed in v1.0.")
                warnings.warn(msg, DeprecationWarning, stacklevel=2)

                task_option = "embed"

            if task_option not in supported_tasks:
                msg = (
                    f"This model does not support the '{task_option}' task. "
                    f"Supported tasks: {supported_tasks}")
                raise ValueError(msg)

            selected_task = task_option

        return supported_tasks, selected_task

    def _parse_quant_hf_config(self):
        quant_cfg = getattr(self.hf_config, "quantization_config", None)
        if quant_cfg is None:
            # compressed-tensors uses a "compression_config" key
            quant_cfg = getattr(self.hf_config, "compression_config", None)
        return quant_cfg

    def _verify_quantization(self) -> None:
        supported_quantization = QUANTIZATION_METHODS
        optimized_quantization_methods = [
            "fp8", "marlin", "modelopt", "gptq_marlin_24", "gptq_marlin",
            "awq_marlin", "fbgemm_fp8", "compressed-tensors", "experts_int8",
            "quark", "modelopt_fp4", "bitblas", "gptq_bitblas"
        ]
        if self.quantization is not None:
            self.quantization = cast(QuantizationMethods, self.quantization)

        # Parse quantization method from the HF model config, if available.
        quant_cfg = self._parse_quant_hf_config()

        if quant_cfg is not None:
            quant_method = quant_cfg.get("quant_method", "").lower()
            quant_method = quant_method.replace("compressed_tensors",
                                                "compressed-tensors")
            quant_cfg["quant_method"] = quant_method

            # Quantization methods which are overrides (i.e. they have a
            # `override_quantization_method` method) must be checked in order
            # of preference (this is particularly important for GPTQ).
            overrides = [
                "marlin",
                "bitblas",
                "gptq_marlin_24",
                "gptq_marlin",
                "gptq_bitblas",
                "awq_marlin",
                "ipex",
                "moe_wna16",
            ]
            quantization_methods = [
                q for q in supported_quantization if q not in overrides
            ]
            # Any custom overrides will be in quantization_methods so we place
            # them at the start of the list so custom overrides have preference
            # over the built in ones.
            quantization_methods = quantization_methods + overrides

            # Detect which checkpoint is it
            for name in quantization_methods:
                method = get_quantization_config(name)
                quantization_override = method.override_quantization_method(
                    quant_cfg, self.quantization)
                if quantization_override is not None:
                    # Raise error if the override is not custom (custom would
                    # be in QUANTIZATION_METHODS but not QuantizationMethods)
                    # and hasn't been added to the overrides list.
                    if (name in get_args(QuantizationMethods)
                            and name not in overrides):
                        raise ValueError(
                            f"Quantization method {name} is an override but "
                            "is has not been added to the `overrides` list "
                            "above. This is necessary to ensure that the "
                            "overrides are checked in order of preference.")
                    quant_method = quantization_override
                    self.quantization = quantization_override
                    break

            # Verify quantization configurations.
            if self.quantization is None:
                self.quantization = quant_method
            elif self.quantization != quant_method:
                raise ValueError(
                    "Quantization method specified in the model config "
                    f"({quant_method}) does not match the quantization "
                    f"method specified in the `quantization` argument "
                    f"({self.quantization}).")

        if self.quantization is not None:
            if self.quantization not in supported_quantization:
                raise ValueError(
                    f"Unknown quantization method: {self.quantization}. Must "
                    f"be one of {supported_quantization}.")
            from vllm.platforms import current_platform
            current_platform.verify_quantization(self.quantization)
            if self.quantization not in optimized_quantization_methods:
                logger.warning(
                    "%s quantization is not fully "
                    "optimized yet. The speed can be slower than "
                    "non-quantized models.", self.quantization)

    def _verify_cuda_graph(self) -> None:
        self.max_seq_len_to_capture = min(self.max_seq_len_to_capture,
                                          self.max_model_len)
        # CUDAGraph capture not supported for enc-dec models and mllama on ROCm
        ROCM_UNSUPPORTED_MODELS = ['mllama']
        unsupported_rocm = (self.hf_config.model_type
                            in ROCM_UNSUPPORTED_MODELS
                            or self.is_encoder_decoder)

        if (unsupported_rocm and not self.enforce_eager
                and current_platform.is_rocm()):
            logger.warning(
                "CUDA graph is not supported for %s on ROCm yet, fallback "
                "to eager mode.", self.hf_config.model_type)
            self.enforce_eager = True

    def _verify_bnb_config(self) -> None:
        """
        The current version of bitsandbytes (0.45.3) with 8-bit models does not
        yet support CUDA graph.
        # TODO Remove this when bitsandbytes supports.
        """
        is_bitsandbytes = self.quantization == "bitsandbytes"
        has_quantization_config = (getattr(self.hf_config,
                                           "quantization_config", None)
                                   is not None)
        is_8bit = (self.hf_config.quantization_config.get(
            "load_in_8bit", False) if has_quantization_config else False)
        if all([
                is_bitsandbytes,
                has_quantization_config,
                is_8bit,
                not self.enforce_eager,
        ]):
            logger.warning(
                "CUDA graph is not supported on BitsAndBytes 8bit yet, "
                "fallback to the eager mode.")

            self.enforce_eager = True

    def _verify_with_expert_parallelism(self) -> None:
        num_expert_names = [
            "moe_num_experts",  # Dbrx
            "num_experts",  # Jamba
            "n_routed_experts",  # DeepSeek
            "num_local_experts",  # Mixtral
        ]
        num_experts = 0
        for name in num_expert_names:
            num_experts = getattr(self.hf_text_config, name, 0)
            if num_experts > 0:
                break
        if num_experts < 1:
            raise ValueError(
                "Number of experts in the model must be greater than 0 "
                "when expert parallelism is enabled.")

    def verify_dual_chunk_attention_config(
        self,
        load_config: "LoadConfig",
    ) -> None:
        if hasattr(self.hf_config, "dual_chunk_attention_config"):
            # Try loading the sparse attention config
            from vllm.model_executor.model_loader.weight_utils import (
                get_sparse_attention_config)
            sparse_attn_config = get_sparse_attention_config(self, load_config)
            if sparse_attn_config:
                self.hf_config.dual_chunk_attention_config[
                    "sparse_attention_config"] = sparse_attn_config
                if "sparse_attention_enabled" not in \
                        self.hf_config.dual_chunk_attention_config:
                    self.hf_config.dual_chunk_attention_config[
                        "sparse_attention_enabled"] = True

    def verify_async_output_proc(self, parallel_config, speculative_config,
                                 device_config) -> None:
        if not self.use_async_output_proc:
            # Nothing to check
            return

        if parallel_config.pipeline_parallel_size > 1:
            self.use_async_output_proc = False
            return

        # Reminder: Please update docs/features/compatibility_matrix.md
        # If the feature combo become valid
        from vllm.platforms import current_platform
        if not current_platform.is_async_output_supported(self.enforce_eager):
            self.use_async_output_proc = False
            return

        if envs.VLLM_USE_RAY_SPMD_WORKER:
            self.use_async_output_proc = False
            return

        # Async postprocessor is not necessary for pooling models
        # since there is no token generation
        if self.runner_type == "pooling":
            self.use_async_output_proc = False

        # Reminder: Please update docs/features/compatibility_matrix.md
        # If the feature combo become valid
        if speculative_config:
            self.use_async_output_proc = False

    def verify_with_parallel_config(
        self,
        parallel_config: "ParallelConfig",
    ) -> None:

        if parallel_config.distributed_executor_backend == "external_launcher":
            assert self.seed is not None, (
                "Seed must be set when using external launcher backend to "
                "make sure sampling results are the same across workers.")

        total_num_attention_heads = getattr(self.hf_text_config,
                                            "num_attention_heads", 0)
        tensor_parallel_size = parallel_config.tensor_parallel_size
        if total_num_attention_heads % tensor_parallel_size != 0:
            raise ValueError(
                f"Total number of attention heads ({total_num_attention_heads})"
                " must be divisible by tensor parallel size "
                f"({tensor_parallel_size}).")

        if parallel_config.enable_expert_parallel:
            self._verify_with_expert_parallelism()

        pipeline_parallel_size = parallel_config.pipeline_parallel_size
        if pipeline_parallel_size > 1:
            if not self.registry.is_pp_supported_model(self.architectures):
                raise NotImplementedError(
                    "Pipeline parallelism is not supported for this model. "
                    "Supported models implement the `SupportsPP` interface.")

            if self.use_async_output_proc:
                self.use_async_output_proc = False

    def get_hf_config_sliding_window(
            self) -> Union[Optional[int], list[Optional[int]]]:
        """Get the sliding window size, or None if disabled."""

        # Some models, like Qwen2 and Qwen1.5, use `use_sliding_window` in
        # addition to sliding window size. We check if that field is present
        # and if it's False, return None.
        if (hasattr(self.hf_text_config, "use_sliding_window")
                and not self.hf_text_config.use_sliding_window):
            return None
        return getattr(self.hf_text_config, "sliding_window", None)

    def get_sliding_window(self) -> Optional[Union[int, list[Optional[int]]]]:
        """Get the sliding window size, or None if disabled.
        """
        # If user disables sliding window, return None.
        if self.disable_sliding_window:
            return None
        # Otherwise get the value from the hf config.
        return self.get_hf_config_sliding_window()

    def get_vocab_size(self) -> int:
        return self.hf_text_config.vocab_size

    def get_hidden_size(self) -> int:
        return self.hf_text_config.hidden_size

    @property
    def is_deepseek_mla(self) -> bool:
        if not hasattr(self.hf_text_config, "model_type"):
            return False
        elif self.hf_text_config.model_type in \
            ('deepseek_v2', 'deepseek_v3', 'deepseek_mtp'):
            return self.hf_text_config.kv_lora_rank is not None
        elif self.hf_text_config.model_type == 'eagle':
            # if the model is an EAGLE module, check for the
            # underlying architecture
            return self.hf_text_config.model.model_type in \
                    ('deepseek_v2', 'deepseek_v3') \
                and self.hf_text_config.kv_lora_rank is not None
        return False

    def get_head_size(self) -> int:
        # TODO remove hard code
        if self.is_deepseek_mla:
            qk_rope_head_dim = getattr(self.hf_text_config, "qk_rope_head_dim",
                                       0)
            if self.use_mla:
                return self.hf_text_config.kv_lora_rank + qk_rope_head_dim
            else:
                qk_nope_head_dim = getattr(self.hf_text_config,
                                           "qk_nope_head_dim", 0)
                if qk_rope_head_dim and qk_nope_head_dim:
                    return qk_rope_head_dim + qk_nope_head_dim

        if hasattr(self.hf_text_config,
                   "model_type") and (self.hf_text_config.model_type
                                      == "zamba2"):
            return self.hf_text_config.attention_head_dim

        if self.is_attention_free:
            return 0

        # NOTE: Some configs may set head_dim=None in the config
        if getattr(self.hf_text_config, "head_dim", None) is not None:
            return self.hf_text_config.head_dim

        # FIXME(woosuk): This may not be true for all models.
        return (self.hf_text_config.hidden_size //
                self.hf_text_config.num_attention_heads)

    def get_total_num_kv_heads(self) -> int:
        """Returns the total number of KV heads."""
        # For GPTBigCode & Falcon:
        # NOTE: for falcon, when new_decoder_architecture is True, the
        # multi_query flag is ignored and we use n_head_kv for the number of
        # KV heads.
        falcon_model_types = ["falcon", "RefinedWeb", "RefinedWebModel"]
        new_decoder_arch_falcon = (
            self.hf_config.model_type in falcon_model_types
            and getattr(self.hf_config, "new_decoder_architecture", False))
        if not new_decoder_arch_falcon and getattr(self.hf_text_config,
                                                   "multi_query", False):
            # Multi-query attention, only one KV head.
            # Currently, tensor parallelism is not supported in this case.
            return 1

        # For DBRX and MPT
        if self.hf_config.model_type == "mpt":
            if "kv_n_heads" in self.hf_config.attn_config:
                return self.hf_config.attn_config["kv_n_heads"]
            return self.hf_config.num_attention_heads
        if self.hf_config.model_type == "dbrx":
            return getattr(self.hf_config.attn_config, "kv_n_heads",
                           self.hf_config.num_attention_heads)

        if self.hf_config.model_type == "nemotron-nas":
            for block in self.hf_config.block_configs:
                if not block.attention.no_op:
                    return self.hf_config.num_attention_heads \
                        // block.attention.n_heads_in_group

            raise RuntimeError("Couldn't determine number of kv heads")

        if self.is_attention_free:
            return 0

        attributes = [
            # For Falcon:
            "n_head_kv",
            "num_kv_heads",
            # For LLaMA-2:
            "num_key_value_heads",
            # For ChatGLM:
            "multi_query_group_num",
        ]
        for attr in attributes:
            num_kv_heads = getattr(self.hf_text_config, attr, None)
            if num_kv_heads is not None:
                return num_kv_heads

        # For non-grouped-query attention models, the number of KV heads is
        # equal to the number of attention heads.
        return self.hf_text_config.num_attention_heads

    def get_num_kv_heads(self, parallel_config: "ParallelConfig") -> int:
        """Returns the number of KV heads per GPU."""
        if self.use_mla:
            # When using MLA during decode it becomes MQA
            return 1

        total_num_kv_heads = self.get_total_num_kv_heads()
        # If tensor parallelism is used, we divide the number of KV heads by
        # the tensor parallel size. We will replicate the KV heads in the
        # case where the number of KV heads is smaller than the tensor
        # parallel size so each GPU has at least one KV head.
        return max(1,
                   total_num_kv_heads // parallel_config.tensor_parallel_size)

    def get_num_attention_heads(self,
                                parallel_config: "ParallelConfig") -> int:
        num_heads = getattr(self.hf_text_config, "num_attention_heads", 0)
        return num_heads // parallel_config.tensor_parallel_size

    def get_layers_start_end_indices(
            self, parallel_config: "ParallelConfig") -> tuple[int, int]:
        from vllm.distributed.utils import get_pp_indices
        if (self.hf_text_config.model_type == "deepseek_mtp"
                or self.hf_config.model_type == "mimo_mtp"):
            total_num_hidden_layers = getattr(self.hf_text_config,
                                              "num_nextn_predict_layers", 0)
        else:
            total_num_hidden_layers = getattr(self.hf_text_config,
                                              "num_hidden_layers", 0)
        # the layout order is: DP x PP x TP
        pp_rank = (parallel_config.rank // parallel_config.tensor_parallel_size
                   ) % parallel_config.pipeline_parallel_size
        pp_size = parallel_config.pipeline_parallel_size
        start, end = get_pp_indices(total_num_hidden_layers, pp_rank, pp_size)
        return start, end

    def get_num_layers(self, parallel_config: "ParallelConfig") -> int:
        start, end = self.get_layers_start_end_indices(parallel_config)
        return end - start

    def get_num_layers_by_block_type(
        self,
        parallel_config: "ParallelConfig",
        block_type: LayerBlockType = LayerBlockType.attention,
    ) -> int:
        # This function relies on 'layers_block_type' in hf_config,
        # for w/o this attribute, we will need to have workarounds like so
        attn_block_type = block_type == LayerBlockType.attention
        is_transformer = not self.is_hybrid and \
                            not self.has_noops and \
                            not self.is_attention_free
        start, end = self.get_layers_start_end_indices(parallel_config)

        if is_transformer:
            # Handle the basic case first
            return end - start if attn_block_type else 0
        elif self.is_attention_free:
            # Attention free
            # Note that this code assumes there
            # is only one type of attention-free block type.
            return 0 if attn_block_type else end - start
        elif self.has_noops:
            block_configs = self.hf_config.block_configs
            return sum(not bc.attention.no_op
                       for bc in block_configs[start:end])
        else:
            # Hybrid model Jamba
            layers_block_type_value = getattr(self.hf_config,
                                              "layers_block_type", None)
            if layers_block_type_value is not None:
                if hasattr(self.hf_text_config,
                           "model_type") and (self.hf_text_config.model_type
                                              == "zamba2"):
                    if attn_block_type:
                        return sum(t == "hybrid"
                                   for t in layers_block_type_value[start:end])
                    else:
                        return self.get_num_layers(parallel_config)
                return sum(t == block_type.value
                           for t in layers_block_type_value[start:end])

            # Hybrid model Minimax
            attn_type_list = getattr(self.hf_config, "attn_type_list", None)
            if attn_type_list:
                return sum(t == 1 for t in attn_type_list[start:end])

            if layers_block_type_value is None and attn_type_list is None:
                raise ValueError(
                    "The model is an hybrid without a"
                    "layers_block_type or an attn_type_list in the hf_config,"
                    "cannot determine the num of "
                    f"{block_type.value} layers")

            return sum(t == 1 for t in attn_type_list[start:end])

    def get_multimodal_config(self) -> "MultiModalConfig":
        """
        Get the multimodal configuration of the model.

        Raises:
            ValueError: If the model is not multimodal.
        """
        if self.multimodal_config is None:
            raise ValueError("The model is not multimodal.")

        return self.multimodal_config

    def try_get_generation_config(self) -> dict[str, Any]:
        if self.generation_config in ("auto", "vllm"):
            config = try_get_generation_config(
                self.hf_config_path or self.model,
                trust_remote_code=self.trust_remote_code,
                revision=self.revision,
            )
        else:
            config = try_get_generation_config(
                self.generation_config,
                trust_remote_code=self.trust_remote_code,
            )

        if config is None:
            return {}

        return config.to_diff_dict()

    def get_diff_sampling_param(self) -> dict[str, Any]:
        """
        This method returns a dictionary containing the parameters
        that differ from the default sampling parameters. If
        `generation_config` is `"vllm"`, an empty dictionary is returned.

        Returns:
            dict[str, Any]: A dictionary with the differing sampling
            parameters, if `generation_config` is `"vllm"` an empty dictionary.
        """
        if self.generation_config == "vllm":
            config = {}
        else:
            config = self.try_get_generation_config()

        # Overriding with given generation config
        config.update(self.override_generation_config)

        available_params = [
            "repetition_penalty",
            "temperature",
            "top_k",
            "top_p",
            "min_p",
            "max_new_tokens",
        ]
        if any(p in config for p in available_params):
            diff_sampling_param = {
                p: config.get(p)
                for p in available_params if config.get(p) is not None
            }
            # Huggingface definition of max_new_tokens is equivalent
            # to vLLM's max_tokens
            if "max_new_tokens" in diff_sampling_param:
                diff_sampling_param["max_tokens"] = diff_sampling_param.pop(
                    "max_new_tokens")
        else:
            diff_sampling_param = {}

        if diff_sampling_param:
            logger.warning_once(
                "Default sampling parameters have been overridden by the "
                "model's Hugging Face generation config recommended from the "
                "model creator. If this is not intended, please relaunch "
                "vLLM instance with `--generation-config vllm`.")
        return diff_sampling_param

    @property
    def is_encoder_decoder(self) -> bool:
        """Extract the HF encoder/decoder model flag."""
        """
        For Mllama, VLLM overrides HF's is_encoder_decoder flag and sets it to
        True to enable cross-attention
        Neuron needs all multimodal data to be in the decoder and does not
        need to explicitly enable cross-attention
        """
        if (current_platform.is_neuron()
                and self.hf_config.model_type == "mllama"):
            return False

        return is_encoder_decoder(self.hf_config)

    @property
    def uses_mrope(self) -> bool:
        return uses_mrope(self.hf_config)

    @property
    def is_multimodal_model(self) -> bool:
        return self.multimodal_config is not None

    @property
    def is_cross_encoder(self) -> bool:
        return self.registry.is_cross_encoder_model(self.architectures)

    @property
    def use_mla(self) -> bool:
        return self.is_deepseek_mla and not envs.VLLM_MLA_DISABLE

    @property
    def supported_runner_types(self) -> set[RunnerType]:
        return {_TASK_RUNNER[task] for task in self.supported_tasks}

    @property
    def runner_type(self) -> RunnerType:
        return _TASK_RUNNER[cast(_ResolvedTask, self.task)]

    @property
    def is_v1_compatible(self) -> bool:
        architectures = getattr(self.hf_config, "architectures", [])
        return ModelRegistry.is_v1_compatible(architectures)

    @property
    def is_matryoshka(self) -> bool:
        return (hasattr(self.hf_config, "matryoshka_dimensions")
                or getattr(self.hf_config, "is_matryoshka", False))

    @property
    def matryoshka_dimensions(self):
        return getattr(self.hf_config, "matryoshka_dimensions", None)

    def get_and_verify_max_len(self, max_model_len: int):
        tokenizer_config = try_get_tokenizer_config(
            self.tokenizer,
            trust_remote_code=self.trust_remote_code,
            revision=self.tokenizer_revision)
        max_model_len = _get_and_verify_max_len(
            hf_config=self.hf_text_config,
            tokenizer_config=tokenizer_config,
            max_model_len=max_model_len,
            disable_sliding_window=self.disable_sliding_window,
            sliding_window_len=self.get_hf_config_sliding_window(),
            spec_target_max_model_len=self.spec_target_max_model_len,
            encoder_config=self.encoder_config)
        logger.info("Using max model len %s", max_model_len)
        return max_model_len

allowed_local_media_path class-attribute instance-attribute

allowed_local_media_path: str = ''

Allowing API requests to read local images or videos from directories specified by the server file system. This is a security risk. Should only be enabled in trusted environments.

architectures property

architectures: list[str]

code_revision class-attribute instance-attribute

code_revision: Optional[str] = None

The specific revision to use for the model code on the Hugging Face Hub. It can be a branch name, a tag name, or a commit id. If unspecified, will use the default version.

config_format class-attribute instance-attribute

config_format: Union[str, ConfigFormat] = AUTO.value

The format of the model config to load:

• "auto" will try to load the config in hf format if available else it will try to load in mistral format.

• "hf" will load the config in hf format.

• "mistral" will load the config in mistral format.

disable_cascade_attn class-attribute instance-attribute

disable_cascade_attn: bool = False

Disable cascade attention for V1. While cascade attention does not change the mathematical correctness, disabling it could be useful for preventing potential numerical issues. Note that even if this is set to False, cascade attention will be only used when the heuristic tells that it's beneficial.

disable_mm_preprocessor_cache class-attribute instance-attribute

disable_mm_preprocessor_cache: bool = False

If True, disable caching of the multi-modal preprocessor/mapper (not recommended).

disable_sliding_window class-attribute instance-attribute

disable_sliding_window: bool = False

Whether to disable sliding window. If True, we will disable the sliding window functionality of the model, capping to sliding window size. If the model does not support sliding window, this argument is ignored.

dtype class-attribute instance-attribute

dtype: Union[ModelDType, dtype] = 'auto'

Data type for model weights and activations:

• "auto" will use FP16 precision for FP32 and FP16 models, and BF16 precision for BF16 models.

• "half" for FP16. Recommended for AWQ quantization.

• "float16" is the same as "half".

• "bfloat16" for a balance between precision and range.

• "float" is shorthand for FP32 precision.

• "float32" for FP32 precision.

enable_prompt_embeds class-attribute instance-attribute

enable_prompt_embeds: bool = False

If True, enables passing text embeddings as inputs via the prompt_embeds key. Note that enabling this will double the time required for graph compilation.

enable_sleep_mode class-attribute instance-attribute

enable_sleep_mode: bool = False

Enable sleep mode for the engine (only cuda platform is supported).

enforce_eager class-attribute instance-attribute

enforce_eager: bool = False

Whether to always use eager-mode PyTorch. If True, we will disable CUDA graph and always execute the model in eager mode. If False, we will use CUDA graph and eager execution in hybrid for maximal performance and flexibility.

generation_config class-attribute instance-attribute

generation_config: str = 'auto'

The folder path to the generation config. Defaults to "auto", the generation config will be loaded from model path. If set to "vllm", no generation config is loaded, vLLM defaults will be used. If set to a folder path, the generation config will be loaded from the specified folder path. If max_new_tokens is specified in generation config, then it sets a server-wide limit on the number of output tokens for all requests.

hf_config_path class-attribute instance-attribute

hf_config_path: Optional[str] = None

Name or path of the Hugging Face config to use. If unspecified, model name or path will be used.

hf_overrides class-attribute instance-attribute

hf_overrides: HfOverrides = field(default_factory=dict)

If a dictionary, contains arguments to be forwarded to the Hugging Face config. If a callable, it is called to update the HuggingFace config.

hf_token class-attribute instance-attribute

hf_token: Optional[Union[bool, str]] = None

The token to use as HTTP bearer authorization for remote files . If True, will use the token generated when running huggingface-cli login (stored in ~/.huggingface).

is_cross_encoder property

is_cross_encoder: bool

is_deepseek_mla property

is_deepseek_mla: bool

is_encoder_decoder property

is_encoder_decoder: bool

Extract the HF encoder/decoder model flag.

is_matryoshka property

is_matryoshka: bool

is_multimodal_model property

is_multimodal_model: bool

is_v1_compatible property

is_v1_compatible: bool

limit_mm_per_prompt class-attribute instance-attribute

limit_mm_per_prompt: dict[str, int] = field(
    default_factory=dict
)

Maximum number of data items per modality per prompt. Only applicable for multimodal models.

logits_processor_pattern class-attribute instance-attribute

logits_processor_pattern: Optional[str] = None

Optional regex pattern specifying valid logits processor qualified names that can be passed with the logits_processors extra completion argument. Defaults to None, which allows no processors.

matryoshka_dimensions property

matryoshka_dimensions

max_logprobs class-attribute instance-attribute

max_logprobs: int = 20

Maximum number of log probabilities to return when logprobs is specified in SamplingParams. The default value comes the default for the OpenAI Chat Completions API.

max_model_len class-attribute instance-attribute

max_model_len: SkipValidation[int] = None

Model context length (prompt and output). If unspecified, will be automatically derived from the model config.

When passing via --max-model-len, supports k/m/g/K/M/G in human-readable format. Examples:

• 1k -> 1000

• 1K -> 1024

• 25.6k -> 25,600

max_seq_len_to_capture class-attribute instance-attribute

max_seq_len_to_capture: int = 8192

Maximum sequence len covered by CUDA graphs. When a sequence has context length larger than this, we fall back to eager mode. Additionally for encoder-decoder models, if the sequence length of the encoder input is larger than this, we fall back to the eager mode.

mm_processor_kwargs class-attribute instance-attribute

mm_processor_kwargs: Optional[dict[str, Any]] = None

Arguments to be forwarded to the model's processor for multi-modal data, e.g., image processor. Overrides for the multi-modal processor obtained from AutoProcessor.from_pretrained. The available overrides depend on the model that is being run. For example, for Phi-3-Vision: {"num_crops": 4}.

model class-attribute instance-attribute

model: str = 'facebook/opt-125m'

Name or path of the Hugging Face model to use. It is also used as the content for model_name tag in metrics output when served_model_name is not specified.

model_impl class-attribute instance-attribute

model_impl: Union[str, ModelImpl] = AUTO.value

Which implementation of the model to use:

• "auto" will try to use the vLLM implementation, if it exists, and fall back to the Transformers implementation if no vLLM implementation is available.

• "vllm" will use the vLLM model implementation.

• "transformers" will use the Transformers model implementation.

override_attention_dtype class-attribute instance-attribute

override_attention_dtype: Optional[str] = None

Override dtype for attention

override_generation_config class-attribute instance-attribute

override_generation_config: dict[str, Any] = field(
    default_factory=dict
)

Overrides or sets generation config. e.g. {"temperature": 0.5}. If used with --generation-config auto, the override parameters will be merged with the default config from the model. If used with --generation-config vllm, only the override parameters are used.

override_neuron_config class-attribute instance-attribute

override_neuron_config: dict[str, Any] = field(
    default_factory=dict
)

Initialize non-default neuron config or override default neuron config that are specific to Neuron devices, this argument will be used to configure the neuron config that can not be gathered from the vllm arguments. e.g. {"cast_logits_dtype": "bfloat16"}.

override_pooler_config class-attribute instance-attribute

override_pooler_config: Optional[
    Union[dict, PoolerConfig]
] = None

Initialize non-default pooling config or override default pooling config for the pooling model. e.g. {"pooling_type": "mean", "normalize": false}.

pooler_config class-attribute instance-attribute

pooler_config: Optional[PoolerConfig] = field(init=False)

Pooler config which controls the behaviour of output pooling in pooling models.

quantization class-attribute instance-attribute

quantization: SkipValidation[
    Optional[QuantizationMethods]
] = None

Method used to quantize the weights. If None, we first check the quantization_config attribute in the model config file. If that is None, we assume the model weights are not quantized and use dtype to determine the data type of the weights.

registry property

registry

revision class-attribute instance-attribute

revision: Optional[str] = None

The specific model version to use. It can be a branch name, a tag name, or a commit id. If unspecified, will use the default version.

rope_scaling class-attribute instance-attribute

rope_scaling: dict[str, Any] = field(default_factory=dict)

RoPE scaling configuration. For example, {"rope_type":"dynamic","factor":2.0}.

rope_theta class-attribute instance-attribute

rope_theta: Optional[float] = None

RoPE theta. Use with rope_scaling. In some cases, changing the RoPE theta improves the performance of the scaled model.

runner_type property

runner_type: RunnerType

seed class-attribute instance-attribute

seed: Optional[int] = None

Random seed for reproducibility. Initialized to None in V0, but initialized to 0 in V1.

served_model_name class-attribute instance-attribute

served_model_name: Optional[Union[str, list[str]]] = None

The model name(s) used in the API. If multiple names are provided, the server will respond to any of the provided names. The model name in the model field of a response will be the first name in this list. If not specified, the model name will be the same as the --model argument. Noted that this name(s) will also be used in model_name tag content of prometheus metrics, if multiple names provided, metrics tag will take the first one.

skip_tokenizer_init class-attribute instance-attribute

skip_tokenizer_init: bool = False

Skip initialization of tokenizer and detokenizer. Expects valid prompt_token_ids and None for prompt from the input. The generated output will contain token ids.

spec_target_max_model_len class-attribute instance-attribute

spec_target_max_model_len: Optional[int] = None

Specify the maximum length for spec decoding draft models.

supported_runner_types property

supported_runner_types: set[RunnerType]

task class-attribute instance-attribute

task: Literal[TaskOption, Literal['draft']] = 'auto'

The task to use the model for. Each vLLM instance only supports one task, even if the same model can be used for multiple tasks. When the model only supports one task, "auto" can be used to select it; otherwise, you must specify explicitly which task to use.

tokenizer class-attribute instance-attribute

tokenizer: SkipValidation[str] = None

Name or path of the Hugging Face tokenizer to use. If unspecified, model name or path will be used.

tokenizer_mode class-attribute instance-attribute

tokenizer_mode: TokenizerMode = 'auto'

Tokenizer mode:

• "auto" will use the fast tokenizer if available.

• "slow" will always use the slow tokenizer.

• "mistral" will always use the tokenizer from mistral_common.

• "custom" will use --tokenizer to select the preregistered tokenizer.

tokenizer_revision class-attribute instance-attribute

tokenizer_revision: Optional[str] = None

The specific revision to use for the tokenizer on the Hugging Face Hub. It can be a branch name, a tag name, or a commit id. If unspecified, will use the default version.

trust_remote_code class-attribute instance-attribute

trust_remote_code: bool = False

Trust remote code (e.g., from HuggingFace) when downloading the model and tokenizer.

use_async_output_proc class-attribute instance-attribute

use_async_output_proc: bool = True

Whether to use async output processor.

use_mla property

use_mla: bool

uses_mrope property

uses_mrope: bool

__post_init__

__post_init__() -> None

Source code in vllm/config.py

def __post_init__(self) -> None:
    # Set the default seed to 0 in V1.
    # NOTE(woosuk): In V0, we set the default seed to None because the
    # driver worker shares the same process as the user process, and thus
    # setting a seed affects the user process as well.
    # In V1, we use separate processes for workers (unless
    # VLLM_ENABLE_V1_MULTIPROCESSING=0), so setting a seed here
    # doesn't affect the user process. However, without a consistent seed,
    # different tensor parallel workers would sample different tokens,
    # leading to inconsistent results.
    if envs.VLLM_USE_V1 and self.seed is None:
        self.seed = 0
        if not envs.VLLM_ENABLE_V1_MULTIPROCESSING:
            logger.warning(
                "The global random seed is set to %d. Since "
                "VLLM_ENABLE_V1_MULTIPROCESSING is set to False, this may "
                "affect the random state of the Python process that "
                "launched vLLM.", self.seed)

    self.model = maybe_model_redirect(self.model)
    # The tokenizer is consistent with the model by default.
    if self.tokenizer is None:
        self.tokenizer = self.model
    if self.tokenizer_revision is None:
        self.tokenizer_revision = self.revision
    self.tokenizer = maybe_model_redirect(self.tokenizer)

    if isinstance(self.hf_config_path, str):
        self.hf_config_path = maybe_model_redirect(self.hf_config_path)

    if callable(self.hf_overrides):
        hf_overrides_kw = {}
        hf_overrides_fn = self.hf_overrides
    else:
        hf_overrides_kw = self.hf_overrides
        hf_overrides_fn = None

    if self.rope_scaling:
        hf_override: dict[str, Any] = {"rope_scaling": self.rope_scaling}
        hf_overrides_kw.update(hf_override)
        hf_overrides_str = json.dumps(hf_overrides_kw)
        msg = (
            "`--rope-scaling` will be removed in a future release. "
            f"'Please instead use `--hf-overrides '{hf_overrides_str}'`")
        warnings.warn(DeprecationWarning(msg), stacklevel=2)
    if self.rope_theta is not None:
        hf_override = {"rope_theta": self.rope_theta}
        hf_overrides_kw.update(hf_override)
        hf_overrides_str = json.dumps(hf_overrides_kw)
        msg = (
            "`--rope-theta` will be removed in a future release. "
            f"'Please instead use `--hf-overrides '{hf_overrides_str}'`")
        warnings.warn(DeprecationWarning(msg), stacklevel=2)

    self.maybe_pull_model_tokenizer_for_s3(self.model, self.tokenizer)

    if (backend := envs.VLLM_ATTENTION_BACKEND
        ) and backend == "FLASHINFER" and find_spec("flashinfer") is None:
        raise ValueError(
            "VLLM_ATTENTION_BACKEND is set to FLASHINFER, but flashinfer "
            "module was not found. See "
            "https://github.com/vllm-project/vllm/blob/main/docker/Dockerfile "  # noqa: E501
            "for instructions on how to install it.")

    from vllm.platforms import current_platform

    if (self.override_attention_dtype is not None
            and not current_platform.is_rocm()):
        warnings.warn(
            "override-attention-dtype is set but not using ROCm platform",
            stacklevel=2)

    if (self.enable_sleep_mode
            and not current_platform.is_sleep_mode_available()):
        raise ValueError(
            "Sleep mode is not supported on current platform.")

    if isinstance(self.config_format, str):
        self.config_format = ConfigFormat(self.config_format)

    hf_config = get_config(self.hf_config_path or self.model,
                           self.trust_remote_code, self.revision,
                           self.code_revision, self.config_format)

    if hf_overrides_kw:
        logger.info("Overriding HF config with %s", hf_overrides_kw)
        hf_config.update(hf_overrides_kw)
    if hf_overrides_fn:
        logger.info("Overriding HF config with %s", hf_overrides_fn)
        hf_config = hf_overrides_fn(hf_config)

    self.hf_config = hf_config

    self.hf_text_config = get_hf_text_config(self.hf_config)
    self.attention_chunk_size = getattr(self.hf_text_config,
                                        "attention_chunk_size", None)
    self.encoder_config = self._get_encoder_config()
    self.hf_image_processor_config = get_hf_image_processor_config(
        self.model, hf_token=self.hf_token, revision=self.revision)

    supported_tasks, task = self._resolve_task(self.task)
    self.supported_tasks = supported_tasks
    self.task = task
    if self.task in ("draft", "generate"):
        self.truncation_side = "left"
    else:
        self.truncation_side = "right"

    self.pooler_config = self._init_pooler_config()

    self.dtype = _get_and_verify_dtype(
        self.model,
        self.hf_config,
        self.dtype,
        is_pooling_model=self.runner_type == "pooling",
        revision=self.revision,
    )

    # Workaround for Gemma 2 which uses interleaved sliding window
    # attention, but it's not specified in its config. TODO: remove this
    # when Gemma 2 is fixed in Transformers.
    if self.hf_text_config.model_type == "gemma2":
        self.hf_text_config.sliding_window_pattern = 2

    sliding_window = getattr(self.hf_text_config, "sliding_window", None)
    sliding_window_pattern = getattr(self.hf_text_config,
                                     "sliding_window_pattern", None)
    has_interleaved_attention = sliding_window_pattern is not None or (
        isinstance(sliding_window, list))

    if not self.disable_sliding_window and has_interleaved_attention:
        if (backend :=
                envs.VLLM_ATTENTION_BACKEND) in ("XFORMERS", "FLASHINFER"):
            sliding_window_len_min = get_min_sliding_window(
                self.hf_text_config.sliding_window)

            logger.warning_once(
                "%s has interleaved attention, which is currently not supported by the %s backend. Disabling sliding window and capping the max length to the sliding window size (%d).",  # noqa: E501
                self.hf_text_config.model_type,
                backend,
                sliding_window_len_min,
            )
            self.disable_sliding_window = True
        else:
            # for a model with interleaved attention,
            # the scheduler and the model treat it as full attention
            # (i.e., not dropping any tokens outside the window).
            # only the attention layer itself is aware of the sliding
            # window, and use the window size to compute the attention.
            self.hf_text_config.interleaved_sliding_window = sliding_window

            if hasattr(self.hf_text_config, "sliding_window"):
                delattr(self.hf_text_config, "sliding_window")

            sliding_window = None

    self.original_max_model_len = self.max_model_len
    self.max_model_len = self.get_and_verify_max_len(self.max_model_len)
    self.served_model_name = get_served_model_name(self.model,
                                                   self.served_model_name)
    self.multimodal_config = self._init_multimodal_config()
    if not self.skip_tokenizer_init:
        self._verify_tokenizer_mode()

    self.is_attention_free = self._init_attention_free()
    self.is_hybrid = self._init_is_hybrid()
    self.has_noops = self._init_has_noops()
    self.has_inner_state = self._init_has_inner_state()

    if (not current_platform.is_neuron() and self.override_neuron_config):
        raise ValueError(
            "`override_neuron_config` is only supported on Neuron.")

    self._verify_quantization()
    self._verify_cuda_graph()
    self._verify_bnb_config()

_get_encoder_config

_get_encoder_config()

Source code in vllm/config.py

def _get_encoder_config(self):
    return get_sentence_transformer_tokenizer_config(
        self.model, self.revision)

_get_preferred_task

_get_preferred_task(
    architectures: list[str],
    supported_tasks: set[_ResolvedTask],
) -> Optional[_ResolvedTask]

Source code in vllm/config.py

def _get_preferred_task(
    self,
    architectures: list[str],
    supported_tasks: set[_ResolvedTask],
) -> Optional[_ResolvedTask]:
    model_id = self.model
    if get_pooling_config(model_id, self.revision):
        return "embed"
    if self.registry.is_cross_encoder_model(architectures):
        return "score"
    if self.registry.is_transcription_model(architectures):
        return "transcription"

    suffix_to_preferred_task: list[tuple[str, _ResolvedTask]] = [
        # Other models follow this pattern
        ("ForCausalLM", "generate"),
        ("ForConditionalGeneration", "generate"),
        ("ForSequenceClassification", "classify"),
        ("ChatModel", "generate"),
        ("LMHeadModel", "generate"),
        ("EmbeddingModel", "embed"),
        ("RewardModel", "reward"),
    ]
    _, arch = self.registry.inspect_model_cls(architectures)

    for suffix, pref_task in suffix_to_preferred_task:
        if arch.endswith(suffix) and pref_task in supported_tasks:
            return pref_task

    return None

_init_attention_free

_init_attention_free() -> bool

Source code in vllm/config.py

def _init_attention_free(self) -> bool:
    return self.registry.is_attention_free_model(self.architectures)

_init_has_inner_state

_init_has_inner_state() -> bool

Source code in vllm/config.py

def _init_has_inner_state(self) -> bool:
    return self.registry.model_has_inner_state(self.architectures)

_init_has_noops

_init_has_noops() -> bool

Source code in vllm/config.py

def _init_has_noops(self) -> bool:
    architectures = getattr(self.hf_config, "architectures", [])
    return self.registry.is_noops_model(architectures)

_init_is_hybrid

_init_is_hybrid() -> bool

Source code in vllm/config.py

def _init_is_hybrid(self) -> bool:
    return self.registry.is_hybrid_model(self.architectures)

_init_multimodal_config

_init_multimodal_config() -> Optional[MultiModalConfig]

Source code in vllm/config.py

def _init_multimodal_config(self) -> Optional["MultiModalConfig"]:
    if self.registry.is_multimodal_model(self.architectures):
        return MultiModalConfig(
            limit_per_prompt=self.limit_mm_per_prompt,
            mm_processor_kwargs=self.mm_processor_kwargs,
            disable_mm_preprocessor_cache=self.
            disable_mm_preprocessor_cache)

    if self.limit_mm_per_prompt:
        raise ValueError("`limit_mm_per_prompt` is only supported for "
                         "multimodal models.")
    if self.mm_processor_kwargs:
        raise ValueError("`mm_processor_kwargs` is only supported for "
                         "multimodal models.")
    if self.disable_mm_preprocessor_cache:
        raise ValueError("`disable_mm_preprocessor_cache` is only "
                         "supported for multimodal models.")

    return None

_init_pooler_config

_init_pooler_config() -> Optional[PoolerConfig]

Source code in vllm/config.py

def _init_pooler_config(self) -> Optional["PoolerConfig"]:
    if self.runner_type == "pooling":
        if isinstance(self.override_pooler_config, dict):
            self.override_pooler_config = PoolerConfig(
                **self.override_pooler_config)

        pooler_config = self.override_pooler_config or PoolerConfig()

        base_config = get_pooling_config(self.model, self.revision)
        if base_config is not None:
            # Only set values that are not overridden by the user
            for k, v in base_config.items():
                if getattr(pooler_config, k) is None:
                    setattr(pooler_config, k, v)

        if self.is_matryoshka:
            if pooler_config.normalize is None:
                pooler_config.normalize = True
            elif not pooler_config.normalize:
                raise ValueError(
                    "`normalize` must be enabled (set to True) "
                    "for models that are compatible with "
                    "Matryoshka Representation.")

        return pooler_config

    return None

_parse_quant_hf_config

_parse_quant_hf_config()

Source code in vllm/config.py

def _parse_quant_hf_config(self):
    quant_cfg = getattr(self.hf_config, "quantization_config", None)
    if quant_cfg is None:
        # compressed-tensors uses a "compression_config" key
        quant_cfg = getattr(self.hf_config, "compression_config", None)
    return quant_cfg

_resolve_task

_resolve_task(
    task_option: Literal[TaskOption, Literal["draft"]],
) -> tuple[set[_ResolvedTask], _ResolvedTask]

Source code in vllm/config.py

def _resolve_task(
    self,
    task_option: Literal[TaskOption, Literal["draft"]],
) -> tuple[set[_ResolvedTask], _ResolvedTask]:
    if task_option == "draft":
        return {"draft"}, "draft"

    registry = self.registry
    architectures = self.architectures

    runner_support: dict[RunnerType, bool] = {
        # NOTE: Listed from highest to lowest priority,
        # in case the model supports multiple of them
        "transcription": registry.is_transcription_model(architectures),
        "generate": registry.is_text_generation_model(architectures),
        "pooling": registry.is_pooling_model(architectures),
    }
    supported_runner_types_lst: list[RunnerType] = [
        runner_type
        for runner_type, is_supported in runner_support.items()
        if is_supported
    ]

    supported_tasks_lst: list[_ResolvedTask] = [
        task for runner_type in supported_runner_types_lst
        for task in _RUNNER_TASKS[runner_type]
    ]
    supported_tasks = set(supported_tasks_lst)

    if task_option == "auto":
        selected_task = next(iter(supported_tasks_lst))

        if len(supported_tasks_lst) > 1:
            preferred_task = self._get_preferred_task(
                architectures, supported_tasks)
            if preferred_task is not None:
                selected_task = preferred_task

            logger.info(
                "This model supports multiple tasks: %s. "
                "Defaulting to '%s'.", supported_tasks, selected_task)
    else:
        # Aliases
        if task_option == "embedding":
            msg = ("The 'embedding' task has been renamed to "
                   "'embed', please use the new name. The old name "
                   "will be removed in v1.0.")
            warnings.warn(msg, DeprecationWarning, stacklevel=2)

            task_option = "embed"

        if task_option not in supported_tasks:
            msg = (
                f"This model does not support the '{task_option}' task. "
                f"Supported tasks: {supported_tasks}")
            raise ValueError(msg)

        selected_task = task_option

    return supported_tasks, selected_task

_verify_bnb_config

_verify_bnb_config() -> None

The current version of bitsandbytes (0.45.3) with 8-bit models does not yet support CUDA graph.

TODO Remove this when bitsandbytes supports.

Source code in vllm/config.py

def _verify_bnb_config(self) -> None:
    """
    The current version of bitsandbytes (0.45.3) with 8-bit models does not
    yet support CUDA graph.
    # TODO Remove this when bitsandbytes supports.
    """
    is_bitsandbytes = self.quantization == "bitsandbytes"
    has_quantization_config = (getattr(self.hf_config,
                                       "quantization_config", None)
                               is not None)
    is_8bit = (self.hf_config.quantization_config.get(
        "load_in_8bit", False) if has_quantization_config else False)
    if all([
            is_bitsandbytes,
            has_quantization_config,
            is_8bit,
            not self.enforce_eager,
    ]):
        logger.warning(
            "CUDA graph is not supported on BitsAndBytes 8bit yet, "
            "fallback to the eager mode.")

        self.enforce_eager = True

_verify_cuda_graph

_verify_cuda_graph() -> None

Source code in vllm/config.py

def _verify_cuda_graph(self) -> None:
    self.max_seq_len_to_capture = min(self.max_seq_len_to_capture,
                                      self.max_model_len)
    # CUDAGraph capture not supported for enc-dec models and mllama on ROCm
    ROCM_UNSUPPORTED_MODELS = ['mllama']
    unsupported_rocm = (self.hf_config.model_type
                        in ROCM_UNSUPPORTED_MODELS
                        or self.is_encoder_decoder)

    if (unsupported_rocm and not self.enforce_eager
            and current_platform.is_rocm()):
        logger.warning(
            "CUDA graph is not supported for %s on ROCm yet, fallback "
            "to eager mode.", self.hf_config.model_type)
        self.enforce_eager = True

_verify_quantization

_verify_quantization() -> None

Source code in vllm/config.py

def _verify_quantization(self) -> None:
    supported_quantization = QUANTIZATION_METHODS
    optimized_quantization_methods = [
        "fp8", "marlin", "modelopt", "gptq_marlin_24", "gptq_marlin",
        "awq_marlin", "fbgemm_fp8", "compressed-tensors", "experts_int8",
        "quark", "modelopt_fp4", "bitblas", "gptq_bitblas"
    ]
    if self.quantization is not None:
        self.quantization = cast(QuantizationMethods, self.quantization)

    # Parse quantization method from the HF model config, if available.
    quant_cfg = self._parse_quant_hf_config()

    if quant_cfg is not None:
        quant_method = quant_cfg.get("quant_method", "").lower()
        quant_method = quant_method.replace("compressed_tensors",
                                            "compressed-tensors")
        quant_cfg["quant_method"] = quant_method

        # Quantization methods which are overrides (i.e. they have a
        # `override_quantization_method` method) must be checked in order
        # of preference (this is particularly important for GPTQ).
        overrides = [
            "marlin",
            "bitblas",
            "gptq_marlin_24",
            "gptq_marlin",
            "gptq_bitblas",
            "awq_marlin",
            "ipex",
            "moe_wna16",
        ]
        quantization_methods = [
            q for q in supported_quantization if q not in overrides
        ]
        # Any custom overrides will be in quantization_methods so we place
        # them at the start of the list so custom overrides have preference
        # over the built in ones.
        quantization_methods = quantization_methods + overrides

        # Detect which checkpoint is it
        for name in quantization_methods:
            method = get_quantization_config(name)
            quantization_override = method.override_quantization_method(
                quant_cfg, self.quantization)
            if quantization_override is not None:
                # Raise error if the override is not custom (custom would
                # be in QUANTIZATION_METHODS but not QuantizationMethods)
                # and hasn't been added to the overrides list.
                if (name in get_args(QuantizationMethods)
                        and name not in overrides):
                    raise ValueError(
                        f"Quantization method {name} is an override but "
                        "is has not been added to the `overrides` list "
                        "above. This is necessary to ensure that the "
                        "overrides are checked in order of preference.")
                quant_method = quantization_override
                self.quantization = quantization_override
                break

        # Verify quantization configurations.
        if self.quantization is None:
            self.quantization = quant_method
        elif self.quantization != quant_method:
            raise ValueError(
                "Quantization method specified in the model config "
                f"({quant_method}) does not match the quantization "
                f"method specified in the `quantization` argument "
                f"({self.quantization}).")

    if self.quantization is not None:
        if self.quantization not in supported_quantization:
            raise ValueError(
                f"Unknown quantization method: {self.quantization}. Must "
                f"be one of {supported_quantization}.")
        from vllm.platforms import current_platform
        current_platform.verify_quantization(self.quantization)
        if self.quantization not in optimized_quantization_methods:
            logger.warning(
                "%s quantization is not fully "
                "optimized yet. The speed can be slower than "
                "non-quantized models.", self.quantization)

_verify_tokenizer_mode

_verify_tokenizer_mode() -> None

Source code in vllm/config.py

def _verify_tokenizer_mode(self) -> None:
    tokenizer_mode = cast(TokenizerMode, self.tokenizer_mode.lower())
    if tokenizer_mode not in get_args(TokenizerMode):
        raise ValueError(
            f"Unknown tokenizer mode: {self.tokenizer_mode}. Must be "
            f"one of {get_args(TokenizerMode)}.")
    self.tokenizer_mode = tokenizer_mode

_verify_with_expert_parallelism

_verify_with_expert_parallelism() -> None

Source code in vllm/config.py

def _verify_with_expert_parallelism(self) -> None:
    num_expert_names = [
        "moe_num_experts",  # Dbrx
        "num_experts",  # Jamba
        "n_routed_experts",  # DeepSeek
        "num_local_experts",  # Mixtral
    ]
    num_experts = 0
    for name in num_expert_names:
        num_experts = getattr(self.hf_text_config, name, 0)
        if num_experts > 0:
            break
    if num_experts < 1:
        raise ValueError(
            "Number of experts in the model must be greater than 0 "
            "when expert parallelism is enabled.")

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    factors: list[Any] = []
    factors.append(self.model)
    factors.append(self.dtype)
    factors.append(self.quantization)
    factors.append(self.revision)
    factors.append(self.code_revision)
    factors.append(self.max_model_len)
    factors.append(self.max_logprobs)
    factors.append(self.disable_sliding_window)
    factors.append(self.trust_remote_code)
    factors.append(self.generation_config)
    factors.append(self.model_impl)
    factors.append(self.override_generation_config)
    factors.append(self.rope_scaling)
    factors.append(self.rope_theta)
    # hf_config can control how the model looks!
    factors.append(self.hf_config.to_json_string())
    str_factors = str(factors)
    assert_hashable(str_factors)
    return hashlib.sha256(str(factors).encode()).hexdigest()

get_and_verify_max_len

get_and_verify_max_len(max_model_len: int)

Source code in vllm/config.py

def get_and_verify_max_len(self, max_model_len: int):
    tokenizer_config = try_get_tokenizer_config(
        self.tokenizer,
        trust_remote_code=self.trust_remote_code,
        revision=self.tokenizer_revision)
    max_model_len = _get_and_verify_max_len(
        hf_config=self.hf_text_config,
        tokenizer_config=tokenizer_config,
        max_model_len=max_model_len,
        disable_sliding_window=self.disable_sliding_window,
        sliding_window_len=self.get_hf_config_sliding_window(),
        spec_target_max_model_len=self.spec_target_max_model_len,
        encoder_config=self.encoder_config)
    logger.info("Using max model len %s", max_model_len)
    return max_model_len

get_diff_sampling_param

get_diff_sampling_param() -> dict[str, Any]

This method returns a dictionary containing the parameters that differ from the default sampling parameters. If generation_config is "vllm", an empty dictionary is returned.

Returns:

Type	Description
dict[str, Any]	dict[str, Any]: A dictionary with the differing sampling
dict[str, Any]	parameters, if generation_config is "vllm" an empty dictionary.

Source code in vllm/config.py

def get_diff_sampling_param(self) -> dict[str, Any]:
    """
    This method returns a dictionary containing the parameters
    that differ from the default sampling parameters. If
    `generation_config` is `"vllm"`, an empty dictionary is returned.

    Returns:
        dict[str, Any]: A dictionary with the differing sampling
        parameters, if `generation_config` is `"vllm"` an empty dictionary.
    """
    if self.generation_config == "vllm":
        config = {}
    else:
        config = self.try_get_generation_config()

    # Overriding with given generation config
    config.update(self.override_generation_config)

    available_params = [
        "repetition_penalty",
        "temperature",
        "top_k",
        "top_p",
        "min_p",
        "max_new_tokens",
    ]
    if any(p in config for p in available_params):
        diff_sampling_param = {
            p: config.get(p)
            for p in available_params if config.get(p) is not None
        }
        # Huggingface definition of max_new_tokens is equivalent
        # to vLLM's max_tokens
        if "max_new_tokens" in diff_sampling_param:
            diff_sampling_param["max_tokens"] = diff_sampling_param.pop(
                "max_new_tokens")
    else:
        diff_sampling_param = {}

    if diff_sampling_param:
        logger.warning_once(
            "Default sampling parameters have been overridden by the "
            "model's Hugging Face generation config recommended from the "
            "model creator. If this is not intended, please relaunch "
            "vLLM instance with `--generation-config vllm`.")
    return diff_sampling_param

get_head_size

get_head_size() -> int

Source code in vllm/config.py

def get_head_size(self) -> int:
    # TODO remove hard code
    if self.is_deepseek_mla:
        qk_rope_head_dim = getattr(self.hf_text_config, "qk_rope_head_dim",
                                   0)
        if self.use_mla:
            return self.hf_text_config.kv_lora_rank + qk_rope_head_dim
        else:
            qk_nope_head_dim = getattr(self.hf_text_config,
                                       "qk_nope_head_dim", 0)
            if qk_rope_head_dim and qk_nope_head_dim:
                return qk_rope_head_dim + qk_nope_head_dim

    if hasattr(self.hf_text_config,
               "model_type") and (self.hf_text_config.model_type
                                  == "zamba2"):
        return self.hf_text_config.attention_head_dim

    if self.is_attention_free:
        return 0

    # NOTE: Some configs may set head_dim=None in the config
    if getattr(self.hf_text_config, "head_dim", None) is not None:
        return self.hf_text_config.head_dim

    # FIXME(woosuk): This may not be true for all models.
    return (self.hf_text_config.hidden_size //
            self.hf_text_config.num_attention_heads)

get_hf_config_sliding_window

get_hf_config_sliding_window() -> Union[
    Optional[int], list[Optional[int]]
]

Get the sliding window size, or None if disabled.

Source code in vllm/config.py

def get_hf_config_sliding_window(
        self) -> Union[Optional[int], list[Optional[int]]]:
    """Get the sliding window size, or None if disabled."""

    # Some models, like Qwen2 and Qwen1.5, use `use_sliding_window` in
    # addition to sliding window size. We check if that field is present
    # and if it's False, return None.
    if (hasattr(self.hf_text_config, "use_sliding_window")
            and not self.hf_text_config.use_sliding_window):
        return None
    return getattr(self.hf_text_config, "sliding_window", None)

get_hidden_size

get_hidden_size() -> int

Source code in vllm/config.py

def get_hidden_size(self) -> int:
    return self.hf_text_config.hidden_size

get_layers_start_end_indices

get_layers_start_end_indices(
    parallel_config: ParallelConfig,
) -> tuple[int, int]

Source code in vllm/config.py

def get_layers_start_end_indices(
        self, parallel_config: "ParallelConfig") -> tuple[int, int]:
    from vllm.distributed.utils import get_pp_indices
    if (self.hf_text_config.model_type == "deepseek_mtp"
            or self.hf_config.model_type == "mimo_mtp"):
        total_num_hidden_layers = getattr(self.hf_text_config,
                                          "num_nextn_predict_layers", 0)
    else:
        total_num_hidden_layers = getattr(self.hf_text_config,
                                          "num_hidden_layers", 0)
    # the layout order is: DP x PP x TP
    pp_rank = (parallel_config.rank // parallel_config.tensor_parallel_size
               ) % parallel_config.pipeline_parallel_size
    pp_size = parallel_config.pipeline_parallel_size
    start, end = get_pp_indices(total_num_hidden_layers, pp_rank, pp_size)
    return start, end

get_multimodal_config

get_multimodal_config() -> MultiModalConfig

Get the multimodal configuration of the model.

Raises:

Type	Description
ValueError	If the model is not multimodal.

Source code in vllm/config.py

def get_multimodal_config(self) -> "MultiModalConfig":
    """
    Get the multimodal configuration of the model.

    Raises:
        ValueError: If the model is not multimodal.
    """
    if self.multimodal_config is None:
        raise ValueError("The model is not multimodal.")

    return self.multimodal_config

get_num_attention_heads

get_num_attention_heads(
    parallel_config: ParallelConfig,
) -> int

Source code in vllm/config.py

def get_num_attention_heads(self,
                            parallel_config: "ParallelConfig") -> int:
    num_heads = getattr(self.hf_text_config, "num_attention_heads", 0)
    return num_heads // parallel_config.tensor_parallel_size

get_num_kv_heads

get_num_kv_heads(parallel_config: ParallelConfig) -> int

Returns the number of KV heads per GPU.

Source code in vllm/config.py

def get_num_kv_heads(self, parallel_config: "ParallelConfig") -> int:
    """Returns the number of KV heads per GPU."""
    if self.use_mla:
        # When using MLA during decode it becomes MQA
        return 1

    total_num_kv_heads = self.get_total_num_kv_heads()
    # If tensor parallelism is used, we divide the number of KV heads by
    # the tensor parallel size. We will replicate the KV heads in the
    # case where the number of KV heads is smaller than the tensor
    # parallel size so each GPU has at least one KV head.
    return max(1,
               total_num_kv_heads // parallel_config.tensor_parallel_size)

get_num_layers

get_num_layers(parallel_config: ParallelConfig) -> int

Source code in vllm/config.py

def get_num_layers(self, parallel_config: "ParallelConfig") -> int:
    start, end = self.get_layers_start_end_indices(parallel_config)
    return end - start

get_num_layers_by_block_type

get_num_layers_by_block_type(
    parallel_config: ParallelConfig,
    block_type: LayerBlockType = attention,
) -> int

Source code in vllm/config.py

def get_num_layers_by_block_type(
    self,
    parallel_config: "ParallelConfig",
    block_type: LayerBlockType = LayerBlockType.attention,
) -> int:
    # This function relies on 'layers_block_type' in hf_config,
    # for w/o this attribute, we will need to have workarounds like so
    attn_block_type = block_type == LayerBlockType.attention
    is_transformer = not self.is_hybrid and \
                        not self.has_noops and \
                        not self.is_attention_free
    start, end = self.get_layers_start_end_indices(parallel_config)

    if is_transformer:
        # Handle the basic case first
        return end - start if attn_block_type else 0
    elif self.is_attention_free:
        # Attention free
        # Note that this code assumes there
        # is only one type of attention-free block type.
        return 0 if attn_block_type else end - start
    elif self.has_noops:
        block_configs = self.hf_config.block_configs
        return sum(not bc.attention.no_op
                   for bc in block_configs[start:end])
    else:
        # Hybrid model Jamba
        layers_block_type_value = getattr(self.hf_config,
                                          "layers_block_type", None)
        if layers_block_type_value is not None:
            if hasattr(self.hf_text_config,
                       "model_type") and (self.hf_text_config.model_type
                                          == "zamba2"):
                if attn_block_type:
                    return sum(t == "hybrid"
                               for t in layers_block_type_value[start:end])
                else:
                    return self.get_num_layers(parallel_config)
            return sum(t == block_type.value
                       for t in layers_block_type_value[start:end])

        # Hybrid model Minimax
        attn_type_list = getattr(self.hf_config, "attn_type_list", None)
        if attn_type_list:
            return sum(t == 1 for t in attn_type_list[start:end])

        if layers_block_type_value is None and attn_type_list is None:
            raise ValueError(
                "The model is an hybrid without a"
                "layers_block_type or an attn_type_list in the hf_config,"
                "cannot determine the num of "
                f"{block_type.value} layers")

        return sum(t == 1 for t in attn_type_list[start:end])

get_sliding_window

get_sliding_window() -> Optional[
    Union[int, list[Optional[int]]]
]

Get the sliding window size, or None if disabled.

Source code in vllm/config.py

def get_sliding_window(self) -> Optional[Union[int, list[Optional[int]]]]:
    """Get the sliding window size, or None if disabled.
    """
    # If user disables sliding window, return None.
    if self.disable_sliding_window:
        return None
    # Otherwise get the value from the hf config.
    return self.get_hf_config_sliding_window()

get_total_num_kv_heads

get_total_num_kv_heads() -> int

Returns the total number of KV heads.

Source code in vllm/config.py

def get_total_num_kv_heads(self) -> int:
    """Returns the total number of KV heads."""
    # For GPTBigCode & Falcon:
    # NOTE: for falcon, when new_decoder_architecture is True, the
    # multi_query flag is ignored and we use n_head_kv for the number of
    # KV heads.
    falcon_model_types = ["falcon", "RefinedWeb", "RefinedWebModel"]
    new_decoder_arch_falcon = (
        self.hf_config.model_type in falcon_model_types
        and getattr(self.hf_config, "new_decoder_architecture", False))
    if not new_decoder_arch_falcon and getattr(self.hf_text_config,
                                               "multi_query", False):
        # Multi-query attention, only one KV head.
        # Currently, tensor parallelism is not supported in this case.
        return 1

    # For DBRX and MPT
    if self.hf_config.model_type == "mpt":
        if "kv_n_heads" in self.hf_config.attn_config:
            return self.hf_config.attn_config["kv_n_heads"]
        return self.hf_config.num_attention_heads
    if self.hf_config.model_type == "dbrx":
        return getattr(self.hf_config.attn_config, "kv_n_heads",
                       self.hf_config.num_attention_heads)

    if self.hf_config.model_type == "nemotron-nas":
        for block in self.hf_config.block_configs:
            if not block.attention.no_op:
                return self.hf_config.num_attention_heads \
                    // block.attention.n_heads_in_group

        raise RuntimeError("Couldn't determine number of kv heads")

    if self.is_attention_free:
        return 0

    attributes = [
        # For Falcon:
        "n_head_kv",
        "num_kv_heads",
        # For LLaMA-2:
        "num_key_value_heads",
        # For ChatGLM:
        "multi_query_group_num",
    ]
    for attr in attributes:
        num_kv_heads = getattr(self.hf_text_config, attr, None)
        if num_kv_heads is not None:
            return num_kv_heads

    # For non-grouped-query attention models, the number of KV heads is
    # equal to the number of attention heads.
    return self.hf_text_config.num_attention_heads

get_vocab_size

get_vocab_size() -> int

Source code in vllm/config.py

def get_vocab_size(self) -> int:
    return self.hf_text_config.vocab_size

maybe_pull_model_tokenizer_for_s3

maybe_pull_model_tokenizer_for_s3(
    model: str, tokenizer: str
) -> None

Pull model/tokenizer from S3 to temporary directory when needed.

Parameters:

Name	Type	Description	Default
model	str	Model name or path	required
tokenizer	str	Tokenizer name or path	required

Source code in vllm/config.py

def maybe_pull_model_tokenizer_for_s3(self, model: str,
                                      tokenizer: str) -> None:
    """Pull model/tokenizer from S3 to temporary directory when needed.

    Args:
        model: Model name or path
        tokenizer: Tokenizer name or path
    """
    if not (is_s3(model) or is_s3(tokenizer)):
        return

    if is_s3(model):
        s3_model = S3Model()
        s3_model.pull_files(model,
                            allow_pattern=["*.model", "*.py", "*.json"])
        self.model_weights = model
        self.model = s3_model.dir

        # If tokenizer is same as model, download to same directory
        if model == tokenizer:
            s3_model.pull_files(
                model, ignore_pattern=["*.pt", "*.safetensors", "*.bin"])
            self.tokenizer = s3_model.dir
            return

    # Only download tokenizer if needed and not already handled
    if is_s3(tokenizer):
        s3_tokenizer = S3Model()
        s3_tokenizer.pull_files(
            model, ignore_pattern=["*.pt", "*.safetensors", "*.bin"])
        self.tokenizer = s3_tokenizer.dir

try_get_generation_config

try_get_generation_config() -> dict[str, Any]

Source code in vllm/config.py

def try_get_generation_config(self) -> dict[str, Any]:
    if self.generation_config in ("auto", "vllm"):
        config = try_get_generation_config(
            self.hf_config_path or self.model,
            trust_remote_code=self.trust_remote_code,
            revision=self.revision,
        )
    else:
        config = try_get_generation_config(
            self.generation_config,
            trust_remote_code=self.trust_remote_code,
        )

    if config is None:
        return {}

    return config.to_diff_dict()

validate_model_config_after

validate_model_config_after() -> ModelConfig

Source code in vllm/config.py

@model_validator(mode="after")
def validate_model_config_after(self: "ModelConfig") -> "ModelConfig":
    if not isinstance(self.tokenizer, str):
        raise ValueError("tokenizer must be a string after __post_init__.")
    if not isinstance(self.max_model_len, int):
        raise ValueError(
            "max_model_len must be an integer after __post_init__.")
    return self

validate_quantization_before classmethod

validate_quantization_before(value: Any) -> Any

Source code in vllm/config.py

@field_validator("quantization", mode="before")
@classmethod
def validate_quantization_before(cls, value: Any) -> Any:
    if isinstance(value, str):
        return value.lower()
    return value

verify_async_output_proc

verify_async_output_proc(
    parallel_config, speculative_config, device_config
) -> None

Source code in vllm/config.py

def verify_async_output_proc(self, parallel_config, speculative_config,
                             device_config) -> None:
    if not self.use_async_output_proc:
        # Nothing to check
        return

    if parallel_config.pipeline_parallel_size > 1:
        self.use_async_output_proc = False
        return

    # Reminder: Please update docs/features/compatibility_matrix.md
    # If the feature combo become valid
    from vllm.platforms import current_platform
    if not current_platform.is_async_output_supported(self.enforce_eager):
        self.use_async_output_proc = False
        return

    if envs.VLLM_USE_RAY_SPMD_WORKER:
        self.use_async_output_proc = False
        return

    # Async postprocessor is not necessary for pooling models
    # since there is no token generation
    if self.runner_type == "pooling":
        self.use_async_output_proc = False

    # Reminder: Please update docs/features/compatibility_matrix.md
    # If the feature combo become valid
    if speculative_config:
        self.use_async_output_proc = False

verify_dual_chunk_attention_config

verify_dual_chunk_attention_config(
    load_config: LoadConfig,
) -> None

Source code in vllm/config.py

def verify_dual_chunk_attention_config(
    self,
    load_config: "LoadConfig",
) -> None:
    if hasattr(self.hf_config, "dual_chunk_attention_config"):
        # Try loading the sparse attention config
        from vllm.model_executor.model_loader.weight_utils import (
            get_sparse_attention_config)
        sparse_attn_config = get_sparse_attention_config(self, load_config)
        if sparse_attn_config:
            self.hf_config.dual_chunk_attention_config[
                "sparse_attention_config"] = sparse_attn_config
            if "sparse_attention_enabled" not in \
                    self.hf_config.dual_chunk_attention_config:
                self.hf_config.dual_chunk_attention_config[
                    "sparse_attention_enabled"] = True

verify_with_parallel_config

verify_with_parallel_config(
    parallel_config: ParallelConfig,
) -> None

Source code in vllm/config.py

def verify_with_parallel_config(
    self,
    parallel_config: "ParallelConfig",
) -> None:

    if parallel_config.distributed_executor_backend == "external_launcher":
        assert self.seed is not None, (
            "Seed must be set when using external launcher backend to "
            "make sure sampling results are the same across workers.")

    total_num_attention_heads = getattr(self.hf_text_config,
                                        "num_attention_heads", 0)
    tensor_parallel_size = parallel_config.tensor_parallel_size
    if total_num_attention_heads % tensor_parallel_size != 0:
        raise ValueError(
            f"Total number of attention heads ({total_num_attention_heads})"
            " must be divisible by tensor parallel size "
            f"({tensor_parallel_size}).")

    if parallel_config.enable_expert_parallel:
        self._verify_with_expert_parallelism()

    pipeline_parallel_size = parallel_config.pipeline_parallel_size
    if pipeline_parallel_size > 1:
        if not self.registry.is_pp_supported_model(self.architectures):
            raise NotImplementedError(
                "Pipeline parallelism is not supported for this model. "
                "Supported models implement the `SupportsPP` interface.")

        if self.use_async_output_proc:
            self.use_async_output_proc = False

ModelImpl

Bases: str, Enum

Source code in vllm/config.py

class ModelImpl(str, enum.Enum):
    AUTO = "auto"
    VLLM = "vllm"
    TRANSFORMERS = "transformers"

AUTO class-attribute instance-attribute

AUTO = 'auto'

TRANSFORMERS class-attribute instance-attribute

TRANSFORMERS = 'transformers'

VLLM class-attribute instance-attribute

VLLM = 'vllm'

MultiModalConfig

Controls the behavior of multimodal models.

Source code in vllm/config.py

@config
@dataclass
class MultiModalConfig:
    """Controls the behavior of multimodal models."""

    limit_per_prompt: dict[str, int] = \
        cast(dict[str, int], get_field(ModelConfig, "limit_mm_per_prompt"))
    """
    The maximum number of input items allowed per prompt for each modality.
    Defaults to 1 (V0) or 999 (V1) for each modality.

    For example, to allow up to 16 images and 2 videos per prompt:
    `{"images": 16, "videos": 2}`
    """

    mm_processor_kwargs: Optional[dict[str, object]] = None
    """
    Overrides for the multi-modal processor obtained from
    `transformers.AutoProcessor.from_pretrained`.

    The available overrides depend on the model that is being run.

    For example, for Phi-3-Vision:
    `{"num_crops": 4}`.
    """

    disable_mm_preprocessor_cache: bool = False
    """
    If `True`, disable caching of the processed multi-modal inputs.
    """

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # this config will not affect the computation graph.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def get_limit_per_prompt(self, modality: str) -> int:
        """
        Get the maximum number of input items allowed per prompt
        for the given modality.
        """
        return self.limit_per_prompt.get(
            modality,
            999 if envs.VLLM_USE_V1 else 1,
        )

disable_mm_preprocessor_cache class-attribute instance-attribute

disable_mm_preprocessor_cache: bool = False

If True, disable caching of the processed multi-modal inputs.

limit_per_prompt class-attribute instance-attribute

limit_per_prompt: dict[str, int] = cast(
    dict[str, int],
    get_field(ModelConfig, "limit_mm_per_prompt"),
)

The maximum number of input items allowed per prompt for each modality. Defaults to 1 (V0) or 999 (V1) for each modality.

For example, to allow up to 16 images and 2 videos per prompt: {"images": 16, "videos": 2}

mm_processor_kwargs class-attribute instance-attribute

mm_processor_kwargs: Optional[dict[str, object]] = None

Overrides for the multi-modal processor obtained from transformers.AutoProcessor.from_pretrained.

The available overrides depend on the model that is being run.

For example, for Phi-3-Vision: {"num_crops": 4}.

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # this config will not affect the computation graph.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

get_limit_per_prompt

get_limit_per_prompt(modality: str) -> int

Get the maximum number of input items allowed per prompt for the given modality.

Source code in vllm/config.py

def get_limit_per_prompt(self, modality: str) -> int:
    """
    Get the maximum number of input items allowed per prompt
    for the given modality.
    """
    return self.limit_per_prompt.get(
        modality,
        999 if envs.VLLM_USE_V1 else 1,
    )

ObservabilityConfig

Configuration for observability - metrics and tracing.

Source code in vllm/config.py

@config
@dataclass
class ObservabilityConfig:
    """Configuration for observability - metrics and tracing."""

    show_hidden_metrics_for_version: Optional[str] = None
    """Enable deprecated Prometheus metrics that have been hidden since the
    specified version. For example, if a previously deprecated metric has been
    hidden since the v0.7.0 release, you use
    `--show-hidden-metrics-for-version=0.7` as a temporary escape hatch while
    you migrate to new metrics. The metric is likely to be removed completely
    in an upcoming release."""

    @cached_property
    def show_hidden_metrics(self) -> bool:
        """Check if the hidden metrics should be shown."""
        if self.show_hidden_metrics_for_version is None:
            return False
        return version._prev_minor_version_was(
            self.show_hidden_metrics_for_version)

    otlp_traces_endpoint: Optional[str] = None
    """Target URL to which OpenTelemetry traces will be sent."""

    collect_detailed_traces: Optional[list[DetailedTraceModules]] = None
    """It makes sense to set this only if `--otlp-traces-endpoint` is set. If
    set, it will collect detailed traces for the specified modules. This
    involves use of possibly costly and or blocking operations and hence might
    have a performance impact.

    Note that collecting detailed timing information for each request can be
    expensive."""

    @cached_property
    def collect_model_forward_time(self) -> bool:
        """Whether to collect model forward time for the request."""
        return (self.collect_detailed_traces is not None
                and ("model" in self.collect_detailed_traces
                     or "all" in self.collect_detailed_traces))

    @cached_property
    def collect_model_execute_time(self) -> bool:
        """Whether to collect model execute time for the request."""
        return (self.collect_detailed_traces is not None
                and ("worker" in self.collect_detailed_traces
                     or "all" in self.collect_detailed_traces))

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # this config will not affect the computation graph.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self):
        if (self.collect_detailed_traces is not None
                and len(self.collect_detailed_traces) == 1
                and "," in self.collect_detailed_traces[0]):
            self._parse_collect_detailed_traces()

        if not is_otel_available() and self.otlp_traces_endpoint is not None:
            raise ValueError(
                "OpenTelemetry is not available. Unable to configure "
                "'otlp_traces_endpoint'. Ensure OpenTelemetry packages are "
                f"installed. Original error:\n{otel_import_error_traceback}")

    def _parse_collect_detailed_traces(self):
        assert isinstance(self.collect_detailed_traces, list)
        self.collect_detailed_traces = cast(
            list[DetailedTraceModules],
            self.collect_detailed_traces[0].split(","))

collect_detailed_traces class-attribute instance-attribute

collect_detailed_traces: Optional[
    list[DetailedTraceModules]
] = None

It makes sense to set this only if --otlp-traces-endpoint is set. If set, it will collect detailed traces for the specified modules. This involves use of possibly costly and or blocking operations and hence might have a performance impact.

Note that collecting detailed timing information for each request can be expensive.

collect_model_execute_time cached property

collect_model_execute_time: bool

Whether to collect model execute time for the request.

collect_model_forward_time cached property

collect_model_forward_time: bool

Whether to collect model forward time for the request.

otlp_traces_endpoint class-attribute instance-attribute

otlp_traces_endpoint: Optional[str] = None

Target URL to which OpenTelemetry traces will be sent.

show_hidden_metrics cached property

show_hidden_metrics: bool

Check if the hidden metrics should be shown.

show_hidden_metrics_for_version class-attribute instance-attribute

show_hidden_metrics_for_version: Optional[str] = None

Enable deprecated Prometheus metrics that have been hidden since the specified version. For example, if a previously deprecated metric has been hidden since the v0.7.0 release, you use --show-hidden-metrics-for-version=0.7 as a temporary escape hatch while you migrate to new metrics. The metric is likely to be removed completely in an upcoming release.

__post_init__

__post_init__()

Source code in vllm/config.py

def __post_init__(self):
    if (self.collect_detailed_traces is not None
            and len(self.collect_detailed_traces) == 1
            and "," in self.collect_detailed_traces[0]):
        self._parse_collect_detailed_traces()

    if not is_otel_available() and self.otlp_traces_endpoint is not None:
        raise ValueError(
            "OpenTelemetry is not available. Unable to configure "
            "'otlp_traces_endpoint'. Ensure OpenTelemetry packages are "
            f"installed. Original error:\n{otel_import_error_traceback}")

_parse_collect_detailed_traces

_parse_collect_detailed_traces()

Source code in vllm/config.py

def _parse_collect_detailed_traces(self):
    assert isinstance(self.collect_detailed_traces, list)
    self.collect_detailed_traces = cast(
        list[DetailedTraceModules],
        self.collect_detailed_traces[0].split(","))

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # this config will not affect the computation graph.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

ParallelConfig

Configuration for the distributed execution.

Source code in vllm/config.py

@config
@dataclass
class ParallelConfig:
    """Configuration for the distributed execution."""

    pipeline_parallel_size: int = 1
    """Number of pipeline parallel groups."""
    tensor_parallel_size: int = 1
    """Number of tensor parallel groups."""
    data_parallel_size: int = 1
    """Number of data parallel groups. MoE layers will be sharded according to
    the product of the tensor parallel size and data parallel size."""
    data_parallel_size_local: int = 1
    """Number of local data parallel groups."""
    data_parallel_rank: int = 0
    """Rank of the data parallel group."""
    data_parallel_rank_local: Optional[int] = None
    """Local rank of the data parallel group,
    set only in SPMD mode."""
    data_parallel_master_ip: str = "127.0.0.1"
    """IP of the data parallel master."""
    data_parallel_rpc_port: int = 29550
    """Port for data parallel messaging."""
    data_parallel_master_port: int = 29500
    """Port of the data parallel master."""
    data_parallel_backend: str = "mp"
    """Backend to use for data parallel, either "mp" or "ray"."""
    enable_expert_parallel: bool = False
    """Use expert parallelism instead of tensor parallelism for MoE layers."""
    max_parallel_loading_workers: Optional[int] = None
    """Maximum number of parallel loading workers when loading model
    sequentially in multiple batches. To avoid RAM OOM when using tensor
    parallel and large models."""

    disable_custom_all_reduce: bool = False
    """Disable the custom all-reduce kernel and fall back to NCCL."""

    tokenizer_pool_config: Optional[TokenizerPoolConfig] = None
    """This parameter is deprecated and will be removed in a future release.
    Please remove it from your configs"""

    ray_workers_use_nsight: bool = False
    """Whether to profile Ray workers with nsight, see https://docs.ray.io/en/latest/ray-observability/user-guides/profiling.html#profiling-nsight-profiler."""

    placement_group: Optional["PlacementGroup"] = None
    """ray distributed model workers placement group."""

    distributed_executor_backend: Optional[Union[DistributedExecutorBackend,
                                                 type["ExecutorBase"]]] = None
    """Backend to use for distributed model
    workers, either "ray" or "mp" (multiprocessing). If the product
    of pipeline_parallel_size and tensor_parallel_size is less than
    or equal to the number of GPUs available, "mp" will be used to
    keep processing on a single host. Otherwise, this will default
    to "ray" if Ray is installed and fail otherwise. Note that tpu
    and hpu only support Ray for distributed inference."""

    worker_cls: str = "auto"
    """The full name of the worker class to use. If "auto", the worker class
    will be determined based on the platform."""
    sd_worker_cls: str = "auto"
    """The full name of the worker class to use for speculative decoding.
    If "auto", the worker class will be determined based on the platform."""
    worker_extension_cls: str = ""
    """The full name of the worker extension class to use. The worker extension
    class is dynamically inherited by the worker class. This is used to inject
    new attributes and methods to the worker class for use in collective_rpc
    calls."""

    world_size: int = field(init=False)
    """world_size is TPxPP, it affects the number of workers we create."""

    rank: int = 0
    """Global rank in distributed setup."""

    enable_multimodal_encoder_data_parallel: bool = False
    """ Use data parallelism instead of tensor parallelism for vision encoder.
    Only support LLama4 for now"""

    @property
    def world_size_across_dp(self) -> int:
        """world_size_across_dp is TPxPPxDP, it is the size of the world
        including data parallelism."""
        return self.world_size * self.data_parallel_size

    def get_next_dp_init_port(self) -> int:
        """
        We might need to initialize process groups in multiple
        processes that is related to data parallelism,
        e.g. both in the worker and in the engine, which
        can live in different processes. To avoid port conflicts, we
        increment the port number each time we need to initialize a
        new process group related to data parallelism.
        """
        answer = self.data_parallel_master_port
        self.data_parallel_master_port += 1
        return answer

    def stateless_init_dp_group(self) -> "ProcessGroup":
        from vllm.distributed.utils import (
            stateless_init_torch_distributed_process_group)

        # use gloo since the engine process might not have cuda device
        dp_group = stateless_init_torch_distributed_process_group(
            self.data_parallel_master_ip,
            self.get_next_dp_init_port(),
            self.data_parallel_rank,
            self.data_parallel_size,
            backend="gloo")

        return dp_group

    @staticmethod
    def has_unfinished_dp(dp_group: "ProcessGroup",
                          has_unfinished: bool) -> bool:
        tensor = torch.tensor([has_unfinished],
                              dtype=torch.int32,
                              device="cpu")
        # dp rank 0: has_unfinished_seqs=True
        # dp rank 1: has_unfinished_seqs=False
        # aggregated: has_unfinished_seqs=True
        # so this is an OR operation, i.e. MAX in integers
        torch.distributed.all_reduce(tensor, op=ReduceOp.MAX, group=dp_group)
        aggregated_has_unfinished = bool(tensor.item())
        return aggregated_has_unfinished

    def compute_hash(self):
        """
        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        factors: list[Any] = []
        factors.append(self.pipeline_parallel_size)
        factors.append(self.tensor_parallel_size)
        factors.append(self.enable_expert_parallel)
        factors.append(self.data_parallel_size)
        factors.append(envs.VLLM_ALL2ALL_BACKEND)
        return hashlib.sha256(str(factors).encode()).hexdigest()

    def __post_init__(self) -> None:
        self.world_size = self.pipeline_parallel_size * \
            self.tensor_parallel_size

        if self.data_parallel_size_local > self.data_parallel_size:
            raise ValueError(
                f"data_parallel_size_local ({self.data_parallel_size_local}) "
                f"must be <= data_parallel_size ({self.data_parallel_size})")

        if self.data_parallel_size > 1 or self.data_parallel_size_local == 0:
            # Data parallel was specified in the engine args.
            self.data_parallel_master_port = get_open_port()
        else:
            # Otherwise fall back to env vars (e.g. for offline SPMD case).
            self.data_parallel_size = envs.VLLM_DP_SIZE
            self.data_parallel_rank = envs.VLLM_DP_RANK
            self.data_parallel_rank_local = envs.VLLM_DP_RANK_LOCAL
            self.data_parallel_master_ip = envs.VLLM_DP_MASTER_IP
            self.data_parallel_master_port = envs.VLLM_DP_MASTER_PORT

        if self.distributed_executor_backend == "external_launcher":
            import os
            os.environ["VLLM_ENABLE_V1_MULTIPROCESSING"] = "0"
            logger.info("Disabling V1 multiprocessing for external launcher.")

        if self.distributed_executor_backend is None and self.world_size > 1:
            # We use multiprocessing by default if world_size fits on the
            # current node and we aren't in a ray placement group.

            from vllm.executor import ray_utils
            backend: DistributedExecutorBackend = "mp"
            ray_found = ray_utils.ray_is_available()
            if current_platform.is_neuron():
                # neuron uses single process to control multiple devices
                backend = "uni"
            elif current_platform.is_tpu() and envs.VLLM_XLA_USE_SPMD:
                backend = "uni"
            elif (current_platform.is_cuda()
                  and cuda_device_count_stateless() < self.world_size):
                if not ray_found:
                    raise ValueError("Unable to load Ray which is "
                                     "required for multi-node inference, "
                                     "please install Ray with `pip install "
                                     "ray`.") from ray_utils.ray_import_err
                backend = "ray"
            elif self.data_parallel_backend == "ray":
                logger.info("Using ray distributed inference because "
                            "data_parallel_backend is ray")
                backend = "ray"
            elif ray_found:
                if self.placement_group:
                    backend = "ray"
                else:
                    from ray import is_initialized as ray_is_initialized
                    if ray_is_initialized():
                        from ray.util import get_current_placement_group
                        if get_current_placement_group():
                            backend = "ray"
            self.distributed_executor_backend = backend
            logger.info("Defaulting to use %s for distributed inference",
                        backend)

        if self.distributed_executor_backend is None and self.world_size == 1:
            self.distributed_executor_backend = "uni"

    @property
    def use_ray(self) -> bool:
        return self.distributed_executor_backend == "ray" or (
            isinstance(self.distributed_executor_backend, type)
            and self.distributed_executor_backend.uses_ray)

    @model_validator(mode='after')
    def _verify_args(self) -> Self:
        # Lazy import to avoid circular import
        from vllm.executor.executor_base import ExecutorBase
        from vllm.platforms import current_platform
        if self.distributed_executor_backend not in (
                "ray", "mp", "uni",
                "external_launcher", None) and not (isinstance(
                    self.distributed_executor_backend, type) and issubclass(
                        self.distributed_executor_backend, ExecutorBase)):
            raise ValueError(
                "Unrecognized distributed executor backend "
                f"{self.distributed_executor_backend}. Supported "
                "values are 'ray', 'mp' 'uni', 'external_launcher' or"
                " custom ExecutorBase subclass.")
        if self.use_ray:
            from vllm.executor import ray_utils
            ray_utils.assert_ray_available()

        if not current_platform.use_custom_allreduce():
            self.disable_custom_all_reduce = True
            logger.debug(
                "Disabled the custom all-reduce kernel because it is not "
                "supported on current platform.")
        if self.ray_workers_use_nsight and not self.use_ray:
            raise ValueError("Unable to use nsight profiling unless workers "
                             "run with Ray.")

        return self

data_parallel_backend class-attribute instance-attribute

data_parallel_backend: str = 'mp'

Backend to use for data parallel, either "mp" or "ray".

data_parallel_master_ip class-attribute instance-attribute

data_parallel_master_ip: str = '127.0.0.1'

IP of the data parallel master.

data_parallel_master_port class-attribute instance-attribute

data_parallel_master_port: int = 29500

Port of the data parallel master.

data_parallel_rank class-attribute instance-attribute

data_parallel_rank: int = 0

Rank of the data parallel group.

data_parallel_rank_local class-attribute instance-attribute

data_parallel_rank_local: Optional[int] = None

Local rank of the data parallel group, set only in SPMD mode.

data_parallel_rpc_port class-attribute instance-attribute

data_parallel_rpc_port: int = 29550

Port for data parallel messaging.

data_parallel_size class-attribute instance-attribute

data_parallel_size: int = 1

Number of data parallel groups. MoE layers will be sharded according to the product of the tensor parallel size and data parallel size.

data_parallel_size_local class-attribute instance-attribute

data_parallel_size_local: int = 1

Number of local data parallel groups.

disable_custom_all_reduce class-attribute instance-attribute

disable_custom_all_reduce: bool = False

Disable the custom all-reduce kernel and fall back to NCCL.

distributed_executor_backend class-attribute instance-attribute

distributed_executor_backend: Optional[
    Union[DistributedExecutorBackend, type[ExecutorBase]]
] = None

Backend to use for distributed model workers, either "ray" or "mp" (multiprocessing). If the product of pipeline_parallel_size and tensor_parallel_size is less than or equal to the number of GPUs available, "mp" will be used to keep processing on a single host. Otherwise, this will default to "ray" if Ray is installed and fail otherwise. Note that tpu and hpu only support Ray for distributed inference.

enable_expert_parallel class-attribute instance-attribute

enable_expert_parallel: bool = False

Use expert parallelism instead of tensor parallelism for MoE layers.

enable_multimodal_encoder_data_parallel class-attribute instance-attribute

enable_multimodal_encoder_data_parallel: bool = False

Use data parallelism instead of tensor parallelism for vision encoder. Only support LLama4 for now

max_parallel_loading_workers class-attribute instance-attribute

max_parallel_loading_workers: Optional[int] = None

Maximum number of parallel loading workers when loading model sequentially in multiple batches. To avoid RAM OOM when using tensor parallel and large models.

pipeline_parallel_size class-attribute instance-attribute

pipeline_parallel_size: int = 1

Number of pipeline parallel groups.

placement_group class-attribute instance-attribute

placement_group: Optional[PlacementGroup] = None

ray distributed model workers placement group.

rank class-attribute instance-attribute

rank: int = 0

Global rank in distributed setup.

ray_workers_use_nsight class-attribute instance-attribute

ray_workers_use_nsight: bool = False

Whether to profile Ray workers with nsight, see https://docs.ray.io/en/latest/ray-observability/user-guides/profiling.html#profiling-nsight-profiler.

sd_worker_cls class-attribute instance-attribute

sd_worker_cls: str = 'auto'

The full name of the worker class to use for speculative decoding. If "auto", the worker class will be determined based on the platform.

tensor_parallel_size class-attribute instance-attribute

tensor_parallel_size: int = 1

Number of tensor parallel groups.

tokenizer_pool_config class-attribute instance-attribute

tokenizer_pool_config: Optional[TokenizerPoolConfig] = None

This parameter is deprecated and will be removed in a future release. Please remove it from your configs

use_ray property

use_ray: bool

worker_cls class-attribute instance-attribute

worker_cls: str = 'auto'

The full name of the worker class to use. If "auto", the worker class will be determined based on the platform.

worker_extension_cls class-attribute instance-attribute

worker_extension_cls: str = ''

The full name of the worker extension class to use. The worker extension class is dynamically inherited by the worker class. This is used to inject new attributes and methods to the worker class for use in collective_rpc calls.

world_size class-attribute instance-attribute

world_size: int = field(init=False)

world_size is TPxPP, it affects the number of workers we create.

world_size_across_dp property

world_size_across_dp: int

world_size_across_dp is TPxPPxDP, it is the size of the world including data parallelism.

__post_init__

__post_init__() -> None

Source code in vllm/config.py

def __post_init__(self) -> None:
    self.world_size = self.pipeline_parallel_size * \
        self.tensor_parallel_size

    if self.data_parallel_size_local > self.data_parallel_size:
        raise ValueError(
            f"data_parallel_size_local ({self.data_parallel_size_local}) "
            f"must be <= data_parallel_size ({self.data_parallel_size})")

    if self.data_parallel_size > 1 or self.data_parallel_size_local == 0:
        # Data parallel was specified in the engine args.
        self.data_parallel_master_port = get_open_port()
    else:
        # Otherwise fall back to env vars (e.g. for offline SPMD case).
        self.data_parallel_size = envs.VLLM_DP_SIZE
        self.data_parallel_rank = envs.VLLM_DP_RANK
        self.data_parallel_rank_local = envs.VLLM_DP_RANK_LOCAL
        self.data_parallel_master_ip = envs.VLLM_DP_MASTER_IP
        self.data_parallel_master_port = envs.VLLM_DP_MASTER_PORT

    if self.distributed_executor_backend == "external_launcher":
        import os
        os.environ["VLLM_ENABLE_V1_MULTIPROCESSING"] = "0"
        logger.info("Disabling V1 multiprocessing for external launcher.")

    if self.distributed_executor_backend is None and self.world_size > 1:
        # We use multiprocessing by default if world_size fits on the
        # current node and we aren't in a ray placement group.

        from vllm.executor import ray_utils
        backend: DistributedExecutorBackend = "mp"
        ray_found = ray_utils.ray_is_available()
        if current_platform.is_neuron():
            # neuron uses single process to control multiple devices
            backend = "uni"
        elif current_platform.is_tpu() and envs.VLLM_XLA_USE_SPMD:
            backend = "uni"
        elif (current_platform.is_cuda()
              and cuda_device_count_stateless() < self.world_size):
            if not ray_found:
                raise ValueError("Unable to load Ray which is "
                                 "required for multi-node inference, "
                                 "please install Ray with `pip install "
                                 "ray`.") from ray_utils.ray_import_err
            backend = "ray"
        elif self.data_parallel_backend == "ray":
            logger.info("Using ray distributed inference because "
                        "data_parallel_backend is ray")
            backend = "ray"
        elif ray_found:
            if self.placement_group:
                backend = "ray"
            else:
                from ray import is_initialized as ray_is_initialized
                if ray_is_initialized():
                    from ray.util import get_current_placement_group
                    if get_current_placement_group():
                        backend = "ray"
        self.distributed_executor_backend = backend
        logger.info("Defaulting to use %s for distributed inference",
                    backend)

    if self.distributed_executor_backend is None and self.world_size == 1:
        self.distributed_executor_backend = "uni"

_verify_args

_verify_args() -> Self

Source code in vllm/config.py

@model_validator(mode='after')
def _verify_args(self) -> Self:
    # Lazy import to avoid circular import
    from vllm.executor.executor_base import ExecutorBase
    from vllm.platforms import current_platform
    if self.distributed_executor_backend not in (
            "ray", "mp", "uni",
            "external_launcher", None) and not (isinstance(
                self.distributed_executor_backend, type) and issubclass(
                    self.distributed_executor_backend, ExecutorBase)):
        raise ValueError(
            "Unrecognized distributed executor backend "
            f"{self.distributed_executor_backend}. Supported "
            "values are 'ray', 'mp' 'uni', 'external_launcher' or"
            " custom ExecutorBase subclass.")
    if self.use_ray:
        from vllm.executor import ray_utils
        ray_utils.assert_ray_available()

    if not current_platform.use_custom_allreduce():
        self.disable_custom_all_reduce = True
        logger.debug(
            "Disabled the custom all-reduce kernel because it is not "
            "supported on current platform.")
    if self.ray_workers_use_nsight and not self.use_ray:
        raise ValueError("Unable to use nsight profiling unless workers "
                         "run with Ray.")

    return self

compute_hash

compute_hash()

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self):
    """
    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    factors: list[Any] = []
    factors.append(self.pipeline_parallel_size)
    factors.append(self.tensor_parallel_size)
    factors.append(self.enable_expert_parallel)
    factors.append(self.data_parallel_size)
    factors.append(envs.VLLM_ALL2ALL_BACKEND)
    return hashlib.sha256(str(factors).encode()).hexdigest()

get_next_dp_init_port

get_next_dp_init_port() -> int

We might need to initialize process groups in multiple processes that is related to data parallelism, e.g. both in the worker and in the engine, which can live in different processes. To avoid port conflicts, we increment the port number each time we need to initialize a new process group related to data parallelism.

Source code in vllm/config.py

def get_next_dp_init_port(self) -> int:
    """
    We might need to initialize process groups in multiple
    processes that is related to data parallelism,
    e.g. both in the worker and in the engine, which
    can live in different processes. To avoid port conflicts, we
    increment the port number each time we need to initialize a
    new process group related to data parallelism.
    """
    answer = self.data_parallel_master_port
    self.data_parallel_master_port += 1
    return answer

has_unfinished_dp staticmethod

has_unfinished_dp(
    dp_group: ProcessGroup, has_unfinished: bool
) -> bool

Source code in vllm/config.py

@staticmethod
def has_unfinished_dp(dp_group: "ProcessGroup",
                      has_unfinished: bool) -> bool:
    tensor = torch.tensor([has_unfinished],
                          dtype=torch.int32,
                          device="cpu")
    # dp rank 0: has_unfinished_seqs=True
    # dp rank 1: has_unfinished_seqs=False
    # aggregated: has_unfinished_seqs=True
    # so this is an OR operation, i.e. MAX in integers
    torch.distributed.all_reduce(tensor, op=ReduceOp.MAX, group=dp_group)
    aggregated_has_unfinished = bool(tensor.item())
    return aggregated_has_unfinished

stateless_init_dp_group

stateless_init_dp_group() -> ProcessGroup

Source code in vllm/config.py

def stateless_init_dp_group(self) -> "ProcessGroup":
    from vllm.distributed.utils import (
        stateless_init_torch_distributed_process_group)

    # use gloo since the engine process might not have cuda device
    dp_group = stateless_init_torch_distributed_process_group(
        self.data_parallel_master_ip,
        self.get_next_dp_init_port(),
        self.data_parallel_rank,
        self.data_parallel_size,
        backend="gloo")

    return dp_group

PassConfig

Configuration for custom Inductor passes.

This is separate from general CompilationConfig so that inductor passes don't all have access to full configuration - that would create a cycle as the PassManager is set as a property of config.

Source code in vllm/config.py

@config
@dataclass
class PassConfig:
    """Configuration for custom Inductor passes.

    This is separate from general `CompilationConfig` so that inductor passes
    don't all have access to full configuration - that would create a cycle as
    the `PassManager` is set as a property of config."""

    dump_graph_stages: list[str] = field(default_factory=list)
    """List of stages for which we want to dump the graph. Each pass defines
    its own stages (before, after, maybe in-between)."""
    dump_graph_dir: Path = Path(".")
    """Directory to dump the graphs."""
    enable_fusion: bool = field(default_factory=lambda: not envs.VLLM_USE_V1)
    """Whether to enable the custom fusion (RMSNorm/SiluMul+quant) pass."""
    enable_attn_fusion: bool = False
    """Whether to enable the custom attention+quant fusion pass."""
    enable_noop: bool = field(default_factory=lambda: not envs.VLLM_USE_V1)
    """Whether to enable the custom no-op elimination pass."""
    enable_sequence_parallelism: bool = False
    """Whether to enable sequence parallelism."""
    enable_async_tp: bool = False
    """Whether to enable async TP."""

    # TODO(luka) better pass enabling system.

    def uuid(self):
        """
        Produces a hash unique to the pass configuration.
        Any new fields that affect compilation should be added to the hash.
        Do not include dump_graph_* in the hash - they don't affect
        compilation.
        """
        exclude = {"dump_graph_stages", "dump_graph_dir"}
        dict_ = {k: v for k, v in asdict(self).items() if k not in exclude}
        return InductorPass.hash_dict(dict_)

    def __post_init__(self) -> None:
        if not self.enable_noop:
            if self.enable_fusion:
                logger.warning_once(
                    "Fusion enabled but reshape elimination disabled. "
                    "RMSNorm/SiluMul + quant (fp8) fusion might not work")
            if self.enable_attn_fusion:
                logger.warning_once(
                    "Fusion enabled but reshape elimination disabled. "
                    "Attention + quant (fp8) fusion might not work")

dump_graph_dir class-attribute instance-attribute

dump_graph_dir: Path = Path('.')

Directory to dump the graphs.

dump_graph_stages class-attribute instance-attribute

dump_graph_stages: list[str] = field(default_factory=list)

List of stages for which we want to dump the graph. Each pass defines its own stages (before, after, maybe in-between).

enable_async_tp class-attribute instance-attribute

enable_async_tp: bool = False

Whether to enable async TP.

enable_attn_fusion class-attribute instance-attribute

enable_attn_fusion: bool = False

Whether to enable the custom attention+quant fusion pass.

enable_fusion class-attribute instance-attribute

enable_fusion: bool = field(
    default_factory=lambda: not VLLM_USE_V1
)

Whether to enable the custom fusion (RMSNorm/SiluMul+quant) pass.

enable_noop class-attribute instance-attribute

enable_noop: bool = field(
    default_factory=lambda: not VLLM_USE_V1
)

Whether to enable the custom no-op elimination pass.

enable_sequence_parallelism class-attribute instance-attribute

enable_sequence_parallelism: bool = False

Whether to enable sequence parallelism.

__post_init__

__post_init__() -> None

Source code in vllm/config.py

def __post_init__(self) -> None:
    if not self.enable_noop:
        if self.enable_fusion:
            logger.warning_once(
                "Fusion enabled but reshape elimination disabled. "
                "RMSNorm/SiluMul + quant (fp8) fusion might not work")
        if self.enable_attn_fusion:
            logger.warning_once(
                "Fusion enabled but reshape elimination disabled. "
                "Attention + quant (fp8) fusion might not work")

uuid

uuid()

Produces a hash unique to the pass configuration. Any new fields that affect compilation should be added to the hash. Do not include dump_graph_* in the hash - they don't affect compilation.

Source code in vllm/config.py

def uuid(self):
    """
    Produces a hash unique to the pass configuration.
    Any new fields that affect compilation should be added to the hash.
    Do not include dump_graph_* in the hash - they don't affect
    compilation.
    """
    exclude = {"dump_graph_stages", "dump_graph_dir"}
    dict_ = {k: v for k, v in asdict(self).items() if k not in exclude}
    return InductorPass.hash_dict(dict_)

PoolerConfig

Controls the behavior of output pooling in pooling models.

Source code in vllm/config.py

@config
@dataclass
class PoolerConfig:
    """Controls the behavior of output pooling in pooling models."""

    pooling_type: Optional[str] = None
    """
    The pooling method of the pooling model. This should be a key in
    [`vllm.model_executor.layers.pooler.PoolingType`][].
    """

    normalize: Optional[bool] = None
    """
    Whether to normalize the pooled outputs. Usually, this should be set to
    ``True`` for embedding outputs.
    """

    softmax: Optional[bool] = None
    """
    Whether to apply softmax to the pooled outputs. Usually, this should be set
    to ``True`` for classification outputs.
    """

    step_tag_id: Optional[int] = None
    """
    If set, only the score corresponding to the ``step_tag_id`` in the
    generated sentence should be returned. Otherwise, the scores for all tokens
    are returned.
    """

    returned_token_ids: Optional[list[int]] = None
    """
    A list of indices for the vocabulary dimensions to be extracted,
    such as the token IDs of ``good_token`` and ``bad_token`` in the
    ``math-shepherd-mistral-7b-prm`` model.
    """

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # this config will not affect the computation graph.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

normalize class-attribute instance-attribute

normalize: Optional[bool] = None

Whether to normalize the pooled outputs. Usually, this should be set to True for embedding outputs.

pooling_type class-attribute instance-attribute

pooling_type: Optional[str] = None

The pooling method of the pooling model. This should be a key in vllm.model_executor.layers.pooler.PoolingType.

returned_token_ids class-attribute instance-attribute

returned_token_ids: Optional[list[int]] = None

A list of indices for the vocabulary dimensions to be extracted, such as the token IDs of good_token and bad_token in the math-shepherd-mistral-7b-prm model.

softmax class-attribute instance-attribute

softmax: Optional[bool] = None

Whether to apply softmax to the pooled outputs. Usually, this should be set to True for classification outputs.

step_tag_id class-attribute instance-attribute

step_tag_id: Optional[int] = None

If set, only the score corresponding to the step_tag_id in the generated sentence should be returned. Otherwise, the scores for all tokens are returned.

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # this config will not affect the computation graph.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

PromptAdapterConfig

Configuration for PromptAdapters.

Source code in vllm/config.py

@config
@dataclass(config=ConfigDict(arbitrary_types_allowed=True))
class PromptAdapterConfig:
    """Configuration for PromptAdapters."""

    max_prompt_adapters: int = 1
    """Max number of PromptAdapters in a batch."""
    max_prompt_adapter_token: int = 0
    """Max number of PromptAdapters tokens."""
    max_cpu_prompt_adapters: Optional[int] = None
    """Maximum number of PromptAdapters to store in CPU memory. Must be >= than
    `max_prompt_adapters`."""
    prompt_adapter_dtype: Union[torch.dtype, str] = "auto"
    """Data type for PromptAdapter. If auto, will default to base model dtype.
    """

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # this config will not affect the computation graph.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self):

        if self.max_prompt_adapters < 1:
            raise ValueError(f"max_prompt_adapters "
                             f"({self.max_prompt_adapters}) must be >= 1.")
        if self.max_prompt_adapter_token == 0:
            raise ValueError("max_prompt_adapter_token must be set.")
        if self.max_cpu_prompt_adapters is None:
            self.max_cpu_prompt_adapters = self.max_prompt_adapters

    def verify_with_model_config(self, model_config: ModelConfig):
        if self.prompt_adapter_dtype == "auto":
            self.prompt_adapter_dtype = model_config.dtype
        elif isinstance(self.prompt_adapter_dtype, str):
            self.prompt_adapter_dtype = getattr(torch,
                                                self.prompt_adapter_dtype)

max_cpu_prompt_adapters class-attribute instance-attribute

max_cpu_prompt_adapters: Optional[int] = None

Maximum number of PromptAdapters to store in CPU memory. Must be >= than max_prompt_adapters.

max_prompt_adapter_token class-attribute instance-attribute

max_prompt_adapter_token: int = 0

Max number of PromptAdapters tokens.

max_prompt_adapters class-attribute instance-attribute

max_prompt_adapters: int = 1

Max number of PromptAdapters in a batch.

prompt_adapter_dtype class-attribute instance-attribute

prompt_adapter_dtype: Union[dtype, str] = 'auto'

Data type for PromptAdapter. If auto, will default to base model dtype.

__post_init__

__post_init__()

Source code in vllm/config.py

def __post_init__(self):

    if self.max_prompt_adapters < 1:
        raise ValueError(f"max_prompt_adapters "
                         f"({self.max_prompt_adapters}) must be >= 1.")
    if self.max_prompt_adapter_token == 0:
        raise ValueError("max_prompt_adapter_token must be set.")
    if self.max_cpu_prompt_adapters is None:
        self.max_cpu_prompt_adapters = self.max_prompt_adapters

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # this config will not affect the computation graph.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

verify_with_model_config

verify_with_model_config(model_config: ModelConfig)

Source code in vllm/config.py

def verify_with_model_config(self, model_config: ModelConfig):
    if self.prompt_adapter_dtype == "auto":
        self.prompt_adapter_dtype = model_config.dtype
    elif isinstance(self.prompt_adapter_dtype, str):
        self.prompt_adapter_dtype = getattr(torch,
                                            self.prompt_adapter_dtype)

SchedulerConfig

Scheduler configuration.

Source code in vllm/config.py

@config
@dataclass
class SchedulerConfig:
    """Scheduler configuration."""

    runner_type: RunnerType = "generate"
    """The runner type to launch for the model."""

    max_num_batched_tokens: SkipValidation[int] = None  # type: ignore
    """Maximum number of tokens to be processed in a single iteration.

    This config has no static default. If left unspecified by the user, it will
    be set in `EngineArgs.create_engine_config` based on the usage context."""

    max_num_seqs: SkipValidation[int] = None  # type: ignore
    """Maximum number of sequences to be processed in a single iteration.

    This config has no static default. If left unspecified by the user, it will
    be set in `EngineArgs.create_engine_config` based on the usage context."""

    max_model_len: SkipValidation[int] = None  # type: ignore
    """Maximum length of a sequence (including prompt and generated text). This
    is primarily set in `ModelConfig` and that value should be manually
    duplicated here."""

    max_num_partial_prefills: int = 1
    """For chunked prefill, the maximum number of sequences that can be
    partially prefilled concurrently."""

    max_long_partial_prefills: int = 1
    """For chunked prefill, the maximum number of prompts longer than
    long_prefill_token_threshold that will be prefilled concurrently. Setting
    this less than max_num_partial_prefills will allow shorter prompts to jump
    the queue in front of longer prompts in some cases, improving latency."""

    long_prefill_token_threshold: int = 0
    """For chunked prefill, a request is considered long if the prompt is
    longer than this number of tokens."""

    num_lookahead_slots: int = 0
    """The number of slots to allocate per sequence per
    step, beyond the known token ids. This is used in speculative
    decoding to store KV activations of tokens which may or may not be
    accepted.

    NOTE: This will be replaced by speculative config in the future; it is
    present to enable correctness tests until then."""

    cuda_graph_sizes: list[int] = field(default_factory=lambda: [512])
    """Cuda graph capture sizes, default is 512.
    1. if one value is provided, then the capture list would follow the
    pattern: [1, 2, 4] + [i for i in range(8, cuda_graph_sizes + 1, 8)]
    2. more than one value (e.g. 1 2 128) is provided, then the capture list
    will follow the provided list."""

    delay_factor: float = 0.0
    """Apply a delay (of delay factor multiplied by previous
    prompt latency) before scheduling next prompt."""

    enable_chunked_prefill: SkipValidation[bool] = None  # type: ignore
    """If True, prefill requests can be chunked based
    on the remaining max_num_batched_tokens."""

    is_multimodal_model: bool = False
    """True if the model is multimodal."""

    # TODO (ywang96): Make this configurable.
    max_num_encoder_input_tokens: int = field(init=False)
    """Multimodal encoder compute budget, only used in V1.

    NOTE: This is not currently configurable. It will be overridden by
    max_num_batched_tokens in case max multimodal embedding size is larger."""

    # TODO (ywang96): Make this configurable.
    encoder_cache_size: int = field(init=False)
    """Multimodal encoder cache size, only used in V1.

    NOTE: This is not currently configurable. It will be overridden by
    max_num_batched_tokens in case max multimodal embedding size is larger."""

    preemption_mode: Optional[PreemptionMode] = None
    """Whether to perform preemption by swapping or
    recomputation. If not specified, we determine the mode as follows:
    We use recomputation by default since it incurs lower overhead than
    swapping. However, when the sequence group has multiple sequences
    (e.g., beam search), recomputation is not currently supported. In
    such a case, we use swapping instead."""

    num_scheduler_steps: int = 1
    """Maximum number of forward steps per scheduler call."""

    multi_step_stream_outputs: bool = True
    """If False, then multi-step will stream outputs at the end of all steps"""

    send_delta_data: bool = False
    """Private API. If used, scheduler sends delta data to
    workers instead of an entire data. It should be enabled only
    when SPMD worker architecture is enabled. I.e.,
    VLLM_USE_RAY_SPMD_WORKER=1"""

    policy: SchedulerPolicy = "fcfs"
    """The scheduling policy to use:\n
    - "fcfs" means first come first served, i.e. requests are handled in order
    of arrival.\n
    - "priority" means requests are handled based on given priority (lower
    value means earlier handling) and time of arrival deciding any ties)."""

    chunked_prefill_enabled: bool = field(init=False)
    """True if chunked prefill is enabled."""

    disable_chunked_mm_input: bool = False
    """If set to true and chunked prefill is enabled, we do not want to
    partially schedule a multimodal item. Only used in V1
    This ensures that if a request has a mixed prompt
    (like text tokens TTTT followed by image tokens IIIIIIIIII) where only
    some image tokens can be scheduled (like TTTTIIIII, leaving IIIII),
    it will be scheduled as TTTT in one step and IIIIIIIIII in the next."""

    # scheduler class or path. "vllm.core.scheduler.Scheduler" (default)
    # or "mod.custom_class".
    scheduler_cls: Union[str, type[object]] = "vllm.core.scheduler.Scheduler"
    """The scheduler class to use. "vllm.core.scheduler.Scheduler" is the
    default scheduler. Can be a class directly or the path to a class of form
    "mod.custom_class"."""

    disable_hybrid_kv_cache_manager: bool = False
    """If set to True, KV cache manager will allocate the same size of KV cache
    for all attention layers even if there are multiple type of attention layers
    like full attention and sliding window attention.
    """

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        # no factors to consider.
        # this config will not affect the computation graph.
        factors: list[Any] = []
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    def __post_init__(self) -> None:
        if self.max_model_len is None:
            self.max_model_len = 8192

        if self.max_num_seqs is None:
            self.max_num_seqs = 128

        if self.max_num_batched_tokens is None:
            if self.enable_chunked_prefill:
                if self.num_scheduler_steps > 1:
                    # Multi-step Chunked-Prefill doesn't allow prompt-chunking
                    # for now. Have max_num_batched_tokens set to max_model_len
                    # so we don't reject sequences on account of a short
                    # max_num_batched_tokens.
                    self.max_num_batched_tokens = max(
                        self.max_model_len, DEFAULT_MAX_NUM_BATCHED_TOKENS)
                else:
                    self.max_num_batched_tokens = (
                        DEFAULT_MAX_NUM_BATCHED_TOKENS)
            else:
                # If max_model_len is too short, use
                # DEFAULT_MAX_NUM_BATCHED_TOKENS as the default value
                # for higher throughput.
                self.max_num_batched_tokens = max(
                    self.max_model_len, DEFAULT_MAX_NUM_BATCHED_TOKENS)

            if self.runner_type == "pooling":
                # Choose specific value for higher throughput
                self.max_num_batched_tokens = max(
                    self.max_num_batched_tokens,
                    POOLING_MODEL_MAX_NUM_BATCHED_TOKENS,
                )
            if self.is_multimodal_model:
                # The value needs to be at least the number of multimodal tokens
                self.max_num_batched_tokens = max(
                    self.max_num_batched_tokens,
                    MULTIMODAL_MODEL_MAX_NUM_BATCHED_TOKENS,
                )

            # When using default settings,
            # Ensure max_num_batched_tokens does not exceed model limit.
            # Some models (e.g., Whisper) have embeddings tied to max length.
            self.max_num_batched_tokens = min(
                self.max_num_seqs * self.max_model_len,
                self.max_num_batched_tokens)

        self.max_num_encoder_input_tokens = self.max_num_batched_tokens
        self.encoder_cache_size = self.max_num_batched_tokens

        if self.enable_chunked_prefill:
            logger.info(
                "Chunked prefill is enabled with max_num_batched_tokens=%d.",
                self.max_num_batched_tokens)

        self.chunked_prefill_enabled = self.enable_chunked_prefill
        if self.max_num_partial_prefills > 1:
            if self.long_prefill_token_threshold == 0:
                self.long_prefill_token_threshold = int(self.max_model_len *
                                                        0.04)

            logger.info(
                "Concurrent partial prefills enabled with "
                "max_num_partial_prefills=%d, max_long_partial_prefills=%d, "
                "long_prefill_token_threshold=%d",
                self.max_num_partial_prefills, self.max_long_partial_prefills,
                self.long_prefill_token_threshold)

    @model_validator(mode='after')
    def _verify_args(self) -> Self:
        if (self.max_num_batched_tokens < self.max_model_len
                and not self.chunked_prefill_enabled):
            raise ValueError(
                f"max_num_batched_tokens ({self.max_num_batched_tokens}) is "
                f"smaller than max_model_len ({self.max_model_len}). "
                "This effectively limits the maximum sequence length to "
                "max_num_batched_tokens and makes vLLM reject longer "
                "sequences. Please increase max_num_batched_tokens or "
                "decrease max_model_len.")

        if self.max_num_batched_tokens < self.max_num_seqs:
            raise ValueError(
                f"max_num_batched_tokens ({self.max_num_batched_tokens}) must "
                "be greater than or equal to max_num_seqs "
                f"({self.max_num_seqs}).")

        if self.max_num_batched_tokens > self.max_num_seqs * self.max_model_len:
            logger.warning(
                "max_num_batched_tokens (%d) exceeds max_num_seqs"
                "* max_model_len (%d). This may lead to unexpected behavior.",
                self.max_num_batched_tokens,
                self.max_num_seqs * self.max_model_len)

        if self.num_lookahead_slots < 0:
            raise ValueError(
                "num_lookahead_slots "
                f"({self.num_lookahead_slots}) must be greater than or "
                "equal to 0.")

        if self.num_scheduler_steps < 1:
            raise ValueError(
                "num_scheduler_steps "
                f"({self.num_scheduler_steps}) must be greater than or "
                "equal to 1.")

        if self.max_num_partial_prefills < 1:
            raise ValueError(
                f"max_num_partial_prefills ({self.max_num_partial_prefills}) "
                "must be greater than or equal to 1.")
        elif self.max_num_partial_prefills > 1:
            if not self.chunked_prefill_enabled:
                raise ValueError("Chunked prefill must be enabled to set "
                                 "max_num_partial_prefills > 1.")

            if self.long_prefill_token_threshold > self.max_model_len:
                raise ValueError(
                    "long_prefill_token_threshold "
                    f"({self.long_prefill_token_threshold}) cannot be greater "
                    f"than the max_model_len ({self.max_model_len}).")

        if (self.max_long_partial_prefills
                < 1) or (self.max_long_partial_prefills
                         > self.max_num_partial_prefills):
            raise ValueError(
                f"max_long_partial_prefills ({self.max_long_partial_prefills}) "
                "must be greater than or equal to 1 and less than or equal to "
                f"max_num_partial_prefills ({self.max_num_partial_prefills}).")

        return self

    @property
    def is_multi_step(self) -> bool:
        return self.num_scheduler_steps > 1

chunked_prefill_enabled class-attribute instance-attribute

chunked_prefill_enabled: bool = field(init=False)

True if chunked prefill is enabled.

cuda_graph_sizes class-attribute instance-attribute

cuda_graph_sizes: list[int] = field(
    default_factory=lambda: [512]
)

Cuda graph capture sizes, default is 512. 1. if one value is provided, then the capture list would follow the pattern: [1, 2, 4] + [i for i in range(8, cuda_graph_sizes + 1, 8)] 2. more than one value (e.g. 1 2 128) is provided, then the capture list will follow the provided list.

delay_factor class-attribute instance-attribute

delay_factor: float = 0.0

Apply a delay (of delay factor multiplied by previous prompt latency) before scheduling next prompt.

disable_chunked_mm_input class-attribute instance-attribute

disable_chunked_mm_input: bool = False

If set to true and chunked prefill is enabled, we do not want to partially schedule a multimodal item. Only used in V1 This ensures that if a request has a mixed prompt (like text tokens TTTT followed by image tokens IIIIIIIIII) where only some image tokens can be scheduled (like TTTTIIIII, leaving IIIII), it will be scheduled as TTTT in one step and IIIIIIIIII in the next.

disable_hybrid_kv_cache_manager class-attribute instance-attribute

disable_hybrid_kv_cache_manager: bool = False

If set to True, KV cache manager will allocate the same size of KV cache for all attention layers even if there are multiple type of attention layers like full attention and sliding window attention.

enable_chunked_prefill class-attribute instance-attribute

enable_chunked_prefill: SkipValidation[bool] = None

If True, prefill requests can be chunked based on the remaining max_num_batched_tokens.

encoder_cache_size class-attribute instance-attribute

encoder_cache_size: int = field(init=False)

Multimodal encoder cache size, only used in V1.

NOTE: This is not currently configurable. It will be overridden by max_num_batched_tokens in case max multimodal embedding size is larger.

is_multi_step property

is_multi_step: bool

is_multimodal_model class-attribute instance-attribute

is_multimodal_model: bool = False

True if the model is multimodal.

long_prefill_token_threshold class-attribute instance-attribute

long_prefill_token_threshold: int = 0

For chunked prefill, a request is considered long if the prompt is longer than this number of tokens.

max_long_partial_prefills class-attribute instance-attribute

max_long_partial_prefills: int = 1

For chunked prefill, the maximum number of prompts longer than long_prefill_token_threshold that will be prefilled concurrently. Setting this less than max_num_partial_prefills will allow shorter prompts to jump the queue in front of longer prompts in some cases, improving latency.

max_model_len class-attribute instance-attribute

max_model_len: SkipValidation[int] = None

Maximum length of a sequence (including prompt and generated text). This is primarily set in ModelConfig and that value should be manually duplicated here.

max_num_batched_tokens class-attribute instance-attribute

max_num_batched_tokens: SkipValidation[int] = None

Maximum number of tokens to be processed in a single iteration.

This config has no static default. If left unspecified by the user, it will be set in EngineArgs.create_engine_config based on the usage context.

max_num_encoder_input_tokens class-attribute instance-attribute

max_num_encoder_input_tokens: int = field(init=False)

Multimodal encoder compute budget, only used in V1.

NOTE: This is not currently configurable. It will be overridden by max_num_batched_tokens in case max multimodal embedding size is larger.

max_num_partial_prefills class-attribute instance-attribute

max_num_partial_prefills: int = 1

For chunked prefill, the maximum number of sequences that can be partially prefilled concurrently.

max_num_seqs class-attribute instance-attribute

max_num_seqs: SkipValidation[int] = None

Maximum number of sequences to be processed in a single iteration.

This config has no static default. If left unspecified by the user, it will be set in EngineArgs.create_engine_config based on the usage context.

multi_step_stream_outputs class-attribute instance-attribute

multi_step_stream_outputs: bool = True

If False, then multi-step will stream outputs at the end of all steps

num_lookahead_slots class-attribute instance-attribute

num_lookahead_slots: int = 0

The number of slots to allocate per sequence per step, beyond the known token ids. This is used in speculative decoding to store KV activations of tokens which may or may not be accepted.

NOTE: This will be replaced by speculative config in the future; it is present to enable correctness tests until then.

num_scheduler_steps class-attribute instance-attribute

num_scheduler_steps: int = 1

Maximum number of forward steps per scheduler call.

policy class-attribute instance-attribute

policy: SchedulerPolicy = 'fcfs'

The scheduling policy to use:

• "fcfs" means first come first served, i.e. requests are handled in order of arrival.

• "priority" means requests are handled based on given priority (lower value means earlier handling) and time of arrival deciding any ties).

preemption_mode class-attribute instance-attribute

preemption_mode: Optional[PreemptionMode] = None

Whether to perform preemption by swapping or recomputation. If not specified, we determine the mode as follows: We use recomputation by default since it incurs lower overhead than swapping. However, when the sequence group has multiple sequences (e.g., beam search), recomputation is not currently supported. In such a case, we use swapping instead.

runner_type class-attribute instance-attribute

runner_type: RunnerType = 'generate'

The runner type to launch for the model.

scheduler_cls class-attribute instance-attribute

scheduler_cls: Union[str, type[object]] = (
    "vllm.core.scheduler.Scheduler"
)

The scheduler class to use. "vllm.core.scheduler.Scheduler" is the default scheduler. Can be a class directly or the path to a class of form "mod.custom_class".

send_delta_data class-attribute instance-attribute

send_delta_data: bool = False

Private API. If used, scheduler sends delta data to workers instead of an entire data. It should be enabled only when SPMD worker architecture is enabled. I.e., VLLM_USE_RAY_SPMD_WORKER=1

__post_init__

__post_init__() -> None

Source code in vllm/config.py

def __post_init__(self) -> None:
    if self.max_model_len is None:
        self.max_model_len = 8192

    if self.max_num_seqs is None:
        self.max_num_seqs = 128

    if self.max_num_batched_tokens is None:
        if self.enable_chunked_prefill:
            if self.num_scheduler_steps > 1:
                # Multi-step Chunked-Prefill doesn't allow prompt-chunking
                # for now. Have max_num_batched_tokens set to max_model_len
                # so we don't reject sequences on account of a short
                # max_num_batched_tokens.
                self.max_num_batched_tokens = max(
                    self.max_model_len, DEFAULT_MAX_NUM_BATCHED_TOKENS)
            else:
                self.max_num_batched_tokens = (
                    DEFAULT_MAX_NUM_BATCHED_TOKENS)
        else:
            # If max_model_len is too short, use
            # DEFAULT_MAX_NUM_BATCHED_TOKENS as the default value
            # for higher throughput.
            self.max_num_batched_tokens = max(
                self.max_model_len, DEFAULT_MAX_NUM_BATCHED_TOKENS)

        if self.runner_type == "pooling":
            # Choose specific value for higher throughput
            self.max_num_batched_tokens = max(
                self.max_num_batched_tokens,
                POOLING_MODEL_MAX_NUM_BATCHED_TOKENS,
            )
        if self.is_multimodal_model:
            # The value needs to be at least the number of multimodal tokens
            self.max_num_batched_tokens = max(
                self.max_num_batched_tokens,
                MULTIMODAL_MODEL_MAX_NUM_BATCHED_TOKENS,
            )

        # When using default settings,
        # Ensure max_num_batched_tokens does not exceed model limit.
        # Some models (e.g., Whisper) have embeddings tied to max length.
        self.max_num_batched_tokens = min(
            self.max_num_seqs * self.max_model_len,
            self.max_num_batched_tokens)

    self.max_num_encoder_input_tokens = self.max_num_batched_tokens
    self.encoder_cache_size = self.max_num_batched_tokens

    if self.enable_chunked_prefill:
        logger.info(
            "Chunked prefill is enabled with max_num_batched_tokens=%d.",
            self.max_num_batched_tokens)

    self.chunked_prefill_enabled = self.enable_chunked_prefill
    if self.max_num_partial_prefills > 1:
        if self.long_prefill_token_threshold == 0:
            self.long_prefill_token_threshold = int(self.max_model_len *
                                                    0.04)

        logger.info(
            "Concurrent partial prefills enabled with "
            "max_num_partial_prefills=%d, max_long_partial_prefills=%d, "
            "long_prefill_token_threshold=%d",
            self.max_num_partial_prefills, self.max_long_partial_prefills,
            self.long_prefill_token_threshold)

_verify_args

_verify_args() -> Self

Source code in vllm/config.py

@model_validator(mode='after')
def _verify_args(self) -> Self:
    if (self.max_num_batched_tokens < self.max_model_len
            and not self.chunked_prefill_enabled):
        raise ValueError(
            f"max_num_batched_tokens ({self.max_num_batched_tokens}) is "
            f"smaller than max_model_len ({self.max_model_len}). "
            "This effectively limits the maximum sequence length to "
            "max_num_batched_tokens and makes vLLM reject longer "
            "sequences. Please increase max_num_batched_tokens or "
            "decrease max_model_len.")

    if self.max_num_batched_tokens < self.max_num_seqs:
        raise ValueError(
            f"max_num_batched_tokens ({self.max_num_batched_tokens}) must "
            "be greater than or equal to max_num_seqs "
            f"({self.max_num_seqs}).")

    if self.max_num_batched_tokens > self.max_num_seqs * self.max_model_len:
        logger.warning(
            "max_num_batched_tokens (%d) exceeds max_num_seqs"
            "* max_model_len (%d). This may lead to unexpected behavior.",
            self.max_num_batched_tokens,
            self.max_num_seqs * self.max_model_len)

    if self.num_lookahead_slots < 0:
        raise ValueError(
            "num_lookahead_slots "
            f"({self.num_lookahead_slots}) must be greater than or "
            "equal to 0.")

    if self.num_scheduler_steps < 1:
        raise ValueError(
            "num_scheduler_steps "
            f"({self.num_scheduler_steps}) must be greater than or "
            "equal to 1.")

    if self.max_num_partial_prefills < 1:
        raise ValueError(
            f"max_num_partial_prefills ({self.max_num_partial_prefills}) "
            "must be greater than or equal to 1.")
    elif self.max_num_partial_prefills > 1:
        if not self.chunked_prefill_enabled:
            raise ValueError("Chunked prefill must be enabled to set "
                             "max_num_partial_prefills > 1.")

        if self.long_prefill_token_threshold > self.max_model_len:
            raise ValueError(
                "long_prefill_token_threshold "
                f"({self.long_prefill_token_threshold}) cannot be greater "
                f"than the max_model_len ({self.max_model_len}).")

    if (self.max_long_partial_prefills
            < 1) or (self.max_long_partial_prefills
                     > self.max_num_partial_prefills):
        raise ValueError(
            f"max_long_partial_prefills ({self.max_long_partial_prefills}) "
            "must be greater than or equal to 1 and less than or equal to "
            f"max_num_partial_prefills ({self.max_num_partial_prefills}).")

    return self

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    # no factors to consider.
    # this config will not affect the computation graph.
    factors: list[Any] = []
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

SpeculativeConfig

Configuration for speculative decoding.

Source code in vllm/config.py

@config
@dataclass
class SpeculativeConfig:
    """Configuration for speculative decoding."""

    # General speculative decoding control
    num_speculative_tokens: SkipValidation[int] = None  # type: ignore
    """The number of speculative tokens, if provided. It will default to the
    number in the draft model config if present, otherwise, it is required."""
    model: Optional[str] = None
    """The name of the draft model, eagle head, or additional weights, if
    provided."""
    method: Optional[SpeculativeMethod] = None
    """The name of the speculative method to use. If users provide and set the
    `model` param, the speculative method type will be detected automatically
    if possible, if `model` param is not provided, the method name must be
    provided.

    If using `ngram` method, the related configuration `prompt_lookup_max` and
    `prompt_lookup_min` should be considered."""
    acceptance_method: SpeculativeAcceptanceMethod = "rejection_sampler"
    """The method to use for accepting draft tokens:\n
    - "rejection_sampler" maps to `RejectionSampler`.\n
    - "typical_acceptance_sampler" maps to `TypicalAcceptanceSampler`.

    If using `typical_acceptance_sampler`, the related configuration
    `posterior_threshold` and `posterior_alpha` should be considered."""
    draft_tensor_parallel_size: Optional[int] = None
    """The degree of the tensor parallelism for the draft model. Can only be 1
    or the same as the target model's tensor parallel size."""
    disable_logprobs: bool = True
    """If set to True, token log probabilities are not returned during
    speculative decoding. If set to False, token log probabilities are returned
    according to the log probability settings in SamplingParams."""

    # Draft model configuration
    quantization: Optional[QuantizationMethods] = None
    """Quantization method that was used to quantize the draft model weights.
    If `None`, we assume the model weights are not quantized. Note that it only
    takes effect when using the draft model-based speculative method."""
    max_model_len: Optional[int] = None
    """The maximum model length of the draft model. Used when testing the
    ability to skip speculation for some sequences."""
    revision: Optional[str] = None
    """The specific model version to use for the draft model. It can be a
    branch name, a tag name, or a commit id. If unspecified, will use the
    default version."""
    code_revision: Optional[str] = None
    """The specific revision to use for the draft model code on Hugging Face
    Hub. It can be a branch name, a tag name, or a commit id. If unspecified,
    will use the default version."""

    # Advanced control
    disable_mqa_scorer: bool = False
    """Disable the MQA scorer and fall back to batch expansion for scoring
    proposals."""
    disable_by_batch_size: Optional[int] = None
    """Disable speculative decoding for new incoming requests when the number
    of enqueued requests is larger than this value, if provided."""

    # Ngram proposer configuration
    prompt_lookup_max: Optional[int] = None
    """Maximum size of ngram token window when using Ngram proposer, required
    when method is set to ngram."""
    prompt_lookup_min: Optional[int] = None
    """Minimum size of ngram token window when using Ngram proposer, if
    provided. Defaults to 1."""

    # Typical acceptance sampler configuration
    posterior_threshold: Optional[float] = None
    """A threshold value that sets a lower bound on the posterior probability
    of a token in the target model for it to be accepted. This threshold is
    used only when we use the `TypicalAcceptanceSampler` for token acceptance.
    """
    posterior_alpha: Optional[float] = None
    """Scaling factor for entropy-based threshold, applied when using
    `TypicalAcceptanceSampler`."""

    speculative_token_tree: Optional[str] = None
    """Specifies the tree structure for speculative token generation.
    """
    # required configuration params passed from engine
    target_model_config: SkipValidation[ModelConfig] = None  # type: ignore
    """The configuration of the target model."""
    target_parallel_config: SkipValidation[
        ParallelConfig] = None  # type: ignore
    """The parallel configuration for the target model."""
    enable_chunked_prefill: SkipValidation[bool] = None  # type: ignore
    """Whether vLLM is configured to use chunked prefill or not. Used for
    raising an error since it's not yet compatible with speculative decode."""
    disable_log_stats: SkipValidation[bool] = None  # type: ignore
    """Whether to disable the periodic printing of stage times in speculative
    decoding."""

    # params generated in the post-init stage
    draft_model_config: SkipValidation[ModelConfig] = None  # type: ignore
    """The configuration of the draft model initialized internal."""
    draft_parallel_config: SkipValidation[
        ParallelConfig] = None  # type: ignore
    """The parallel configuration for the draft model initialized internal."""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        factors: list[Any] = []
        # Eagle3 affects the computation graph because it returns intermediate
        # hidden states in addition to the final hidden state.
        factors.append(self.method == "eagle3")
        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()
        return hash_str

    @classmethod
    def from_dict(cls, dict_value: dict) -> "SpeculativeConfig":
        """Parse the CLI value for the speculative config."""
        return cls(**dict_value)

    @staticmethod
    def hf_config_override(hf_config: PretrainedConfig) -> PretrainedConfig:
        if hf_config.model_type == "deepseek_v3":
            hf_config.model_type = "deepseek_mtp"
        if hf_config.model_type == "deepseek_mtp":
            n_predict = getattr(hf_config, "num_nextn_predict_layers", None)
            hf_config.update({
                "n_predict": n_predict,
                "architectures": ["DeepSeekMTPModel"]
            })

        if hf_config.architectures[0] == "MiMoForCausalLM":
            hf_config.model_type = "mimo_mtp"
            n_predict = getattr(hf_config, "num_nextn_predict_layers", None)
            hf_config.update({
                "num_hidden_layers": 0,
                "n_predict": n_predict,
                "architectures": ["MiMoMTPModel"]
            })
            return hf_config

        return hf_config

    def __post_init__(self):

        # Note: "method" is a new parameter that helps to extend the
        # configuration of non-model-based proposers, and the "model" parameter
        # will be used to set the draft model, eagle head, or additional weight
        # when needed. If users do not specify "method", the speculative method
        # will be detected automatically if possible. If the speculative method
        # can not be detected, it will be considered as the "draft_model" by
        # default.

        if self.model is None and self.num_speculative_tokens is not None:
            # TODO(Shangming): Refactor mtp configuration logic when supporting
            # mtp acceleration for more models besides deepseek_v3
            if self.target_model_config and \
                (self.target_model_config.hf_text_config.model_type \
                        == "deepseek_v3" or
                    self.target_model_config.hf_text_config.model_type \
                        == "mimo"):
                # use the draft model from the same model:
                self.model = self.target_model_config.model
            elif self.method in ("ngram", "[ngram]"):
                self.model = "ngram"
            else:
                raise ValueError("num_speculative_tokens was provided without "
                                 "speculative model.")

        # Automatically configure the method for ngram when "model" is used
        # instead of "method"
        if self.method is None and (self.model is not None
                                    and self.model in ("ngram", "[ngram]")):
            self.method = "ngram"

        if self.method in ("ngram", "[ngram]"):
            # Unified to "ngram" internally
            self.method = "ngram"
            # Set default values if not provided
            if (self.prompt_lookup_min is None
                    and self.prompt_lookup_max is None):
                # TODO(woosuk): Tune these values. They are arbitrarily chosen.
                self.prompt_lookup_min = 5
                self.prompt_lookup_max = 5
            elif self.prompt_lookup_min is None:
                assert self.prompt_lookup_max is not None
                self.prompt_lookup_min = self.prompt_lookup_max
            elif self.prompt_lookup_max is None:
                assert self.prompt_lookup_min is not None
                self.prompt_lookup_max = self.prompt_lookup_min

            # Validate values
            if self.prompt_lookup_min < 1:
                raise ValueError(
                    f"prompt_lookup_min={self.prompt_lookup_min} must be > 0")
            if self.prompt_lookup_max < 1:
                raise ValueError(
                    f"prompt_lookup_max={self.prompt_lookup_max} must be > 0")
            if self.prompt_lookup_min > self.prompt_lookup_max:
                raise ValueError(
                    f"prompt_lookup_min={self.prompt_lookup_min} must "
                    f"be <= prompt_lookup_max={self.prompt_lookup_max}")

            # TODO: current we still need extract vocab_size from target model
            # config, in future, we may try refactor it out, and set
            # draft related config as None here.
            self.draft_model_config = self.target_model_config
            self.draft_parallel_config = self.target_parallel_config
        else:
            self.prompt_lookup_max = 0
            self.prompt_lookup_min = 0

            if self.model is not None:
                self.draft_model_config = ModelConfig(
                    model=self.model,
                    task="draft",
                    tokenizer=self.target_model_config.tokenizer,
                    tokenizer_mode=self.target_model_config.tokenizer_mode,
                    trust_remote_code=self.target_model_config.
                    trust_remote_code,
                    allowed_local_media_path=self.target_model_config.
                    allowed_local_media_path,
                    dtype=self.target_model_config.dtype,
                    seed=self.target_model_config.seed,
                    revision=self.revision,
                    code_revision=self.code_revision,
                    tokenizer_revision=self.target_model_config.
                    tokenizer_revision,
                    spec_target_max_model_len=self.target_model_config.
                    max_model_len,
                    quantization=self.quantization,
                    enforce_eager=self.target_model_config.enforce_eager,
                    max_seq_len_to_capture=self.target_model_config.
                    max_seq_len_to_capture,
                    max_logprobs=self.target_model_config.max_logprobs,
                    hf_overrides=SpeculativeConfig.hf_config_override,
                )

                # Automatically detect the method
                if self.method in ('eagle', 'eagle3'):
                    pass
                elif "eagle-" in self.draft_model_config.model.lower() or \
                        "eagle3-" in self.draft_model_config.model.lower():
                    self.method = "eagle"
                elif self.draft_model_config.hf_config.model_type == "medusa":
                    self.method = "medusa"
                elif (self.draft_model_config.hf_config.model_type ==
                      "mlp_speculator"):
                    self.method = "mlp_speculator"
                elif (self.draft_model_config.hf_config.model_type ==
                      "deepseek_mtp"):
                    self.method = "deepseek_mtp"
                    if self.num_speculative_tokens > 1:
                        logger.warning(
                                "All Deepseek MTP models only have " \
                                "one layer. Might need some code changes " \
                                "to support multiple layers."
                            )
                else:
                    self.method = "draft_model"

                # Replace hf_config for EAGLE draft_model
                if self.method in ("eagle", "eagle3"):
                    if self.enable_chunked_prefill and not envs.VLLM_USE_V1:
                        raise ValueError(
                            "Chunked prefill and EAGLE are not compatible "
                            "when using V0.")

                    from vllm.transformers_utils.configs.eagle import (
                        EAGLEConfig)
                    if isinstance(self.draft_model_config.hf_config,
                                  EAGLEConfig):
                        pass
                    else:
                        eagle_config = EAGLEConfig(
                            self.draft_model_config.hf_config,
                            method=self.method,
                            model_type="eagle")
                        self.draft_model_config.hf_config = eagle_config

                if (self.num_speculative_tokens is not None
                        and hasattr(self.draft_model_config.hf_config,
                                    "num_lookahead_tokens")):
                    self.draft_model_config.hf_config.num_lookahead_tokens = \
                    self.num_speculative_tokens

                n_predict = getattr(self.draft_model_config.hf_config,
                                    "n_predict", None)
                if n_predict is not None:
                    if self.num_speculative_tokens is None:
                        # Default to max value defined in draft model config.
                        self.num_speculative_tokens = n_predict
                    elif self.num_speculative_tokens > n_predict and \
                            self.num_speculative_tokens % n_predict != 0:
                        # Ensure divisibility for MTP module reuse.
                        raise ValueError(
                            f"num_speculative_tokens:{self.num_speculative_tokens}"
                            f" must be divisible by {n_predict=}")

                self.draft_tensor_parallel_size = \
                    SpeculativeConfig._verify_and_get_draft_tp(
                        self.target_parallel_config,
                        self.draft_tensor_parallel_size,
                        self.draft_model_config.hf_config
                )

                self.draft_model_config.max_model_len = (
                    SpeculativeConfig._maybe_override_draft_max_model_len(
                        self.max_model_len,
                        self.draft_model_config.max_model_len,
                        self.target_model_config.max_model_len,
                    ))

                self.draft_parallel_config = (
                    SpeculativeConfig.create_draft_parallel_config(
                        self.target_parallel_config,
                        self.draft_tensor_parallel_size))

        if self.acceptance_method == "typical_acceptance_sampler":
            if self.posterior_threshold is None:
                self.posterior_threshold = 0.09
            if self.posterior_alpha is None:
                self.posterior_alpha = 0.3

    @staticmethod
    def _maybe_override_draft_max_model_len(
        speculative_max_model_len: Optional[int],
        draft_max_model_len: int,
        target_max_model_len: int,
    ) -> int:
        """Determine the max sequence len for the draft model. This is usually
        the draft_max_model_len, but may be the target_max_model_len if it is
        less than the draft_max_model_len, or may be speculative_max_model_len
        if it is specified.

        This is necessary so that sequences do not exceed the capacity of the
        draft model or the target model.

        speculative_max_model_len is mainly used for testing that sequences can
        skip speculation.
        """

        if speculative_max_model_len is not None:

            if speculative_max_model_len > draft_max_model_len:
                raise ValueError(f"{speculative_max_model_len=} cannot be "
                                 f"larger than {draft_max_model_len=}")

            if speculative_max_model_len > target_max_model_len:
                raise ValueError(f"{speculative_max_model_len=} cannot be "
                                 f"larger than {target_max_model_len=}")

            return speculative_max_model_len

        return min(
            draft_max_model_len,
            target_max_model_len,
        )

    @staticmethod
    def _verify_and_get_draft_tp(
            target_parallel_config: ParallelConfig,
            speculative_draft_tensor_parallel_size: Optional[int],
            draft_hf_config: PretrainedConfig) -> int:
        """
        Verifies and adjusts the tensor parallel size for a draft model
        specified using speculative_draft_tensor_parallel_size.
        """
        # If speculative_draft_tensor_parallel_size is unset then set it
        # appropriately else verify that it is set correctly.
        if speculative_draft_tensor_parallel_size is None:
            if draft_hf_config.model_type == "mlp_speculator":
                speculative_draft_tensor_parallel_size = 1
                if target_parallel_config.tensor_parallel_size > 1:
                    logger.warning(
                        "%s cannot currently be run with tp>1; "
                        "setting speculative_draft_tensor_parallel_size=1",
                        draft_hf_config.model_type)
            else:
                speculative_draft_tensor_parallel_size = \
                    target_parallel_config.tensor_parallel_size
        elif speculative_draft_tensor_parallel_size not in (
                1, target_parallel_config.tensor_parallel_size):
            raise ValueError(
                f"{speculative_draft_tensor_parallel_size=} cannot be "
                f"other value than 1 or target model tensor_parallel_size")
        return speculative_draft_tensor_parallel_size

    @staticmethod
    def create_draft_parallel_config(
        target_parallel_config: ParallelConfig,
        speculative_draft_tensor_parallel_size: int,
    ) -> ParallelConfig:
        """Create a parallel config for use by the draft worker.

        This is mostly a copy of the target parallel config, except the tp_size.
        """
        draft_parallel_config = ParallelConfig(
            pipeline_parallel_size=target_parallel_config.
            pipeline_parallel_size,
            tensor_parallel_size=speculative_draft_tensor_parallel_size,
            distributed_executor_backend=target_parallel_config.
            distributed_executor_backend,
            max_parallel_loading_workers=target_parallel_config.
            max_parallel_loading_workers,
            disable_custom_all_reduce=target_parallel_config.
            disable_custom_all_reduce,
            ray_workers_use_nsight=target_parallel_config.
            ray_workers_use_nsight,
            placement_group=target_parallel_config.placement_group,
        )

        return draft_parallel_config

    @model_validator(mode='after')
    def _verify_args(self) -> Self:
        if self.num_speculative_tokens is None:
            raise ValueError(
                "num_speculative_tokens must be provided with "
                "speculative model unless the draft model config contains an "
                "n_predict parameter.")

        if self.num_speculative_tokens <= 0:
            raise ValueError("Expected num_speculative_tokens to be greater "
                             f"than zero ({self.num_speculative_tokens}).")

        if self.draft_model_config:
            self.draft_model_config.verify_with_parallel_config(
                self.draft_parallel_config)
            # Validate and set draft token acceptance related settings.

        if self.acceptance_method is None:
            raise ValueError("acceptance_method is not set. "
                             "Expected values are rejection_sampler or "
                             "typical_acceptance_sampler.")

        if (self.acceptance_method != 'rejection_sampler'
                and self.acceptance_method != 'typical_acceptance_sampler'):
            raise ValueError(
                "Expected acceptance_method to be either "
                "rejection_sampler or typical_acceptance_sampler. Instead it "
                f"is {self.acceptance_method}")

        if self.acceptance_method == "typical_acceptance_sampler" and (
            (self.posterior_threshold is not None
             and self.posterior_threshold < 0) or
            (self.posterior_alpha is not None and self.posterior_alpha < 0)):
            raise ValueError(
                "Expected the posterior_threshold and posterior_alpha of "
                "typical_acceptance_sampler to be > 0. "
                "Instead found posterior_threshold = "
                f"{self.posterior_threshold} and posterior_alpha = "
                f"{self.posterior_alpha}")

        if (self.disable_by_batch_size is not None
                and self.disable_by_batch_size < 2):
            raise ValueError("Expect the batch size threshold of disabling "
                             "speculative decoding is > 1, but got "
                             f"{self.disable_by_batch_size=}")

        if self.method == "eagle3" and self.target_model_config and \
            "llama" not in self.target_model_config.hf_text_config.model_type:
            raise ValueError(
                "Eagle3 is only supported for Llama models. "
                f"Got {self.target_model_config.hf_text_config.model_type=}")

        return self

    @property
    def num_lookahead_slots(self) -> int:
        """The number of additional slots the scheduler should allocate per
        step, in addition to the slots allocated for each known token.

        This is equal to the number of speculative tokens, as each speculative
        token must be scored.
        """
        return self.num_speculative_tokens

    def use_eagle(self) -> bool:
        return self.method in ("eagle", "eagle3", "deepseek_mtp")

    def __repr__(self) -> str:
        method = self.method
        model = None if method == "ngram" else self.draft_model_config.model
        num_spec_tokens = self.num_speculative_tokens
        return f"SpeculativeConfig({method=}, {model=}, {num_spec_tokens=})"

acceptance_method class-attribute instance-attribute

acceptance_method: SpeculativeAcceptanceMethod = (
    "rejection_sampler"
)

The method to use for accepting draft tokens:

• "rejection_sampler" maps to RejectionSampler.

• "typical_acceptance_sampler" maps to TypicalAcceptanceSampler.

If using typical_acceptance_sampler, the related configuration posterior_threshold and posterior_alpha should be considered.

code_revision class-attribute instance-attribute

code_revision: Optional[str] = None

The specific revision to use for the draft model code on Hugging Face Hub. It can be a branch name, a tag name, or a commit id. If unspecified, will use the default version.

disable_by_batch_size class-attribute instance-attribute

disable_by_batch_size: Optional[int] = None

Disable speculative decoding for new incoming requests when the number of enqueued requests is larger than this value, if provided.

disable_log_stats class-attribute instance-attribute

disable_log_stats: SkipValidation[bool] = None

Whether to disable the periodic printing of stage times in speculative decoding.

disable_logprobs class-attribute instance-attribute

disable_logprobs: bool = True

If set to True, token log probabilities are not returned during speculative decoding. If set to False, token log probabilities are returned according to the log probability settings in SamplingParams.

disable_mqa_scorer class-attribute instance-attribute

disable_mqa_scorer: bool = False

Disable the MQA scorer and fall back to batch expansion for scoring proposals.

draft_model_config class-attribute instance-attribute

draft_model_config: SkipValidation[ModelConfig] = None

The configuration of the draft model initialized internal.

draft_parallel_config class-attribute instance-attribute

draft_parallel_config: SkipValidation[ParallelConfig] = None

The parallel configuration for the draft model initialized internal.

draft_tensor_parallel_size class-attribute instance-attribute

draft_tensor_parallel_size: Optional[int] = None

The degree of the tensor parallelism for the draft model. Can only be 1 or the same as the target model's tensor parallel size.

enable_chunked_prefill class-attribute instance-attribute

enable_chunked_prefill: SkipValidation[bool] = None

Whether vLLM is configured to use chunked prefill or not. Used for raising an error since it's not yet compatible with speculative decode.

max_model_len class-attribute instance-attribute

max_model_len: Optional[int] = None

The maximum model length of the draft model. Used when testing the ability to skip speculation for some sequences.

method class-attribute instance-attribute

method: Optional[SpeculativeMethod] = None

The name of the speculative method to use. If users provide and set the model param, the speculative method type will be detected automatically if possible, if model param is not provided, the method name must be provided.

If using ngram method, the related configuration prompt_lookup_max and prompt_lookup_min should be considered.

model class-attribute instance-attribute

model: Optional[str] = None

The name of the draft model, eagle head, or additional weights, if provided.

num_lookahead_slots property

num_lookahead_slots: int

The number of additional slots the scheduler should allocate per step, in addition to the slots allocated for each known token.

This is equal to the number of speculative tokens, as each speculative token must be scored.

num_speculative_tokens class-attribute instance-attribute

num_speculative_tokens: SkipValidation[int] = None

The number of speculative tokens, if provided. It will default to the number in the draft model config if present, otherwise, it is required.

posterior_alpha class-attribute instance-attribute

posterior_alpha: Optional[float] = None

Scaling factor for entropy-based threshold, applied when using TypicalAcceptanceSampler.

posterior_threshold class-attribute instance-attribute

posterior_threshold: Optional[float] = None

A threshold value that sets a lower bound on the posterior probability of a token in the target model for it to be accepted. This threshold is used only when we use the TypicalAcceptanceSampler for token acceptance.

prompt_lookup_max class-attribute instance-attribute

prompt_lookup_max: Optional[int] = None

Maximum size of ngram token window when using Ngram proposer, required when method is set to ngram.

prompt_lookup_min class-attribute instance-attribute

prompt_lookup_min: Optional[int] = None

Minimum size of ngram token window when using Ngram proposer, if provided. Defaults to 1.

quantization class-attribute instance-attribute

quantization: Optional[QuantizationMethods] = None

Quantization method that was used to quantize the draft model weights. If None, we assume the model weights are not quantized. Note that it only takes effect when using the draft model-based speculative method.

revision class-attribute instance-attribute

revision: Optional[str] = None

The specific model version to use for the draft model. It can be a branch name, a tag name, or a commit id. If unspecified, will use the default version.

speculative_token_tree class-attribute instance-attribute

speculative_token_tree: Optional[str] = None

Specifies the tree structure for speculative token generation.

target_model_config class-attribute instance-attribute

target_model_config: SkipValidation[ModelConfig] = None

The configuration of the target model.

target_parallel_config class-attribute instance-attribute

target_parallel_config: SkipValidation[ParallelConfig] = (
    None
)

The parallel configuration for the target model.

__post_init__

__post_init__()

Source code in vllm/config.py

def __post_init__(self):

    # Note: "method" is a new parameter that helps to extend the
    # configuration of non-model-based proposers, and the "model" parameter
    # will be used to set the draft model, eagle head, or additional weight
    # when needed. If users do not specify "method", the speculative method
    # will be detected automatically if possible. If the speculative method
    # can not be detected, it will be considered as the "draft_model" by
    # default.

    if self.model is None and self.num_speculative_tokens is not None:
        # TODO(Shangming): Refactor mtp configuration logic when supporting
        # mtp acceleration for more models besides deepseek_v3
        if self.target_model_config and \
            (self.target_model_config.hf_text_config.model_type \
                    == "deepseek_v3" or
                self.target_model_config.hf_text_config.model_type \
                    == "mimo"):
            # use the draft model from the same model:
            self.model = self.target_model_config.model
        elif self.method in ("ngram", "[ngram]"):
            self.model = "ngram"
        else:
            raise ValueError("num_speculative_tokens was provided without "
                             "speculative model.")

    # Automatically configure the method for ngram when "model" is used
    # instead of "method"
    if self.method is None and (self.model is not None
                                and self.model in ("ngram", "[ngram]")):
        self.method = "ngram"

    if self.method in ("ngram", "[ngram]"):
        # Unified to "ngram" internally
        self.method = "ngram"
        # Set default values if not provided
        if (self.prompt_lookup_min is None
                and self.prompt_lookup_max is None):
            # TODO(woosuk): Tune these values. They are arbitrarily chosen.
            self.prompt_lookup_min = 5
            self.prompt_lookup_max = 5
        elif self.prompt_lookup_min is None:
            assert self.prompt_lookup_max is not None
            self.prompt_lookup_min = self.prompt_lookup_max
        elif self.prompt_lookup_max is None:
            assert self.prompt_lookup_min is not None
            self.prompt_lookup_max = self.prompt_lookup_min

        # Validate values
        if self.prompt_lookup_min < 1:
            raise ValueError(
                f"prompt_lookup_min={self.prompt_lookup_min} must be > 0")
        if self.prompt_lookup_max < 1:
            raise ValueError(
                f"prompt_lookup_max={self.prompt_lookup_max} must be > 0")
        if self.prompt_lookup_min > self.prompt_lookup_max:
            raise ValueError(
                f"prompt_lookup_min={self.prompt_lookup_min} must "
                f"be <= prompt_lookup_max={self.prompt_lookup_max}")

        # TODO: current we still need extract vocab_size from target model
        # config, in future, we may try refactor it out, and set
        # draft related config as None here.
        self.draft_model_config = self.target_model_config
        self.draft_parallel_config = self.target_parallel_config
    else:
        self.prompt_lookup_max = 0
        self.prompt_lookup_min = 0

        if self.model is not None:
            self.draft_model_config = ModelConfig(
                model=self.model,
                task="draft",
                tokenizer=self.target_model_config.tokenizer,
                tokenizer_mode=self.target_model_config.tokenizer_mode,
                trust_remote_code=self.target_model_config.
                trust_remote_code,
                allowed_local_media_path=self.target_model_config.
                allowed_local_media_path,
                dtype=self.target_model_config.dtype,
                seed=self.target_model_config.seed,
                revision=self.revision,
                code_revision=self.code_revision,
                tokenizer_revision=self.target_model_config.
                tokenizer_revision,
                spec_target_max_model_len=self.target_model_config.
                max_model_len,
                quantization=self.quantization,
                enforce_eager=self.target_model_config.enforce_eager,
                max_seq_len_to_capture=self.target_model_config.
                max_seq_len_to_capture,
                max_logprobs=self.target_model_config.max_logprobs,
                hf_overrides=SpeculativeConfig.hf_config_override,
            )

            # Automatically detect the method
            if self.method in ('eagle', 'eagle3'):
                pass
            elif "eagle-" in self.draft_model_config.model.lower() or \
                    "eagle3-" in self.draft_model_config.model.lower():
                self.method = "eagle"
            elif self.draft_model_config.hf_config.model_type == "medusa":
                self.method = "medusa"
            elif (self.draft_model_config.hf_config.model_type ==
                  "mlp_speculator"):
                self.method = "mlp_speculator"
            elif (self.draft_model_config.hf_config.model_type ==
                  "deepseek_mtp"):
                self.method = "deepseek_mtp"
                if self.num_speculative_tokens > 1:
                    logger.warning(
                            "All Deepseek MTP models only have " \
                            "one layer. Might need some code changes " \
                            "to support multiple layers."
                        )
            else:
                self.method = "draft_model"

            # Replace hf_config for EAGLE draft_model
            if self.method in ("eagle", "eagle3"):
                if self.enable_chunked_prefill and not envs.VLLM_USE_V1:
                    raise ValueError(
                        "Chunked prefill and EAGLE are not compatible "
                        "when using V0.")

                from vllm.transformers_utils.configs.eagle import (
                    EAGLEConfig)
                if isinstance(self.draft_model_config.hf_config,
                              EAGLEConfig):
                    pass
                else:
                    eagle_config = EAGLEConfig(
                        self.draft_model_config.hf_config,
                        method=self.method,
                        model_type="eagle")
                    self.draft_model_config.hf_config = eagle_config

            if (self.num_speculative_tokens is not None
                    and hasattr(self.draft_model_config.hf_config,
                                "num_lookahead_tokens")):
                self.draft_model_config.hf_config.num_lookahead_tokens = \
                self.num_speculative_tokens

            n_predict = getattr(self.draft_model_config.hf_config,
                                "n_predict", None)
            if n_predict is not None:
                if self.num_speculative_tokens is None:
                    # Default to max value defined in draft model config.
                    self.num_speculative_tokens = n_predict
                elif self.num_speculative_tokens > n_predict and \
                        self.num_speculative_tokens % n_predict != 0:
                    # Ensure divisibility for MTP module reuse.
                    raise ValueError(
                        f"num_speculative_tokens:{self.num_speculative_tokens}"
                        f" must be divisible by {n_predict=}")

            self.draft_tensor_parallel_size = \
                SpeculativeConfig._verify_and_get_draft_tp(
                    self.target_parallel_config,
                    self.draft_tensor_parallel_size,
                    self.draft_model_config.hf_config
            )

            self.draft_model_config.max_model_len = (
                SpeculativeConfig._maybe_override_draft_max_model_len(
                    self.max_model_len,
                    self.draft_model_config.max_model_len,
                    self.target_model_config.max_model_len,
                ))

            self.draft_parallel_config = (
                SpeculativeConfig.create_draft_parallel_config(
                    self.target_parallel_config,
                    self.draft_tensor_parallel_size))

    if self.acceptance_method == "typical_acceptance_sampler":
        if self.posterior_threshold is None:
            self.posterior_threshold = 0.09
        if self.posterior_alpha is None:
            self.posterior_alpha = 0.3

__repr__

__repr__() -> str

Source code in vllm/config.py

def __repr__(self) -> str:
    method = self.method
    model = None if method == "ngram" else self.draft_model_config.model
    num_spec_tokens = self.num_speculative_tokens
    return f"SpeculativeConfig({method=}, {model=}, {num_spec_tokens=})"

_maybe_override_draft_max_model_len staticmethod

_maybe_override_draft_max_model_len(
    speculative_max_model_len: Optional[int],
    draft_max_model_len: int,
    target_max_model_len: int,
) -> int

Determine the max sequence len for the draft model. This is usually the draft_max_model_len, but may be the target_max_model_len if it is less than the draft_max_model_len, or may be speculative_max_model_len if it is specified.

This is necessary so that sequences do not exceed the capacity of the draft model or the target model.

speculative_max_model_len is mainly used for testing that sequences can skip speculation.

Source code in vllm/config.py

@staticmethod
def _maybe_override_draft_max_model_len(
    speculative_max_model_len: Optional[int],
    draft_max_model_len: int,
    target_max_model_len: int,
) -> int:
    """Determine the max sequence len for the draft model. This is usually
    the draft_max_model_len, but may be the target_max_model_len if it is
    less than the draft_max_model_len, or may be speculative_max_model_len
    if it is specified.

    This is necessary so that sequences do not exceed the capacity of the
    draft model or the target model.

    speculative_max_model_len is mainly used for testing that sequences can
    skip speculation.
    """

    if speculative_max_model_len is not None:

        if speculative_max_model_len > draft_max_model_len:
            raise ValueError(f"{speculative_max_model_len=} cannot be "
                             f"larger than {draft_max_model_len=}")

        if speculative_max_model_len > target_max_model_len:
            raise ValueError(f"{speculative_max_model_len=} cannot be "
                             f"larger than {target_max_model_len=}")

        return speculative_max_model_len

    return min(
        draft_max_model_len,
        target_max_model_len,
    )

_verify_and_get_draft_tp staticmethod

_verify_and_get_draft_tp(
    target_parallel_config: ParallelConfig,
    speculative_draft_tensor_parallel_size: Optional[int],
    draft_hf_config: PretrainedConfig,
) -> int

Verifies and adjusts the tensor parallel size for a draft model specified using speculative_draft_tensor_parallel_size.

Source code in vllm/config.py

@staticmethod
def _verify_and_get_draft_tp(
        target_parallel_config: ParallelConfig,
        speculative_draft_tensor_parallel_size: Optional[int],
        draft_hf_config: PretrainedConfig) -> int:
    """
    Verifies and adjusts the tensor parallel size for a draft model
    specified using speculative_draft_tensor_parallel_size.
    """
    # If speculative_draft_tensor_parallel_size is unset then set it
    # appropriately else verify that it is set correctly.
    if speculative_draft_tensor_parallel_size is None:
        if draft_hf_config.model_type == "mlp_speculator":
            speculative_draft_tensor_parallel_size = 1
            if target_parallel_config.tensor_parallel_size > 1:
                logger.warning(
                    "%s cannot currently be run with tp>1; "
                    "setting speculative_draft_tensor_parallel_size=1",
                    draft_hf_config.model_type)
        else:
            speculative_draft_tensor_parallel_size = \
                target_parallel_config.tensor_parallel_size
    elif speculative_draft_tensor_parallel_size not in (
            1, target_parallel_config.tensor_parallel_size):
        raise ValueError(
            f"{speculative_draft_tensor_parallel_size=} cannot be "
            f"other value than 1 or target model tensor_parallel_size")
    return speculative_draft_tensor_parallel_size

_verify_args

_verify_args() -> Self

Source code in vllm/config.py

@model_validator(mode='after')
def _verify_args(self) -> Self:
    if self.num_speculative_tokens is None:
        raise ValueError(
            "num_speculative_tokens must be provided with "
            "speculative model unless the draft model config contains an "
            "n_predict parameter.")

    if self.num_speculative_tokens <= 0:
        raise ValueError("Expected num_speculative_tokens to be greater "
                         f"than zero ({self.num_speculative_tokens}).")

    if self.draft_model_config:
        self.draft_model_config.verify_with_parallel_config(
            self.draft_parallel_config)
        # Validate and set draft token acceptance related settings.

    if self.acceptance_method is None:
        raise ValueError("acceptance_method is not set. "
                         "Expected values are rejection_sampler or "
                         "typical_acceptance_sampler.")

    if (self.acceptance_method != 'rejection_sampler'
            and self.acceptance_method != 'typical_acceptance_sampler'):
        raise ValueError(
            "Expected acceptance_method to be either "
            "rejection_sampler or typical_acceptance_sampler. Instead it "
            f"is {self.acceptance_method}")

    if self.acceptance_method == "typical_acceptance_sampler" and (
        (self.posterior_threshold is not None
         and self.posterior_threshold < 0) or
        (self.posterior_alpha is not None and self.posterior_alpha < 0)):
        raise ValueError(
            "Expected the posterior_threshold and posterior_alpha of "
            "typical_acceptance_sampler to be > 0. "
            "Instead found posterior_threshold = "
            f"{self.posterior_threshold} and posterior_alpha = "
            f"{self.posterior_alpha}")

    if (self.disable_by_batch_size is not None
            and self.disable_by_batch_size < 2):
        raise ValueError("Expect the batch size threshold of disabling "
                         "speculative decoding is > 1, but got "
                         f"{self.disable_by_batch_size=}")

    if self.method == "eagle3" and self.target_model_config and \
        "llama" not in self.target_model_config.hf_text_config.model_type:
        raise ValueError(
            "Eagle3 is only supported for Llama models. "
            f"Got {self.target_model_config.hf_text_config.model_type=}")

    return self

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    factors: list[Any] = []
    # Eagle3 affects the computation graph because it returns intermediate
    # hidden states in addition to the final hidden state.
    factors.append(self.method == "eagle3")
    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()
    return hash_str

create_draft_parallel_config staticmethod

create_draft_parallel_config(
    target_parallel_config: ParallelConfig,
    speculative_draft_tensor_parallel_size: int,
) -> ParallelConfig

Create a parallel config for use by the draft worker.

This is mostly a copy of the target parallel config, except the tp_size.

Source code in vllm/config.py

@staticmethod
def create_draft_parallel_config(
    target_parallel_config: ParallelConfig,
    speculative_draft_tensor_parallel_size: int,
) -> ParallelConfig:
    """Create a parallel config for use by the draft worker.

    This is mostly a copy of the target parallel config, except the tp_size.
    """
    draft_parallel_config = ParallelConfig(
        pipeline_parallel_size=target_parallel_config.
        pipeline_parallel_size,
        tensor_parallel_size=speculative_draft_tensor_parallel_size,
        distributed_executor_backend=target_parallel_config.
        distributed_executor_backend,
        max_parallel_loading_workers=target_parallel_config.
        max_parallel_loading_workers,
        disable_custom_all_reduce=target_parallel_config.
        disable_custom_all_reduce,
        ray_workers_use_nsight=target_parallel_config.
        ray_workers_use_nsight,
        placement_group=target_parallel_config.placement_group,
    )

    return draft_parallel_config

from_dict classmethod

from_dict(dict_value: dict) -> SpeculativeConfig

Parse the CLI value for the speculative config.

Source code in vllm/config.py

@classmethod
def from_dict(cls, dict_value: dict) -> "SpeculativeConfig":
    """Parse the CLI value for the speculative config."""
    return cls(**dict_value)

hf_config_override staticmethod

hf_config_override(
    hf_config: PretrainedConfig,
) -> PretrainedConfig

Source code in vllm/config.py

@staticmethod
def hf_config_override(hf_config: PretrainedConfig) -> PretrainedConfig:
    if hf_config.model_type == "deepseek_v3":
        hf_config.model_type = "deepseek_mtp"
    if hf_config.model_type == "deepseek_mtp":
        n_predict = getattr(hf_config, "num_nextn_predict_layers", None)
        hf_config.update({
            "n_predict": n_predict,
            "architectures": ["DeepSeekMTPModel"]
        })

    if hf_config.architectures[0] == "MiMoForCausalLM":
        hf_config.model_type = "mimo_mtp"
        n_predict = getattr(hf_config, "num_nextn_predict_layers", None)
        hf_config.update({
            "num_hidden_layers": 0,
            "n_predict": n_predict,
            "architectures": ["MiMoMTPModel"]
        })
        return hf_config

    return hf_config

use_eagle

use_eagle() -> bool

Source code in vllm/config.py

def use_eagle(self) -> bool:
    return self.method in ("eagle", "eagle3", "deepseek_mtp")

SupportsHash

Bases: Protocol

Source code in vllm/config.py

@runtime_checkable
class SupportsHash(Protocol):

    def compute_hash(self) -> str:
        ...

compute_hash

compute_hash() -> str

Source code in vllm/config.py

def compute_hash(self) -> str:
    ...

SupportsMetricsInfo

Bases: Protocol

Source code in vllm/config.py

class SupportsMetricsInfo(Protocol):

    def metrics_info(self) -> dict[str, str]:
        ...

metrics_info

metrics_info() -> dict[str, str]

Source code in vllm/config.py

def metrics_info(self) -> dict[str, str]:
    ...

TokenizerPoolConfig

This config is deprecated and will be removed in a future release.

Passing these parameters will have no effect. Please remove them from your configurations.

Source code in vllm/config.py

@config
@dataclass
class TokenizerPoolConfig:
    """This config is deprecated and will be removed in a future release.

    Passing these parameters will have no effect. Please remove them from your
    configurations.
    """

    pool_size: int = 0
    """This parameter is deprecated and will be removed in a future release.
    Passing this parameter will have no effect. Please remove it from your
    configurations."""
    pool_type: str = "ray"
    """This parameter is deprecated and will be removed in a future release.
    Passing this parameter will have no effect. Please remove it from your
    configurations."""
    extra_config: dict = field(default_factory=dict)
    """This parameter is deprecated and will be removed in a future release.
    Passing this parameter will have no effect. Please remove it from your
    configurations."""

    def __post_init__(self) -> None:
        logger.warning_once(
            "TokenizerPoolConfig is deprecated and will be removed in a "
            "future release. Passing this parameter will have no effect. "
            "Please remove it from your configurations.")

extra_config class-attribute instance-attribute

extra_config: dict = field(default_factory=dict)

This parameter is deprecated and will be removed in a future release. Passing this parameter will have no effect. Please remove it from your configurations.

pool_size class-attribute instance-attribute

pool_size: int = 0

This parameter is deprecated and will be removed in a future release. Passing this parameter will have no effect. Please remove it from your configurations.

pool_type class-attribute instance-attribute

pool_type: str = 'ray'

This parameter is deprecated and will be removed in a future release. Passing this parameter will have no effect. Please remove it from your configurations.

__post_init__

__post_init__() -> None

Source code in vllm/config.py

def __post_init__(self) -> None:
    logger.warning_once(
        "TokenizerPoolConfig is deprecated and will be removed in a "
        "future release. Passing this parameter will have no effect. "
        "Please remove it from your configurations.")

VllmConfig

Dataclass which contains all vllm-related configuration. This simplifies passing around the distinct configurations in the codebase.

Source code in vllm/config.py

@config
@dataclass(config=ConfigDict(arbitrary_types_allowed=True))
class VllmConfig:
    """Dataclass which contains all vllm-related configuration. This
    simplifies passing around the distinct configurations in the codebase.
    """

    # TODO: use default_factory once default constructing ModelConfig doesn't
    # try to download a model
    model_config: ModelConfig = None  # type: ignore
    """Model configuration."""
    cache_config: CacheConfig = field(default_factory=CacheConfig)
    """Cache configuration."""
    parallel_config: ParallelConfig = field(default_factory=ParallelConfig)
    """Parallel configuration."""
    scheduler_config: SchedulerConfig = field(default_factory=SchedulerConfig)
    """Scheduler configuration."""
    device_config: DeviceConfig = field(default_factory=DeviceConfig)
    """Device configuration."""
    load_config: LoadConfig = field(default_factory=LoadConfig)
    """Load configuration."""
    lora_config: Optional[LoRAConfig] = None
    """LoRA configuration."""
    speculative_config: Optional[SpeculativeConfig] = None
    """Speculative decoding configuration."""
    decoding_config: DecodingConfig = field(default_factory=DecodingConfig)
    """Decoding configuration."""
    observability_config: Optional[ObservabilityConfig] = None
    """Observability configuration."""
    prompt_adapter_config: Optional[PromptAdapterConfig] = None
    """Prompt adapter configuration."""
    quant_config: Optional[QuantizationConfig] = None
    """Quantization configuration."""
    compilation_config: CompilationConfig = field(
        default_factory=CompilationConfig)
    """`torch.compile` configuration for the model.

    When it is a number (0, 1, 2, 3), it will be interpreted as the
    optimization level.

    NOTE: level 0 is the default level without any optimization. level 1 and 2
    are for internal testing only. level 3 is the recommended level for
    production.

    Following the convention of traditional compilers, using `-O` without space
    is also supported. `-O3` is equivalent to `-O 3`.

    You can specify the full compilation config like so:
    `{"level": 3, "cudagraph_capture_sizes": [1, 2, 4, 8]}`
    """
    kv_transfer_config: Optional[KVTransferConfig] = None
    """The configurations for distributed KV cache transfer."""
    kv_events_config: Optional[KVEventsConfig] = None
    """The configurations for event publishing."""
    # some opaque config, only used to provide additional information
    # for the hash computation, mainly used for testing, debugging or out of
    # tree config registration.
    additional_config: Union[dict, SupportsHash] = field(default_factory=dict)
    """Additional config for specified platform. Different platforms may
    support different configs. Make sure the configs are valid for the platform
    you are using. Contents must be hashable."""
    instance_id: str = ""
    """The ID of the vLLM instance."""

    def compute_hash(self) -> str:
        """
        WARNING: Whenever a new field is added to this config,
        ensure that it is included in the factors list if
        it affects the computation graph.

        Provide a hash that uniquely identifies all the configs
        that affect the structure of the computation
        graph from input ids/embeddings to the final hidden states,
        excluding anything before input ids/embeddings and after
        the final hidden states.
        """
        factors: list[Any] = []

        # summarize vllm config
        vllm_factors: list[Any] = []
        from vllm import __version__
        vllm_factors.append(__version__)
        vllm_factors.append(envs.VLLM_USE_V1)
        if self.model_config:
            vllm_factors.append(self.model_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.cache_config:
            vllm_factors.append(self.cache_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.parallel_config:
            vllm_factors.append(self.parallel_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.scheduler_config:
            vllm_factors.append(self.scheduler_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.device_config:
            vllm_factors.append(self.device_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.load_config:
            vllm_factors.append(self.load_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.lora_config:
            vllm_factors.append(self.lora_config.compute_hash())
            # LoRA creates static buffers based on max_num_batched_tokens.
            # The tensor sizes and strides get captured in the torch.compile
            # graph explicitly.
            vllm_factors.append(
                str(self.scheduler_config.max_num_batched_tokens))
        else:
            vllm_factors.append("None")
        if self.speculative_config:
            vllm_factors.append(self.speculative_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.decoding_config:
            vllm_factors.append(self.decoding_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.observability_config:
            vllm_factors.append(self.observability_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.prompt_adapter_config:
            vllm_factors.append(self.prompt_adapter_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.quant_config:
            pass  # should be captured by model_config.quantization
        if self.compilation_config:
            vllm_factors.append(self.compilation_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.kv_transfer_config:
            vllm_factors.append(self.kv_transfer_config.compute_hash())
        else:
            vllm_factors.append("None")
        if self.additional_config:
            if isinstance(additional_config := self.additional_config, dict):
                additional_config_hash = hashlib.md5(
                    json.dumps(additional_config, sort_keys=True).encode(),
                    usedforsecurity=False,
                ).hexdigest()
            else:
                additional_config_hash = additional_config.compute_hash()
            vllm_factors.append(additional_config_hash)
        else:
            vllm_factors.append("None")
        factors.append(vllm_factors)

        hash_str = hashlib.md5(str(factors).encode(),
                               usedforsecurity=False).hexdigest()[:10]
        return hash_str

    def pad_for_cudagraph(self, batch_size: int) -> int:
        # if batch_size > self.compilation_config.max_capture_size,
        # it should raise an IndexError.
        # the caller should make sure the batch_size is within the range,
        # i.e., batch_size <= self.compilation_config.max_capture_size
        return self.compilation_config.bs_to_padded_graph_size[batch_size]

    @staticmethod
    def _get_quantization_config(
            model_config: ModelConfig,
            load_config: LoadConfig) -> Optional[QuantizationConfig]:
        """Get the quantization config."""
        from vllm.platforms import current_platform
        if model_config.quantization is not None:
            from vllm.model_executor.model_loader.weight_utils import (
                get_quant_config)
            quant_config = get_quant_config(model_config, load_config)
            capability_tuple = current_platform.get_device_capability()

            if capability_tuple is not None:
                capability = capability_tuple.to_int()
                if capability < quant_config.get_min_capability():
                    raise ValueError(
                        f"The quantization method {model_config.quantization} "
                        "is not supported for the current GPU. Minimum "
                        f"capability: {quant_config.get_min_capability()}. "
                        f"Current capability: {capability}.")
            supported_dtypes = quant_config.get_supported_act_dtypes()
            if model_config.dtype not in supported_dtypes:
                raise ValueError(
                    f"{model_config.dtype} is not supported for quantization "
                    f"method {model_config.quantization}. Supported dtypes: "
                    f"{supported_dtypes}")
            return quant_config
        return None

    @staticmethod
    def get_quantization_config(
            model_config: ModelConfig,
            load_config: LoadConfig) -> Optional[QuantizationConfig]:
        import copy

        # For some reason, the _ version of this modifies the model_config
        # object, so using deepcopy to avoid this problem.
        return VllmConfig._get_quantization_config(copy.deepcopy(model_config),
                                                   load_config)

    def with_hf_config(
        self,
        hf_config: PretrainedConfig,
        architectures: Optional[list[str]] = None,
    ) -> "VllmConfig":
        if architectures is not None:
            hf_config = copy.deepcopy(hf_config)
            hf_config.architectures = architectures

        model_config = copy.deepcopy(self.model_config)
        model_config.hf_config = hf_config

        return replace(self, model_config=model_config)

    def __post_init__(self):
        """Verify configs are valid & consistent with each other.
        """
        if self.model_config is not None:
            self.model_config.verify_async_output_proc(self.parallel_config,
                                                       self.speculative_config,
                                                       self.device_config)
            self.model_config.verify_with_parallel_config(self.parallel_config)
            self.model_config.verify_dual_chunk_attention_config(
                self.load_config)

        self.cache_config.verify_with_parallel_config(self.parallel_config)

        if self.lora_config is not None:
            self.lora_config.verify_with_cache_config(self.cache_config)
            self.lora_config.verify_with_model_config(self.model_config)
            self.lora_config.verify_lora_support()
        if self.prompt_adapter_config is not None:
            self.prompt_adapter_config.verify_with_model_config(
                self.model_config)

        if self.quant_config is None and self.model_config is not None:
            self.quant_config = VllmConfig._get_quantization_config(
                self.model_config, self.load_config)

        from vllm.platforms import current_platform
        if self.model_config is not None and \
            self.scheduler_config.chunked_prefill_enabled and \
            self.model_config.dtype == torch.float32 and \
            current_platform.get_device_capability() == (7, 5):
            logger.warning_once(
                "Turing devices tensor cores do not support float32 matmul. "
                "To workaround this limitation, vLLM will set 'ieee' input "
                "precision for chunked prefill triton kernels.")

        # async tp is built on top of sequence parallelism
        # and requires it to be enabled.
        if self.compilation_config.pass_config.enable_async_tp:
            self.compilation_config.pass_config.enable_sequence_parallelism = \
                True
        if self.compilation_config.pass_config.enable_sequence_parallelism:
            self.compilation_config.custom_ops.append("+rms_norm")
        if envs.VLLM_USE_V1 and self.model_config is not None and \
            not self.model_config.enforce_eager:
            # By default, V1 uses piecewise CUDA graphs. If full_cuda_graph
            # is set to True, full CUDA graphs will be used.
            self.compilation_config.cudagraph_num_of_warmups = 1
            self.compilation_config.level = CompilationLevel.PIECEWISE
            self.compilation_config.set_splitting_ops_for_v1()

            # The behavior of custom ops with inductor depends on the config:
            # - If use_inductor=True and custom_ops is empty:
            #   Inductor generates Triton kernels for all registered custom ops
            #   (default behavior)
            # - If use_inductor=True and custom_ops is non-empty:
            #   Custom CUDA kernels are used for specified ops while inductor
            #   generates Triton kernels for remaining ops, including misc torch
            #   ops in the model.
            if (not self.compilation_config.custom_ops
                    and self.compilation_config.use_inductor):
                # Let inductor generate Triton kernels for the custom ops.
                self.compilation_config.custom_ops = ["none"]

        self._set_cudagraph_sizes()

        if self.cache_config.cpu_offload_gb > 0 and \
            self.compilation_config.level != CompilationLevel.NO_COMPILATION \
                and not envs.VLLM_USE_V1:
            logger.warning(
                "CPU offload is not supported with `torch.compile` in v0 yet."
                " Disabling `torch.compile`.")
            self.compilation_config.level = CompilationLevel.NO_COMPILATION

        if ((not envs.VLLM_USE_V1) and self.lora_config is not None
                and self.compilation_config.level
                != CompilationLevel.NO_COMPILATION):
            logger.warning(
                "LoRA for V0 is not supported with `torch.compile` yet. "
                "Disabling `torch.compile`.")
            self.compilation_config.level = CompilationLevel.NO_COMPILATION

        if self.compilation_config.full_cuda_graph and \
            not self.model_config.disable_cascade_attn:
            logger.info("full_cuda_graph is not supported with "
                        "cascade attention. Disabling cascade attention.")
            self.model_config.disable_cascade_attn = True

        disable_chunked_prefill_reasons: list[str] = []

        if self.model_config and self.model_config.pooler_config:
            pooling_type = self.model_config.pooler_config.pooling_type
            if pooling_type is None or pooling_type.lower() != "last":
                disable_chunked_prefill_reasons.append(
                    "Only \"last\" pooling supports chunked "
                    "prefill and prefix caching; disabling both.")

        if disable_chunked_prefill_reasons:
            for reason in disable_chunked_prefill_reasons:
                logger.info(reason)
            self.scheduler_config.chunked_prefill_enabled = False
            self.scheduler_config.long_prefill_token_threshold = 0
            self.scheduler_config.max_num_batched_tokens = max(
                self.scheduler_config.max_model_len,
                DEFAULT_MAX_NUM_BATCHED_TOKENS)

            if self.cache_config is not None:
                self.cache_config.enable_prefix_caching = False

        if (self.kv_events_config is not None
                and self.kv_events_config.enable_kv_cache_events
                and not self.cache_config.enable_prefix_caching):
            logger.warning(
                "KV cache events are on, but prefix caching is not enabled."
                "Use --enable-prefix-caching to enable.")
        if (self.kv_events_config is not None
                and self.kv_events_config.publisher != "null"
                and not self.kv_events_config.enable_kv_cache_events):
            logger.warning("KV cache events are disabled,"
                           "but the scheduler is configured to publish them."
                           "Modify KVEventsConfig.enable_kv_cache_events"
                           "to True to enable.")
        current_platform.check_and_update_config(self)

        if not self.instance_id:
            self.instance_id = random_uuid()[:5]

        if (envs.VLLM_USE_V1
                and not self.scheduler_config.disable_hybrid_kv_cache_manager):
            # logger should only print warning message for hybrid models. As we
            # can't know whether the model is hybrid or not now, so we don't log
            # warning message here and will log it later.
            if not (current_platform.is_cuda() or current_platform.is_rocm()):
                # Hybrid KV cache manager is not supported on non-GPU platforms.
                self.scheduler_config.disable_hybrid_kv_cache_manager = True
            if self.kv_transfer_config is not None:
                # Hybrid KV cache manager is not compatible with KV transfer.
                self.scheduler_config.disable_hybrid_kv_cache_manager = True
            if self.kv_events_config is not None:
                # Hybrid KV cache manager is not compatible with KV events.
                self.scheduler_config.disable_hybrid_kv_cache_manager = True

    def update_sizes_for_sequence_parallelism(self,
                                              possible_sizes: list) -> list:
        # remove the sizes that not multiple of tp_size when
        # enable sequence parallelism
        removed_sizes = [
            size for size in possible_sizes
            if size % self.parallel_config.tensor_parallel_size != 0
        ]
        if removed_sizes:
            logger.warning(
                "Batch sizes %s are removed because they are not "
                "multiple of tp_size %d when "
                "sequence parallelism is enabled", removed_sizes,
                self.parallel_config.tensor_parallel_size)

        return [
            size for size in possible_sizes
            if size % self.parallel_config.tensor_parallel_size == 0
        ]

    def _set_cudagraph_sizes(self):
        """
        cudagraph batchsize padding logic:

        `[1, 2, 4] + [8 * i for i in range(1, 1025)]` is a list of all possible
        batch sizes that cudagraph will capture.

        Depending on the engine's configuration of `max_num_seqs`, the
        candidate batch sizes to capture cudagraph will shrink to the subset
        which just cover the range of `[1, max_num_seqs]`. In the common case,
        `max_num_seqs` is 256, and the cudagraph batch sizes will be
        `[1, 2, 4, 8, 16, 24, 32, 40, ..., 256]`.

        However, if users specify the cudagraph capture sizes through
        compilation config, we will use the specified sizes instead.

        In the end, `vllm_config.compilation_config.cudagraph_capture_sizes`
        will be the final sizes to capture cudagraph (in descending order).

        During runtime, if batchsize is larger than
        `vllm_config.compilation_config.cudagraph_capture_sizes`,
        no cudagraph will be used.
        If the batch size is no larger than
        `vllm_config.compilation_config.cudagraph_capture_sizes`,
        we can quickly find the padded graph size for a given batch size by
        looking up `vllm_config.compilation_config.bs_to_padded_graph_size`.
        """

        # calculate the default `batch_size_capture_list`
        if not envs.VLLM_USE_V1:
            batch_size_capture_list = []
            max_batchsize_to_capture = 0
            if self.scheduler_config is not None and \
                self.model_config is not None and \
                    not self.model_config.enforce_eager:

                possible_sizes = [1, 2, 4] + [8 * i for i in range(1, 1025)]
                if self.parallel_config.tensor_parallel_size > 1 and \
                    self.compilation_config.pass_config.enable_sequence_parallelism:
                    possible_sizes = self.update_sizes_for_sequence_parallelism(
                        possible_sizes)

                # find the minimum size that is larger than max_num_seqs,
                # which then becomes the max_batchsize_to_capture
                larger_sizes = [
                    x for x in possible_sizes
                    if x >= self.scheduler_config.max_num_seqs
                ]
                if larger_sizes:
                    max_batchsize_to_capture = larger_sizes[0]
                else:
                    max_batchsize_to_capture = possible_sizes[-1]

                # filter out the sizes that are
                # larger than max_batchsize_to_capture
                batch_size_capture_list = [
                    size for size in possible_sizes
                    if size <= max_batchsize_to_capture
                ]
        else:
            batch_size_capture_list = []
            if self.model_config is not None and \
                not self.model_config.enforce_eager:
                cuda_graph_sizes = self.scheduler_config.cuda_graph_sizes
                if len(cuda_graph_sizes) == 1:
                    batch_size_capture_list = [1, 2, 4] + [
                        i for i in range(8, cuda_graph_sizes[0] + 1, 8)
                    ]
                elif len(cuda_graph_sizes) > 1:
                    batch_size_capture_list = sorted(cuda_graph_sizes)
                else:
                    raise TypeError(f"Invalid value for {cuda_graph_sizes=}.")
                if self.parallel_config.tensor_parallel_size > 1 and \
                    self.compilation_config.pass_config.enable_sequence_parallelism:
                    batch_size_capture_list = \
                        self.update_sizes_for_sequence_parallelism(batch_size_capture_list)
                max_num_tokens = self.scheduler_config.max_num_batched_tokens
                batch_size_capture_list = [
                    size for size in batch_size_capture_list
                    if size <= max_num_tokens
                ]

        self.compilation_config.init_with_cudagraph_sizes(
            batch_size_capture_list)

    def recalculate_max_model_len(self, max_model_len: int):
        model_config = self.model_config
        max_model_len = model_config.get_and_verify_max_len(max_model_len)
        self.model_config.max_model_len = max_model_len
        self.scheduler_config.max_model_len = max_model_len
        self.compute_hash()

    def __str__(self):
        return (
            f"model={self.model_config.model!r},"
            f" speculative_config={self.speculative_config!r},"
            f" tokenizer={self.model_config.tokenizer!r}, "
            f"skip_tokenizer_init={self.model_config.skip_tokenizer_init},"
            f" tokenizer_mode={self.model_config.tokenizer_mode}, "
            f"revision={self.model_config.revision}, "
            f"override_neuron_config={self.model_config.override_neuron_config},"
            f" tokenizer_revision={self.model_config.tokenizer_revision}, "
            f"trust_remote_code={self.model_config.trust_remote_code}, "
            f"dtype={self.model_config.dtype}, "
            f"max_seq_len={self.model_config.max_model_len},"
            f" download_dir={self.load_config.download_dir!r}, "
            f"load_format={self.load_config.load_format}, "
            f"tensor_parallel_size={self.parallel_config.tensor_parallel_size},"
            f" pipeline_parallel_size={self.parallel_config.pipeline_parallel_size}, "  # noqa
            f"disable_custom_all_reduce={self.parallel_config.disable_custom_all_reduce}, "  # noqa
            f"quantization={self.model_config.quantization}, "
            f"enforce_eager={self.model_config.enforce_eager}, "
            f"kv_cache_dtype={self.cache_config.cache_dtype}, "
            f" device_config={self.device_config.device}, "
            f"decoding_config={self.decoding_config!r}, "
            f"observability_config={self.observability_config!r}, "
            f"seed={self.model_config.seed}, "
            f"served_model_name={self.model_config.served_model_name}, "
            f"num_scheduler_steps={self.scheduler_config.num_scheduler_steps}, "
            f"multi_step_stream_outputs={self.scheduler_config.multi_step_stream_outputs}, "  # noqa
            f"enable_prefix_caching={self.cache_config.enable_prefix_caching}, "
            f"chunked_prefill_enabled={self.scheduler_config.chunked_prefill_enabled}, "  # noqa
            f"use_async_output_proc={self.model_config.use_async_output_proc}, "
            f"pooler_config={self.model_config.pooler_config!r}, "
            f"compilation_config={self.compilation_config!r}")

additional_config class-attribute instance-attribute

additional_config: Union[dict, SupportsHash] = field(
    default_factory=dict
)

Additional config for specified platform. Different platforms may support different configs. Make sure the configs are valid for the platform you are using. Contents must be hashable.

cache_config class-attribute instance-attribute

cache_config: CacheConfig = field(
    default_factory=CacheConfig
)

Cache configuration.

compilation_config class-attribute instance-attribute

compilation_config: CompilationConfig = field(
    default_factory=CompilationConfig
)

torch.compile configuration for the model.

When it is a number (0, 1, 2, 3), it will be interpreted as the optimization level.

NOTE: level 0 is the default level without any optimization. level 1 and 2 are for internal testing only. level 3 is the recommended level for production.

Following the convention of traditional compilers, using -O without space is also supported. -O3 is equivalent to -O 3.

You can specify the full compilation config like so: {"level": 3, "cudagraph_capture_sizes": [1, 2, 4, 8]}

decoding_config class-attribute instance-attribute

decoding_config: DecodingConfig = field(
    default_factory=DecodingConfig
)

Decoding configuration.

device_config class-attribute instance-attribute

device_config: DeviceConfig = field(
    default_factory=DeviceConfig
)

Device configuration.

instance_id class-attribute instance-attribute

instance_id: str = ''

The ID of the vLLM instance.

kv_events_config class-attribute instance-attribute

kv_events_config: Optional[KVEventsConfig] = None

The configurations for event publishing.

kv_transfer_config class-attribute instance-attribute

kv_transfer_config: Optional[KVTransferConfig] = None

The configurations for distributed KV cache transfer.

load_config class-attribute instance-attribute

load_config: LoadConfig = field(default_factory=LoadConfig)

Load configuration.

lora_config class-attribute instance-attribute

lora_config: Optional[LoRAConfig] = None

LoRA configuration.

model_config class-attribute instance-attribute

model_config: ModelConfig = None

Model configuration.

observability_config class-attribute instance-attribute

observability_config: Optional[ObservabilityConfig] = None

Observability configuration.

parallel_config class-attribute instance-attribute

parallel_config: ParallelConfig = field(
    default_factory=ParallelConfig
)

Parallel configuration.

prompt_adapter_config class-attribute instance-attribute

prompt_adapter_config: Optional[PromptAdapterConfig] = None

Prompt adapter configuration.

quant_config class-attribute instance-attribute

quant_config: Optional[QuantizationConfig] = None

Quantization configuration.

scheduler_config class-attribute instance-attribute

scheduler_config: SchedulerConfig = field(
    default_factory=SchedulerConfig
)

Scheduler configuration.

speculative_config class-attribute instance-attribute

speculative_config: Optional[SpeculativeConfig] = None

Speculative decoding configuration.

__post_init__

__post_init__()

Verify configs are valid & consistent with each other.

Source code in vllm/config.py

def __post_init__(self):
    """Verify configs are valid & consistent with each other.
    """
    if self.model_config is not None:
        self.model_config.verify_async_output_proc(self.parallel_config,
                                                   self.speculative_config,
                                                   self.device_config)
        self.model_config.verify_with_parallel_config(self.parallel_config)
        self.model_config.verify_dual_chunk_attention_config(
            self.load_config)

    self.cache_config.verify_with_parallel_config(self.parallel_config)

    if self.lora_config is not None:
        self.lora_config.verify_with_cache_config(self.cache_config)
        self.lora_config.verify_with_model_config(self.model_config)
        self.lora_config.verify_lora_support()
    if self.prompt_adapter_config is not None:
        self.prompt_adapter_config.verify_with_model_config(
            self.model_config)

    if self.quant_config is None and self.model_config is not None:
        self.quant_config = VllmConfig._get_quantization_config(
            self.model_config, self.load_config)

    from vllm.platforms import current_platform
    if self.model_config is not None and \
        self.scheduler_config.chunked_prefill_enabled and \
        self.model_config.dtype == torch.float32 and \
        current_platform.get_device_capability() == (7, 5):
        logger.warning_once(
            "Turing devices tensor cores do not support float32 matmul. "
            "To workaround this limitation, vLLM will set 'ieee' input "
            "precision for chunked prefill triton kernels.")

    # async tp is built on top of sequence parallelism
    # and requires it to be enabled.
    if self.compilation_config.pass_config.enable_async_tp:
        self.compilation_config.pass_config.enable_sequence_parallelism = \
            True
    if self.compilation_config.pass_config.enable_sequence_parallelism:
        self.compilation_config.custom_ops.append("+rms_norm")
    if envs.VLLM_USE_V1 and self.model_config is not None and \
        not self.model_config.enforce_eager:
        # By default, V1 uses piecewise CUDA graphs. If full_cuda_graph
        # is set to True, full CUDA graphs will be used.
        self.compilation_config.cudagraph_num_of_warmups = 1
        self.compilation_config.level = CompilationLevel.PIECEWISE
        self.compilation_config.set_splitting_ops_for_v1()

        # The behavior of custom ops with inductor depends on the config:
        # - If use_inductor=True and custom_ops is empty:
        #   Inductor generates Triton kernels for all registered custom ops
        #   (default behavior)
        # - If use_inductor=True and custom_ops is non-empty:
        #   Custom CUDA kernels are used for specified ops while inductor
        #   generates Triton kernels for remaining ops, including misc torch
        #   ops in the model.
        if (not self.compilation_config.custom_ops
                and self.compilation_config.use_inductor):
            # Let inductor generate Triton kernels for the custom ops.
            self.compilation_config.custom_ops = ["none"]

    self._set_cudagraph_sizes()

    if self.cache_config.cpu_offload_gb > 0 and \
        self.compilation_config.level != CompilationLevel.NO_COMPILATION \
            and not envs.VLLM_USE_V1:
        logger.warning(
            "CPU offload is not supported with `torch.compile` in v0 yet."
            " Disabling `torch.compile`.")
        self.compilation_config.level = CompilationLevel.NO_COMPILATION

    if ((not envs.VLLM_USE_V1) and self.lora_config is not None
            and self.compilation_config.level
            != CompilationLevel.NO_COMPILATION):
        logger.warning(
            "LoRA for V0 is not supported with `torch.compile` yet. "
            "Disabling `torch.compile`.")
        self.compilation_config.level = CompilationLevel.NO_COMPILATION

    if self.compilation_config.full_cuda_graph and \
        not self.model_config.disable_cascade_attn:
        logger.info("full_cuda_graph is not supported with "
                    "cascade attention. Disabling cascade attention.")
        self.model_config.disable_cascade_attn = True

    disable_chunked_prefill_reasons: list[str] = []

    if self.model_config and self.model_config.pooler_config:
        pooling_type = self.model_config.pooler_config.pooling_type
        if pooling_type is None or pooling_type.lower() != "last":
            disable_chunked_prefill_reasons.append(
                "Only \"last\" pooling supports chunked "
                "prefill and prefix caching; disabling both.")

    if disable_chunked_prefill_reasons:
        for reason in disable_chunked_prefill_reasons:
            logger.info(reason)
        self.scheduler_config.chunked_prefill_enabled = False
        self.scheduler_config.long_prefill_token_threshold = 0
        self.scheduler_config.max_num_batched_tokens = max(
            self.scheduler_config.max_model_len,
            DEFAULT_MAX_NUM_BATCHED_TOKENS)

        if self.cache_config is not None:
            self.cache_config.enable_prefix_caching = False

    if (self.kv_events_config is not None
            and self.kv_events_config.enable_kv_cache_events
            and not self.cache_config.enable_prefix_caching):
        logger.warning(
            "KV cache events are on, but prefix caching is not enabled."
            "Use --enable-prefix-caching to enable.")
    if (self.kv_events_config is not None
            and self.kv_events_config.publisher != "null"
            and not self.kv_events_config.enable_kv_cache_events):
        logger.warning("KV cache events are disabled,"
                       "but the scheduler is configured to publish them."
                       "Modify KVEventsConfig.enable_kv_cache_events"
                       "to True to enable.")
    current_platform.check_and_update_config(self)

    if not self.instance_id:
        self.instance_id = random_uuid()[:5]

    if (envs.VLLM_USE_V1
            and not self.scheduler_config.disable_hybrid_kv_cache_manager):
        # logger should only print warning message for hybrid models. As we
        # can't know whether the model is hybrid or not now, so we don't log
        # warning message here and will log it later.
        if not (current_platform.is_cuda() or current_platform.is_rocm()):
            # Hybrid KV cache manager is not supported on non-GPU platforms.
            self.scheduler_config.disable_hybrid_kv_cache_manager = True
        if self.kv_transfer_config is not None:
            # Hybrid KV cache manager is not compatible with KV transfer.
            self.scheduler_config.disable_hybrid_kv_cache_manager = True
        if self.kv_events_config is not None:
            # Hybrid KV cache manager is not compatible with KV events.
            self.scheduler_config.disable_hybrid_kv_cache_manager = True

__str__

__str__()

Source code in vllm/config.py

def __str__(self):
    return (
        f"model={self.model_config.model!r},"
        f" speculative_config={self.speculative_config!r},"
        f" tokenizer={self.model_config.tokenizer!r}, "
        f"skip_tokenizer_init={self.model_config.skip_tokenizer_init},"
        f" tokenizer_mode={self.model_config.tokenizer_mode}, "
        f"revision={self.model_config.revision}, "
        f"override_neuron_config={self.model_config.override_neuron_config},"
        f" tokenizer_revision={self.model_config.tokenizer_revision}, "
        f"trust_remote_code={self.model_config.trust_remote_code}, "
        f"dtype={self.model_config.dtype}, "
        f"max_seq_len={self.model_config.max_model_len},"
        f" download_dir={self.load_config.download_dir!r}, "
        f"load_format={self.load_config.load_format}, "
        f"tensor_parallel_size={self.parallel_config.tensor_parallel_size},"
        f" pipeline_parallel_size={self.parallel_config.pipeline_parallel_size}, "  # noqa
        f"disable_custom_all_reduce={self.parallel_config.disable_custom_all_reduce}, "  # noqa
        f"quantization={self.model_config.quantization}, "
        f"enforce_eager={self.model_config.enforce_eager}, "
        f"kv_cache_dtype={self.cache_config.cache_dtype}, "
        f" device_config={self.device_config.device}, "
        f"decoding_config={self.decoding_config!r}, "
        f"observability_config={self.observability_config!r}, "
        f"seed={self.model_config.seed}, "
        f"served_model_name={self.model_config.served_model_name}, "
        f"num_scheduler_steps={self.scheduler_config.num_scheduler_steps}, "
        f"multi_step_stream_outputs={self.scheduler_config.multi_step_stream_outputs}, "  # noqa
        f"enable_prefix_caching={self.cache_config.enable_prefix_caching}, "
        f"chunked_prefill_enabled={self.scheduler_config.chunked_prefill_enabled}, "  # noqa
        f"use_async_output_proc={self.model_config.use_async_output_proc}, "
        f"pooler_config={self.model_config.pooler_config!r}, "
        f"compilation_config={self.compilation_config!r}")

_get_quantization_config staticmethod

_get_quantization_config(
    model_config: ModelConfig, load_config: LoadConfig
) -> Optional[QuantizationConfig]

Get the quantization config.

Source code in vllm/config.py

@staticmethod
def _get_quantization_config(
        model_config: ModelConfig,
        load_config: LoadConfig) -> Optional[QuantizationConfig]:
    """Get the quantization config."""
    from vllm.platforms import current_platform
    if model_config.quantization is not None:
        from vllm.model_executor.model_loader.weight_utils import (
            get_quant_config)
        quant_config = get_quant_config(model_config, load_config)
        capability_tuple = current_platform.get_device_capability()

        if capability_tuple is not None:
            capability = capability_tuple.to_int()
            if capability < quant_config.get_min_capability():
                raise ValueError(
                    f"The quantization method {model_config.quantization} "
                    "is not supported for the current GPU. Minimum "
                    f"capability: {quant_config.get_min_capability()}. "
                    f"Current capability: {capability}.")
        supported_dtypes = quant_config.get_supported_act_dtypes()
        if model_config.dtype not in supported_dtypes:
            raise ValueError(
                f"{model_config.dtype} is not supported for quantization "
                f"method {model_config.quantization}. Supported dtypes: "
                f"{supported_dtypes}")
        return quant_config
    return None

_set_cudagraph_sizes

_set_cudagraph_sizes()

cudagraph batchsize padding logic:

[1, 2, 4] + [8 * i for i in range(1, 1025)] is a list of all possible batch sizes that cudagraph will capture.

Depending on the engine's configuration of max_num_seqs, the candidate batch sizes to capture cudagraph will shrink to the subset which just cover the range of [1, max_num_seqs]. In the common case, max_num_seqs is 256, and the cudagraph batch sizes will be [1, 2, 4, 8, 16, 24, 32, 40, ..., 256].

However, if users specify the cudagraph capture sizes through compilation config, we will use the specified sizes instead.

In the end, vllm_config.compilation_config.cudagraph_capture_sizes will be the final sizes to capture cudagraph (in descending order).

During runtime, if batchsize is larger than vllm_config.compilation_config.cudagraph_capture_sizes, no cudagraph will be used. If the batch size is no larger than vllm_config.compilation_config.cudagraph_capture_sizes, we can quickly find the padded graph size for a given batch size by looking up vllm_config.compilation_config.bs_to_padded_graph_size.

Source code in vllm/config.py

def _set_cudagraph_sizes(self):
    """
    cudagraph batchsize padding logic:

    `[1, 2, 4] + [8 * i for i in range(1, 1025)]` is a list of all possible
    batch sizes that cudagraph will capture.

    Depending on the engine's configuration of `max_num_seqs`, the
    candidate batch sizes to capture cudagraph will shrink to the subset
    which just cover the range of `[1, max_num_seqs]`. In the common case,
    `max_num_seqs` is 256, and the cudagraph batch sizes will be
    `[1, 2, 4, 8, 16, 24, 32, 40, ..., 256]`.

    However, if users specify the cudagraph capture sizes through
    compilation config, we will use the specified sizes instead.

    In the end, `vllm_config.compilation_config.cudagraph_capture_sizes`
    will be the final sizes to capture cudagraph (in descending order).

    During runtime, if batchsize is larger than
    `vllm_config.compilation_config.cudagraph_capture_sizes`,
    no cudagraph will be used.
    If the batch size is no larger than
    `vllm_config.compilation_config.cudagraph_capture_sizes`,
    we can quickly find the padded graph size for a given batch size by
    looking up `vllm_config.compilation_config.bs_to_padded_graph_size`.
    """

    # calculate the default `batch_size_capture_list`
    if not envs.VLLM_USE_V1:
        batch_size_capture_list = []
        max_batchsize_to_capture = 0
        if self.scheduler_config is not None and \
            self.model_config is not None and \
                not self.model_config.enforce_eager:

            possible_sizes = [1, 2, 4] + [8 * i for i in range(1, 1025)]
            if self.parallel_config.tensor_parallel_size > 1 and \
                self.compilation_config.pass_config.enable_sequence_parallelism:
                possible_sizes = self.update_sizes_for_sequence_parallelism(
                    possible_sizes)

            # find the minimum size that is larger than max_num_seqs,
            # which then becomes the max_batchsize_to_capture
            larger_sizes = [
                x for x in possible_sizes
                if x >= self.scheduler_config.max_num_seqs
            ]
            if larger_sizes:
                max_batchsize_to_capture = larger_sizes[0]
            else:
                max_batchsize_to_capture = possible_sizes[-1]

            # filter out the sizes that are
            # larger than max_batchsize_to_capture
            batch_size_capture_list = [
                size for size in possible_sizes
                if size <= max_batchsize_to_capture
            ]
    else:
        batch_size_capture_list = []
        if self.model_config is not None and \
            not self.model_config.enforce_eager:
            cuda_graph_sizes = self.scheduler_config.cuda_graph_sizes
            if len(cuda_graph_sizes) == 1:
                batch_size_capture_list = [1, 2, 4] + [
                    i for i in range(8, cuda_graph_sizes[0] + 1, 8)
                ]
            elif len(cuda_graph_sizes) > 1:
                batch_size_capture_list = sorted(cuda_graph_sizes)
            else:
                raise TypeError(f"Invalid value for {cuda_graph_sizes=}.")
            if self.parallel_config.tensor_parallel_size > 1 and \
                self.compilation_config.pass_config.enable_sequence_parallelism:
                batch_size_capture_list = \
                    self.update_sizes_for_sequence_parallelism(batch_size_capture_list)
            max_num_tokens = self.scheduler_config.max_num_batched_tokens
            batch_size_capture_list = [
                size for size in batch_size_capture_list
                if size <= max_num_tokens
            ]

    self.compilation_config.init_with_cudagraph_sizes(
        batch_size_capture_list)

compute_hash

compute_hash() -> str

WARNING: Whenever a new field is added to this config, ensure that it is included in the factors list if it affects the computation graph.

Provide a hash that uniquely identifies all the configs that affect the structure of the computation graph from input ids/embeddings to the final hidden states, excluding anything before input ids/embeddings and after the final hidden states.

Source code in vllm/config.py

def compute_hash(self) -> str:
    """
    WARNING: Whenever a new field is added to this config,
    ensure that it is included in the factors list if
    it affects the computation graph.

    Provide a hash that uniquely identifies all the configs
    that affect the structure of the computation
    graph from input ids/embeddings to the final hidden states,
    excluding anything before input ids/embeddings and after
    the final hidden states.
    """
    factors: list[Any] = []

    # summarize vllm config
    vllm_factors: list[Any] = []
    from vllm import __version__
    vllm_factors.append(__version__)
    vllm_factors.append(envs.VLLM_USE_V1)
    if self.model_config:
        vllm_factors.append(self.model_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.cache_config:
        vllm_factors.append(self.cache_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.parallel_config:
        vllm_factors.append(self.parallel_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.scheduler_config:
        vllm_factors.append(self.scheduler_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.device_config:
        vllm_factors.append(self.device_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.load_config:
        vllm_factors.append(self.load_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.lora_config:
        vllm_factors.append(self.lora_config.compute_hash())
        # LoRA creates static buffers based on max_num_batched_tokens.
        # The tensor sizes and strides get captured in the torch.compile
        # graph explicitly.
        vllm_factors.append(
            str(self.scheduler_config.max_num_batched_tokens))
    else:
        vllm_factors.append("None")
    if self.speculative_config:
        vllm_factors.append(self.speculative_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.decoding_config:
        vllm_factors.append(self.decoding_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.observability_config:
        vllm_factors.append(self.observability_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.prompt_adapter_config:
        vllm_factors.append(self.prompt_adapter_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.quant_config:
        pass  # should be captured by model_config.quantization
    if self.compilation_config:
        vllm_factors.append(self.compilation_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.kv_transfer_config:
        vllm_factors.append(self.kv_transfer_config.compute_hash())
    else:
        vllm_factors.append("None")
    if self.additional_config:
        if isinstance(additional_config := self.additional_config, dict):
            additional_config_hash = hashlib.md5(
                json.dumps(additional_config, sort_keys=True).encode(),
                usedforsecurity=False,
            ).hexdigest()
        else:
            additional_config_hash = additional_config.compute_hash()
        vllm_factors.append(additional_config_hash)
    else:
        vllm_factors.append("None")
    factors.append(vllm_factors)

    hash_str = hashlib.md5(str(factors).encode(),
                           usedforsecurity=False).hexdigest()[:10]
    return hash_str

get_quantization_config staticmethod

get_quantization_config(
    model_config: ModelConfig, load_config: LoadConfig
) -> Optional[QuantizationConfig]

Source code in vllm/config.py

@staticmethod
def get_quantization_config(
        model_config: ModelConfig,
        load_config: LoadConfig) -> Optional[QuantizationConfig]:
    import copy

    # For some reason, the _ version of this modifies the model_config
    # object, so using deepcopy to avoid this problem.
    return VllmConfig._get_quantization_config(copy.deepcopy(model_config),
                                               load_config)

pad_for_cudagraph

pad_for_cudagraph(batch_size: int) -> int

Source code in vllm/config.py

def pad_for_cudagraph(self, batch_size: int) -> int:
    # if batch_size > self.compilation_config.max_capture_size,
    # it should raise an IndexError.
    # the caller should make sure the batch_size is within the range,
    # i.e., batch_size <= self.compilation_config.max_capture_size
    return self.compilation_config.bs_to_padded_graph_size[batch_size]

recalculate_max_model_len

recalculate_max_model_len(max_model_len: int)

Source code in vllm/config.py

def recalculate_max_model_len(self, max_model_len: int):
    model_config = self.model_config
    max_model_len = model_config.get_and_verify_max_len(max_model_len)
    self.model_config.max_model_len = max_model_len
    self.scheduler_config.max_model_len = max_model_len
    self.compute_hash()

update_sizes_for_sequence_parallelism

update_sizes_for_sequence_parallelism(
    possible_sizes: list,
) -> list

Source code in vllm/config.py

def update_sizes_for_sequence_parallelism(self,
                                          possible_sizes: list) -> list:
    # remove the sizes that not multiple of tp_size when
    # enable sequence parallelism
    removed_sizes = [
        size for size in possible_sizes
        if size % self.parallel_config.tensor_parallel_size != 0
    ]
    if removed_sizes:
        logger.warning(
            "Batch sizes %s are removed because they are not "
            "multiple of tp_size %d when "
            "sequence parallelism is enabled", removed_sizes,
            self.parallel_config.tensor_parallel_size)

    return [
        size for size in possible_sizes
        if size % self.parallel_config.tensor_parallel_size == 0
    ]

with_hf_config

with_hf_config(
    hf_config: PretrainedConfig,
    architectures: Optional[list[str]] = None,
) -> VllmConfig

Source code in vllm/config.py

def with_hf_config(
    self,
    hf_config: PretrainedConfig,
    architectures: Optional[list[str]] = None,
) -> "VllmConfig":
    if architectures is not None:
        hf_config = copy.deepcopy(hf_config)
        hf_config.architectures = architectures

    model_config = copy.deepcopy(self.model_config)
    model_config.hf_config = hf_config

    return replace(self, model_config=model_config)

_check_valid_dtype

_check_valid_dtype(model_type: str, dtype: dtype)

Source code in vllm/config.py

def _check_valid_dtype(model_type: str, dtype: torch.dtype):
    if model_type in _FLOAT16_NOT_SUPPORTED_MODELS and dtype == torch.float16:
        reason = _FLOAT16_NOT_SUPPORTED_MODELS[model_type]
        raise ValueError(f"The model type {model_type!r} "
                         f"does not support float16. Reason: {reason}")

    return True

_find_dtype

_find_dtype(
    model_id: str,
    config: PretrainedConfig,
    *,
    revision: Optional[str],
)

Source code in vllm/config.py

def _find_dtype(
    model_id: str,
    config: PretrainedConfig,
    *,
    revision: Optional[str],
):
    # NOTE: getattr(config, "torch_dtype", torch.float32) is not correct
    # because config.torch_dtype can be None.
    config_dtype = getattr(config, "torch_dtype", None)

    # Fallbacks for multi-modal models if the root config
    # does not define torch_dtype
    if config_dtype is None:
        config_dtype = getattr(config.get_text_config(), "torch_dtype", None)
    if config_dtype is None and hasattr(config, "vision_config"):
        config_dtype = getattr(config.vision_config, "torch_dtype", None)
    if config_dtype is None and hasattr(config, "encoder_config"):
        config_dtype = getattr(config.encoder_config, "torch_dtype", None)

    # Try to read the dtype of the weights if they are in safetensors format
    if config_dtype is None:
        repo_mt = try_get_safetensors_metadata(model_id, revision=revision)

        if repo_mt and (files_mt := repo_mt.files_metadata):
            param_dtypes: set[torch.dtype] = {
                _SAFETENSORS_TO_TORCH_DTYPE[dtype_str]
                for file_mt in files_mt.values()
                for dtype_str in file_mt.parameter_count
                if dtype_str in _SAFETENSORS_TO_TORCH_DTYPE
            }

            if param_dtypes:
                return common_broadcastable_dtype(param_dtypes)

    if config_dtype is None:
        config_dtype = torch.float32

    return config_dtype

_get_and_verify_dtype

_get_and_verify_dtype(
    model_id: str,
    config: PretrainedConfig,
    dtype: Union[str, dtype],
    *,
    is_pooling_model: bool,
    revision: Optional[str] = None,
) -> dtype

Source code in vllm/config.py

def _get_and_verify_dtype(
    model_id: str,
    config: PretrainedConfig,
    dtype: Union[str, torch.dtype],
    *,
    is_pooling_model: bool,
    revision: Optional[str] = None,
) -> torch.dtype:
    config_dtype = _find_dtype(model_id, config, revision=revision)
    model_type = config.model_type

    if isinstance(dtype, str):
        dtype = dtype.lower()
        if dtype == "auto":
            # Set default dtype from model config
            torch_dtype = _resolve_auto_dtype(
                model_type,
                config_dtype,
                is_pooling_model=is_pooling_model,
            )
        else:
            if dtype not in _STR_DTYPE_TO_TORCH_DTYPE:
                raise ValueError(f"Unknown dtype: {dtype!r}")
            torch_dtype = _STR_DTYPE_TO_TORCH_DTYPE[dtype]
    elif isinstance(dtype, torch.dtype):
        torch_dtype = dtype
    else:
        raise ValueError(f"Unknown dtype: {dtype}")

    _check_valid_dtype(model_type, torch_dtype)

    if torch_dtype != config_dtype:
        if torch_dtype == torch.float32:
            # Upcasting to float32 is allowed.
            logger.info("Upcasting %s to %s.", config_dtype, torch_dtype)
        elif config_dtype == torch.float32:
            # Downcasting from float32 to float16 or bfloat16 is allowed.
            logger.info("Downcasting %s to %s.", config_dtype, torch_dtype)
        else:
            # Casting between float16 and bfloat16 is allowed with a warning.
            logger.warning("Casting %s to %s.", config_dtype, torch_dtype)

    return torch_dtype

_get_and_verify_max_len

_get_and_verify_max_len(
    hf_config: PretrainedConfig,
    tokenizer_config: Optional[dict],
    max_model_len: Optional[int],
    disable_sliding_window: bool,
    sliding_window_len: Optional[
        Union[int, list[Optional[int]]]
    ],
    spec_target_max_model_len: Optional[int] = None,
    encoder_config: Optional[Any] = None,
) -> int

Get and verify the model's maximum length.

Source code in vllm/config.py

def _get_and_verify_max_len(
    hf_config: PretrainedConfig,
    tokenizer_config: Optional[dict],
    max_model_len: Optional[int],
    disable_sliding_window: bool,
    sliding_window_len: Optional[Union[int, list[Optional[int]]]],
    spec_target_max_model_len: Optional[int] = None,
    encoder_config: Optional[Any] = None,
) -> int:
    """Get and verify the model's maximum length."""
    derived_max_model_len = float("inf")
    possible_keys = [
        # OPT
        "max_position_embeddings",
        # GPT-2
        "n_positions",
        # MPT
        "max_seq_len",
        # ChatGLM2
        "seq_length",
        # Command-R
        "model_max_length",
        # Whisper
        "max_target_positions",
        # Others
        "max_sequence_length",
        "max_seq_length",
        "seq_len",
    ]
    # Choose the smallest "max_length" from the possible keys
    max_len_key = None
    for key in possible_keys:
        max_len = getattr(hf_config, key, None)
        if max_len is not None:
            max_len_key = key if max_len < derived_max_model_len \
                else max_len_key
            derived_max_model_len = min(derived_max_model_len, max_len)
    # For Command-R / Cohere, Cohere2 / Aya Vision models
    if tmp_max_len := getattr(hf_config, "model_max_length", None):
        max_len_key = "model_max_length"
        derived_max_model_len = tmp_max_len

    # If sliding window is manually disabled, max_length should be less
    # than the sliding window length in the model config.
    if disable_sliding_window and sliding_window_len is not None:

        sliding_window_len_min = get_min_sliding_window(sliding_window_len)
        max_len_key = "sliding_window" \
            if sliding_window_len_min < derived_max_model_len else max_len_key
        derived_max_model_len = min(derived_max_model_len,
                                    sliding_window_len_min)

    # Consider model_max_length in tokenizer_config
    if tokenizer_config:
        tokenizer_model_max_length = tokenizer_config.get(
            "model_max_length", derived_max_model_len)
        derived_max_model_len = min(derived_max_model_len,
                                    tokenizer_model_max_length)

    # If none of the keys were found in the config, use a default and
    # log a warning.
    if derived_max_model_len == float("inf"):
        if max_model_len is not None:
            # If max_model_len is specified, we use it.
            return max_model_len

        if spec_target_max_model_len is not None:
            # If this is a speculative draft model, we use the max model len
            # from the target model.
            return spec_target_max_model_len

        default_max_len = 2048
        logger.warning(
            "The model's config.json does not contain any of the following "
            "keys to determine the original maximum length of the model: "
            "%s. Assuming the model's maximum length is %d.", possible_keys,
            default_max_len)
        derived_max_model_len = default_max_len

    rope_scaling = getattr(hf_config, "rope_scaling", None)
    # NOTE(woosuk): Gemma3's max_model_len (128K) is already scaled by RoPE
    # scaling, so we skip applying the scaling factor again.
    if rope_scaling is not None and "gemma3" not in hf_config.model_type:
        # No need to consider "type" key because of patch_rope_scaling when
        # loading HF config
        rope_type = rope_scaling["rope_type"]

        if rope_type not in ("su", "longrope", "llama3"):
            if disable_sliding_window:
                # TODO(robertgshaw): Find a model that supports rope_scaling
                # with sliding window to see if this case should be allowed.
                raise NotImplementedError(
                    "Disabling sliding window is not supported for models "
                    "with rope_scaling. Please raise an issue so we can "
                    "investigate.")

            # NOTE: rope_type == "default" does not define factor
            # https://github.com/huggingface/transformers/blob/v4.45.2/src/transformers/modeling_rope_utils.py
            scaling_factor = rope_scaling.get("factor", 1.0)

            if rope_type == "yarn":
                derived_max_model_len = rope_scaling[
                    "original_max_position_embeddings"]
            derived_max_model_len *= scaling_factor

    if encoder_config and "max_seq_length" in encoder_config:
        derived_max_model_len = encoder_config["max_seq_length"]

    # If the user specified a max length, make sure it is smaller than the
    # derived length from the HF model config.
    if max_model_len is None:
        max_model_len = int(derived_max_model_len)
        if current_platform.is_tpu():
            logger.warning(
                "--max-model-len is not specified, "
                "it's currently using model's default length %s, "
                "which might be too large."
                "Please input with --max-model-len based on your "
                "request input length and output length, to avoid "
                "unnecessary degradation.", max_model_len)
    elif max_model_len > derived_max_model_len:
        # Some models might have a separate key for specifying model_max_length
        # that will be bigger than derived_max_model_len. We compare user input
        # with model_max_length and allow this override when it's smaller.
        model_max_length = getattr(hf_config, "model_max_length", None)
        if model_max_length is not None and max_model_len <= model_max_length:
            if disable_sliding_window:
                # TODO(robertgshaw): Find a model that has model_max_length
                # with sliding window to see if this case should be allowed.
                raise NotImplementedError(
                    "Disabling sliding window is not supported for models "
                    "model_max_length in the config. Please raise an issue "
                    "so we can investigate.")
        else:
            msg = (
                f"User-specified max_model_len ({max_model_len}) is greater "
                f"than the derived max_model_len ({max_len_key}="
                f"{derived_max_model_len} or model_max_length="
                f"{model_max_length} in model's config.json). This may lead "
                "to incorrect model outputs or CUDA errors.")
            if envs.VLLM_ALLOW_LONG_MAX_MODEL_LEN:
                logger.warning(
                    "%s Make sure the value is correct and within the "
                    "model context size.", msg)
            else:
                raise ValueError(
                    f"{msg} To allow overriding this maximum, set "
                    "the env var VLLM_ALLOW_LONG_MAX_MODEL_LEN=1")
    return int(max_model_len)

_is_valid_dtype

_is_valid_dtype(model_type: str, dtype: dtype)

Source code in vllm/config.py

def _is_valid_dtype(model_type: str, dtype: torch.dtype):
    if model_type in _FLOAT16_NOT_SUPPORTED_MODELS and dtype == torch.float16:  # noqa: E501, SIM103
        return False

    return True

_resolve_auto_dtype

_resolve_auto_dtype(
    model_type: str,
    config_dtype: dtype,
    *,
    is_pooling_model: bool,
)

Source code in vllm/config.py

def _resolve_auto_dtype(
    model_type: str,
    config_dtype: torch.dtype,
    *,
    is_pooling_model: bool,
):
    from vllm.platforms import current_platform

    supported_dtypes = [
        dtype for dtype in current_platform.supported_dtypes
        if _is_valid_dtype(model_type, dtype)
    ]

    if is_pooling_model and torch.float16 in supported_dtypes:
        preferred_dtype = torch.float16
    else:
        preferred_dtype = supported_dtypes[0]

    # Downcast for float32 models
    if config_dtype == torch.float32:
        config_dtype = preferred_dtype

    if config_dtype in supported_dtypes:
        return config_dtype

    # Ensure device compatibility
    device_name = current_platform.get_device_name()
    device_capability = current_platform.get_device_capability()

    if device_capability is None:
        device_str = f"{device_name!r}"
    else:
        version_str = device_capability.as_version_str()
        device_str = f"{device_name!r} (with compute capability {version_str})"

    logger.warning(
        "Your device %s doesn't support %s. "
        "Falling back to %s for compatibility.",
        device_str,
        config_dtype,
        preferred_dtype,
    )

    return preferred_dtype

assert_hashable

assert_hashable(text)

Source code in vllm/config.py

def assert_hashable(text):
    if not contains_object_print(text):
        return True
    raise AssertionError(
        f"vLLM tried to hash some configs that may have Python objects ids "
        f"in them. This is a bug, please file an issue. "
        f"Text being hashed: {text}")

config

config(cls: ConfigT) -> ConfigT

A decorator that ensures all fields in a dataclass have default values and that each field has a docstring.

If a ConfigT is used as a CLI argument itself, the default value provided by get_kwargs will be the result parsing a JSON string as the kwargs (i.e. ConfigT(**json.loads(cli_arg))). However, if a particular ConfigT requires custom construction from CLI (i.e. CompilationConfig), it can have a from_cli method, which will be called instead.

Source code in vllm/config.py

def config(cls: ConfigT) -> ConfigT:
    """
    A decorator that ensures all fields in a dataclass have default values
    and that each field has a docstring.

    If a `ConfigT` is used as a CLI argument itself, the default value provided
    by `get_kwargs` will be the result parsing a JSON string as the kwargs
    (i.e. `ConfigT(**json.loads(cli_arg))`). However, if a particular `ConfigT`
    requires custom construction from CLI (i.e. `CompilationConfig`), it can
    have a `from_cli` method, which will be called instead.
    """
    if not is_dataclass(cls):
        raise TypeError("The decorated class must be a dataclass.")
    attr_docs = get_attr_docs(cls)
    for f in fields(cls):
        if f.init and f.default is MISSING and f.default_factory is MISSING:
            raise ValueError(
                f"Field '{f.name}' in {cls.__name__} must have a default value."
            )

        if f.name not in attr_docs:
            raise ValueError(
                f"Field '{f.name}' in {cls.__name__} must have a docstring.")

        if get_origin(f.type) is Union:
            args = get_args(f.type)
            literal_args = [arg for arg in args if get_origin(arg) is Literal]
            if len(literal_args) > 1:
                raise ValueError(
                    f"Field '{f.name}' in {cls.__name__} must use a single "
                    "Literal type. Please use 'Literal[Literal1, Literal2]' "
                    "instead of 'Union[Literal1, Literal2]'.")
    return cls

contains_object_print

contains_object_print(text)

Check if the text looks like a printed Python object, e.g. contains any substring matching the pattern: "at 0xFFFFFFF>" We match against 0x followed by 2-16 hex chars (there's a max of 16 on a 64 bit system).

Parameters:

Name	Type	Description	Default
text	str	The text to check	required

Returns:

Name	Type	Description
result	bool	True if a match is found, False otherwise.

Source code in vllm/config.py

def contains_object_print(text):
    """
    Check if the text looks like a printed Python object, e.g.
    contains any substring matching the pattern: "at 0xFFFFFFF>"
    We match against 0x followed by 2-16 hex chars (there's
    a max of 16 on a 64 bit system).

    Args:
        text (str): The text to check

    Returns:
        result (bool): `True` if a match is found, `False` otherwise.
    """
    pattern = r'at 0x[a-fA-F0-9]{2,16}>'
    match = re.search(pattern, text)
    return match is not None

get_attr_docs

get_attr_docs(cls: type[Any]) -> dict[str, str]

Get any docstrings placed after attribute assignments in a class body.

https://davidism.com/mit-license/

Source code in vllm/config.py

def get_attr_docs(cls: type[Any]) -> dict[str, str]:
    """
    Get any docstrings placed after attribute assignments in a class body.

    https://davidism.com/mit-license/
    """

    def pairwise(iterable):
        """
        Manually implement https://docs.python.org/3/library/itertools.html#itertools.pairwise

        Can be removed when Python 3.9 support is dropped.
        """
        iterator = iter(iterable)
        a = next(iterator, None)

        for b in iterator:
            yield a, b
            a = b

    cls_node = ast.parse(textwrap.dedent(inspect.getsource(cls))).body[0]

    if not isinstance(cls_node, ast.ClassDef):
        raise TypeError("Given object was not a class.")

    out = {}

    # Consider each pair of nodes.
    for a, b in pairwise(cls_node.body):
        # Must be an assignment then a constant string.
        if (not isinstance(a, (ast.Assign, ast.AnnAssign))
                or not isinstance(b, ast.Expr)
                or not isinstance(b.value, ast.Constant)
                or not isinstance(b.value.value, str)):
            continue

        doc = inspect.cleandoc(b.value.value)

        # An assignment can have multiple targets (a = b = v), but an
        # annotated assignment only has one target.
        targets = a.targets if isinstance(a, ast.Assign) else [a.target]

        for target in targets:
            # Must be assigning to a plain name.
            if not isinstance(target, ast.Name):
                continue

            out[target.id] = doc

    return out

get_current_model_prefix

get_current_model_prefix() -> str

Get the prefix of the model that's currently being initialized.

Source code in vllm/config.py

def get_current_model_prefix() -> str:
    """
    Get the prefix of the model that's currently being initialized.
    """
    assert _current_prefix is not None, \
        "Current model prefix is not set. "
    return _current_prefix

get_current_vllm_config

get_current_vllm_config() -> VllmConfig

Source code in vllm/config.py

def get_current_vllm_config() -> VllmConfig:
    if _current_vllm_config is None:
        # in ci, usually when we test custom ops/modules directly,
        # we don't set the vllm config. In that case, we set a default
        # config.
        logger.warning("Current vLLM config is not set.")
        from vllm.config import VllmConfig
        return VllmConfig()
    return _current_vllm_config

get_field

get_field(cls: ConfigType, name: str) -> Field

Get the default factory field of a dataclass by name. Used for getting default factory fields in EngineArgs.

Source code in vllm/config.py

def get_field(cls: ConfigType, name: str) -> Field:
    """Get the default factory field of a dataclass by name. Used for getting
    default factory fields in `EngineArgs`."""
    if not is_dataclass(cls):
        raise TypeError("The given class is not a dataclass.")
    cls_fields = {f.name: f for f in fields(cls)}
    if name not in cls_fields:
        raise ValueError(f"Field '{name}' not found in {cls.__name__}.")
    named_field: Field = cls_fields[name]
    if (default_factory := named_field.default_factory) is not MISSING:
        return field(default_factory=default_factory)
    if (default := named_field.default) is not MISSING:
        return field(default=default)
    raise ValueError(
        f"{cls.__name__}.{name} must have a default value or default factory.")

get_layers_from_vllm_config

get_layers_from_vllm_config(
    vllm_config: VllmConfig, layer_type: type[T]
) -> dict[str, T]

Source code in vllm/config.py

def get_layers_from_vllm_config(vllm_config: VllmConfig,
                                layer_type: type[T]) -> dict[str, T]:
    return {
        layer_name: layer
        for layer_name, layer in
        vllm_config.compilation_config.static_forward_context.items()
        if isinstance(layer, layer_type)
    }

get_min_sliding_window

get_min_sliding_window(
    sliding_window: Union[int, list[Optional[int]]],
) -> int

Source code in vllm/config.py

def get_min_sliding_window(
        sliding_window: Union[int, list[Optional[int]]]) -> int:
    if isinstance(sliding_window, list):
        return min(s for s in sliding_window if s is not None)

    return sliding_window

get_served_model_name

get_served_model_name(
    model: str,
    served_model_name: Optional[Union[str, list[str]]],
)

If the input is a non-empty list, the first model_name in served_model_name is taken. If the input is a non-empty string, it is used directly. For cases where the input is either an empty string or an empty list, the fallback is to use self.model.

Source code in vllm/config.py

def get_served_model_name(model: str,
                          served_model_name: Optional[Union[str, list[str]]]):
    """
    If the input is a non-empty list, the first model_name in
    `served_model_name` is taken.
    If the input is a non-empty string, it is used directly.
    For cases where the input is either an empty string or an
    empty list, the fallback is to use `self.model`.
    """
    if not served_model_name:
        return model
    if isinstance(served_model_name, list):
        return served_model_name[0]
    return served_model_name

is_init_field

is_init_field(cls: ConfigType, name: str) -> bool

Source code in vllm/config.py

def is_init_field(cls: ConfigType, name: str) -> bool:
    return next(f for f in fields(cls) if f.name == name).init

set_current_vllm_config

set_current_vllm_config(
    vllm_config: VllmConfig,
    check_compile=False,
    prefix: Optional[str] = None,
)

Temporarily set the current vLLM config. Used during model initialization. We save the current vLLM config in a global variable, so that all modules can access it, e.g. custom ops can access the vLLM config to determine how to dispatch.

Source code in vllm/config.py

@contextmanager
def set_current_vllm_config(vllm_config: VllmConfig,
                            check_compile=False,
                            prefix: Optional[str] = None):
    """
    Temporarily set the current vLLM config.
    Used during model initialization.
    We save the current vLLM config in a global variable,
    so that all modules can access it, e.g. custom ops
    can access the vLLM config to determine how to dispatch.
    """
    global _current_vllm_config, _current_prefix
    old_vllm_config = _current_vllm_config
    old_prefix = _current_prefix
    from vllm.compilation.counter import compilation_counter
    num_models_seen = compilation_counter.num_models_seen
    try:
        _current_vllm_config = vllm_config
        _current_prefix = prefix
        yield
    except Exception:
        raise
    else:
        logger.debug("enabled custom ops: %s",
                     vllm_config.compilation_config.enabled_custom_ops)
        logger.debug("disabled custom ops: %s",
                     vllm_config.compilation_config.disabled_custom_ops)
        if check_compile and \
            vllm_config.compilation_config.level == CompilationLevel.PIECEWISE \
            and compilation_counter.num_models_seen == num_models_seen:
            # If the model supports compilation,
            # compilation_counter.num_models_seen should be increased
            # by at least 1.
            # If it is not increased, it means the model does not support
            # compilation (does not have @support_torch_compile decorator).
            logger.warning(
                "`torch.compile` is turned on, but the model %s"
                " does not support it. Please open an issue on GitHub"
                " if you want it to be supported.",
                vllm_config.model_config.model)
    finally:
        _current_vllm_config = old_vllm_config
        _current_prefix = old_prefix