Chapter 2. vLLM server usage

Run in headless mode. See multi-node data parallel documentation for more details.

Default: False

How many API server processes to run.

Default: 1

Read CLI options from a config file. Must be a YAML with the following options: https://docs.vllm.ai/en/latest/configuration/serve_args.html

Default: None

Disable logging statistics.

Default: False

Enable logging requests.

Default: False

Disable logging requests.

Default: True

Host name.

Default: None

Port number.

Default: 8000

Unix domain socket path. If set, host and port arguments are ignored.

Default: None

Possible choices: critical, debug, error, info, trace, warning

Log level for uvicorn.

Default: info

Disable uvicorn access log.

Default: False

Allow credentials.

Default: False

Allowed origins.

Default: ['*']

Allowed methods.

Default: ['*']

Allowed headers.

Default: ['*']

If provided, the server will require one of these keys to be presented in the header.

Default: None

LoRA modules configurations in either 'name=path' format or JSON format or JSON list format. Example (old format): 'name=path' Example (new format): {"name": "name", "path": "lora_path", "base_model_name": "id"}

Default: None

The file path to the chat template, or the template in single-line form for the specified model.

Default: None

Possible choices: auto, openai, string

The format to render message content within a chat template.

Default: auto

Whether to trust the chat template provided in the request. If False, the server will always use the chat template specified by --chat-template or the ones from tokenizer.

Default: False

The role name to return if request.add_generation_prompt=true.

Default: assistant

The file path to the SSL key file.

Default: None

The file path to the SSL cert file.

Default: None

The CA certificates file.

Default: None

Refresh SSL Context when SSL certificate files change

Default: False

Whether client certificate is required (see stdlib ssl module’s).

Default: 0

FastAPI root_path when app is behind a path based routing proxy.

Default: None

Additional ASGI middleware to apply to the app. We accept multiple --middleware arguments. The value should be an import path. If a function is provided, vLLM will add it to the server using @app.middleware('http'). If a class is provided, vLLM will add it to the server using app.add_middleware().

Default: []

When --max-logprobs is specified, represents single tokens as strings of the form 'token_id:{token_id}' so that tokens that are not JSON-encodable can be identified.

Default: False

If specified, will run the OpenAI frontend server in the same process as the model serving engine.

Default: False

If specified, API server will add X-Request-Id header to responses.

Default: False

Enable auto tool choice for supported models. Use --tool-call-parser to specify which parser to use.

Default: False

If specified, exclude tool definitions in prompts when tool_choice='none'.

Default: False

Select the tool call parser depending on the model that you’re using. This is used to parse the model-generated tool call into OpenAI API format. Required for --enable-auto-tool-choice. You can choose any option from the built-in parsers or register a plugin via --tool-parser-plugin.

Default: None

Special the tool parser plugin write to parse the model-generated tool into OpenAI API format, the name register in this plugin can be used in --tool-call-parser.

Default: ``

Comma-separated list of host:port pairs (IPv4, IPv6, or hostname). Examples: 127.0.0.1:8000, [::1]:8000, localhost:1234. Or demo for demo purpose.

Default: None

Path to logging config JSON file for both vllm and uvicorn

Default: None

Max number of prompt characters or prompt ID numbers being printed in log. The default of None means unlimited.

Default: None

Disable FastAPI’s OpenAPI schema, Swagger UI, and ReDoc endpoint.

Default: False

If set to True, enable prompt_tokens_details in usage.

Default: False

If set to True, enable tracking server_load_metrics in the app state.

Default: False

If set to True, including usage on every request.

Default: False

Enable the /get_tokenizer_info endpoint. May expose chat templates and other tokenizer configuration.

Default: False

If True, log model outputs (generations). Requires --enable-log-requests.

Default: False

Maximum size (bytes) of an incomplete HTTP event (header or body) for h11 parser. Helps mitigate header abuse.

Default: 4194304 (4 MB)

Maximum number of HTTP headers allowed in a request for h11 parser. Helps mitigate header abuse.s

Default: 256

If set to True, log the stack trace of error responses

Default: False

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.

Default: Qwen/Qwen3-0.6B

Possible choices: auto, draft, generate, pooling

The type of model runner to use. Each vLLM instance only supports one model runner, even if the same model can be used for multiple types.

Default: auto

Possible choices: auto, classify, embed, none, reward

Convert the model using adapters defined in [vllm.model_executor.models.adapters][]. The most common use case is to adapt a text generation model to be used for pooling tasks.

Default: auto

Possible choices: auto, classify, draft, embed, embedding, generate, reward, score, transcription, None

The task to use the model for. If the model supports more than one model runner, this is used to select which model runner to run.

Note that the model may support other tasks using the same model runner.

Default: None

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

Default: None

Possible choices: auto, custom, mistral, slow

Tokenizer mode:

Default: auto

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

Default: False

Possible choices: auto, bfloat16, float, float16, float32, half

Data type for model weights and activations:

Default: auto

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

Default: None

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

Default: None

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.

Default: ``

If set, only media URLs that belong to this domain can be used for multi-modal inputs.

Default: 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.

Default: 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.

Default: None

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

Should either be a valid JSON string or JSON keys passed individually.

Default: {}

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

Default: 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.

Default: 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:

Parse human-readable integers like '1k', '2M', etc. Including decimal values with decimal multipliers.

Examples:
- '1k' -> 1,000
- '1K' -> 1,024
- '25.6k' -> 25,600

Examples:
- '1k' -> 1,000
- '1K' -> 1,024
- '25.6k' -> 25,600

Default: 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.

Default: None

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.

Default: False

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. -1 means no cap, i.e. all (output_length * vocab_size) logprobs are allowed to be returned and it may cause OOM.

Default: 20

Possible choices: processed_logits, processed_logprobs, raw_logits, raw_logprobs

Indicates the content returned in the logprobs and prompt_logprobs. Supported mode: 1) raw_logprobs, 2) processed_logprobs, 3) raw_logits, 4) processed_logits. Raw means the values before applying any logit processors, like bad words. Processed means the values after applying all processors, including temperature and top_k/top_p.

Default: raw_logprobs

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.

Default: 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.

Default: 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.

Default: 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.

Default: False

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.

Default: None

Possible choices: auto, hf, mistral

The format of the model config to load:

Default: auto

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).

Default: None

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

Default: {}

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

Should either be a valid JSON string or JSON keys passed individually.

Default: None

Use pooler_config instead. This field will be removed in v0.12.0 or v1.0.0, whichever is sooner.

Should either be a valid JSON string or JSON keys passed individually.

Default: 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.

Default: None

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.

Default: auto

Overrides or sets generation config. For example, {"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.

Should either be a valid JSON string or JSON keys passed individually.

Default: {}

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

Default: False

Possible choices: auto, terratorch, transformers, vllm

Which implementation of the model to use:

Default: auto

Override dtype for attention

Default: None

One or more logits processors' fully-qualified class names or class definitions

Default: None

IOProcessor plugin name to load at model startup

Default: None

The format of the model weights to load:

Default: auto

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

Default: None

Specifies the loading strategy for safetensors weights.

Default: lazy

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

Default: {}

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

Default: None

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

Default: True

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

Default: cpu

Possible choices: deepseek_r1, glm45, openai_gptoss, granite, hunyuan_a13b, mistral, qwen3, seed_oss, step3

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

Default: ``

--guided-decoding-backend will be removed in v0.12.0.

Default: None

--guided-decoding-disable-fallback will be removed in v0.12.0.

Default: None

--guided-decoding-disable-any-whitespace will be removed in v0.12.0.

Default: None

--guided-decoding-disable-additional-properties will be removed in v0.12.0.

Default: None

Possible choices: external_launcher, mp, ray, uni

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 only support Ray for distributed inference.

Default: None

Number of pipeline parallel groups.

Default: 1

Number of tensor parallel groups.

Default: 1

Number of decode context parallel groups, because the world size does not change by dcp, it simply reuse the GPUs of TP group, and tp_size needs to be divisible by dcp_size.

Default: 1

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

Default: 1

Data parallel rank of this instance. When set, enables external load balancer mode.

Default: None

Starting data parallel rank for secondary nodes.

Default: None

Number of data parallel replicas to run on this node.

Default: None

Address of data parallel cluster head-node.

Default: None

Port for data parallel RPC communication.

Default: None

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

Default: mp

Whether to use "hybrid" DP LB mode. Applies only to online serving and when data_parallel_size > 0. Enables running an AsyncLLM and API server on a "per-node" basis where vLLM load balances between local data parallel ranks, but an external LB balances between vLLM nodes/replicas. Set explicitly in conjunction with --data-parallel-start-rank.

Default: False

Use expert parallelism instead of tensor parallelism for MoE layers.

Default: False

Enable dual batch overlap for the model executor.

Default: False

The threshold for dual batch overlap for batches only containing decodes. If the number of tokens in the request is greater than this threshold, microbatching will be used. Otherwise, the request will be processed in a single batch.

Default: 32

The threshold for dual batch overlap for batches that contain one or more prefills. If the number of tokens in the request is greater than this threshold, microbatching will be used. Otherwise, the request will be processed in a single batch.

Default: 512

Enable expert parallelism load balancing for MoE layers.

Default: False

Expert parallelism configuration.

Should either be a valid JSON string or JSON keys passed individually.

Default: EPLBConfig(window_size=1000, step_interval=3000, num_redundant_experts=0, log_balancedness=False)

Possible choices: linear, round_robin

The expert placement strategy for MoE layers:

Default: linear

--num-redundant-experts will be removed in v0.12.0.

Default: None

--eplb-window-size will be removed in v0.12.0.

Default: None

--eplb-step-interval will be removed in v0.12.0.

Default: None

--eplb-log-balancedness will be removed in v0.12.0.

Default: 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.

Default: None

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

Default: False

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

Default: False

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

Default: auto

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.

Default: ``

Default: False

Possible choices: 1, 8, 16, 32, 64, 128

Size of a contiguous cache block in number of tokens. On CUDA devices, only block sizes up to 32 are supported.

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

Default: None

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.

Default: 0.9

Size of KV Cache per GPU in bytes. By default, this is set to None and vllm can automatically infer the kv cache size based on gpu_memory_utilization. However, users may want to manually specify the kv cache memory size. kv_cache_memory_bytes allows more fine-grain control of how much memory gets used when compared with using gpu_memory_memory_utilization. Note that kv_cache_memory_bytes (when not-None) ignores gpu_memory_utilization

Parse human-readable integers like '1k', '2M', etc. Including decimal values with decimal multipliers.

Examples:
- '1k' -> 1,000
- '1K' -> 1,024
- '25.6k' -> 25,600

Examples:
- '1k' -> 1,000
- '1K' -> 1,024
- '25.6k' -> 25,600

Default: None

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

Default: 4

Possible choices: auto, bfloat16, fp8, fp8_e4m3, fp8_e5m2, fp8_inc

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). Intel Gaudi (HPU) supports fp8 (using fp8_inc). Some models (namely DeepSeekV3.2) default to fp8, set to bfloat16 to use bfloat16 instead, this is an invalid option for models that do not default to fp8.

Default: auto

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

Default: None

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

Default: None

Possible choices: sha256, sha256_cbor

Set the hash algorithm for prefix caching:

Default: sha256

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.

Default: 0

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.

Default: False

This feature is work in progress and no prefill optimization takes place with this flag enabled currently.

In some KV sharing setups, e.g. YOCO (https://arxiv.org/abs/2405.05254), some layers can skip tokens corresponding to prefill. This flag enables attention metadata for eligible layers to be overridden with metadata necessary for implementing this optimization in some models (e.g. Gemma3n)

Default: False

Possible choices: auto, float32

The data type to use for the Mamba cache (both the conv as well as the ssm state). If set to 'auto', the data type will be inferred from the model config.

Default: auto

Possible choices: auto, float32

The data type to use for the Mamba cache (ssm state only, conv state will still be controlled by mamba_cache_dtype). If set to 'auto', the data type for the ssm state will be determined by mamba_cache_dtype.

Default: auto

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: {"image": 16, "video": 2}

Should either be a valid JSON string or JSON keys passed individually.

Default: {}

Additional args passed to process media inputs, keyed by modalities. For example, to set num_frames for video, set --media-io-kwargs '{"video": {"num_frames": 40} }'

Should either be a valid JSON string or JSON keys passed individually.

Default: {}

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 transformers.AutoProcessor.from_pretrained.

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

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

Should either be a valid JSON string or JSON keys passed individually.

Default: None

The size (in GiB) of the multi-modal processor cache, which is used to avoid re-processing past multi-modal inputs.

This cache is duplicated for each API process and engine core process, resulting in a total memory usage of mm_processor_cache_gb * (api_server_count + data_parallel_size).

Set to 0 to disable this cache completely (not recommended).

Default: 4

Default: False

Possible choices: lru, shm

Type of cache to use for the multi-modal preprocessor/mapper. If shm, use shared memory FIFO cache. If lru, use mirrored LRU cache.

Default: lru

Size limit (in MiB) for each object stored in the multi-modal processor shared memory cache. Only effective when mm_processor_cache_type is "shm".

Default: 128

Possible choices: data, weights

Indicates how to optimize multi-modal encoder inference using tensor parallelism (TP).

Default: weights

Enable fully interleaved support for multimodal prompts, while using --chat-template-content-format=string.

Default: False

When enabled, skips multimodal memory profiling and only profiles with language backbone model during engine initialization.

This reduces engine startup time but shifts the responsibility to users for estimating the peak memory usage of the activation of multimodal encoder and embedding cache.

Default: False

Sets pruning rate for video pruning via Efficient Video Sampling. Value sits in range [0;1) and determines fraction of media tokens from each video to be pruned.

Default: None

If True, enable handling of LoRA adapters.

Default: None

Enable bias for LoRA adapters. This option will be removed in v0.12.0.

Default: False

Max number of LoRAs in a single batch.

Default: 1

Max LoRA rank.

Default: 16

(Deprecated) Maximum size of extra vocabulary that can be present in a LoRA adapter. Will be removed in v0.12.0.

Default: 256

Possible choices: auto, bfloat16, float16

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

Default: auto

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

Default: None

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.

Default: False

Dictionary mapping specific modalities to LoRA model paths; this field is only applicable to multimodal models and should be leveraged when a model always expects a LoRA to be active when a given modality is present. Note that currently, if a request provides multiple additional modalities, each of which have their own LoRA, we do NOT apply default_mm_loras because we currently only support one lora adapter per prompt. When run in offline mode, the lora IDs for n modalities will be automatically assigned to 1-n with the names of the modalities in alphabetic order.

Should either be a valid JSON string or JSON keys passed individually.

Default: 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.

Default: None

Target URL to which OpenTelemetry traces will be sent.

Default: None

Possible choices: all, model, worker, None, model,worker, model,all, worker,model, worker,all, all,model, all,worker

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.

Default: 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.

Parse human-readable integers like '1k', '2M', etc. Including decimal values with decimal multipliers.

Examples:
- '1k' -> 1,000
- '1K' -> 1,024
- '25.6k' -> 25,600

Examples:
- '1k' -> 1,000
- '1K' -> 1,024
- '25.6k' -> 25,600

Default: 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.

Default: None

For chunked prefill, the maximum number of sequences that can be partially prefilled concurrently.

Default: 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.

Default: 1

Cuda graph capture sizes

Default: []

For chunked prefill, a request is considered long if the prompt is longer than this number of tokens.

Default: 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.

Default: 0

Possible choices: fcfs, priority

The scheduling policy to use:

Default: fcfs

If True, prefill requests can be chunked based on the remaining max_num_batched_tokens.

Default: None

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.

Default: False

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".

Default: vllm.core.scheduler.Scheduler

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.

Default: False

If set to True, perform async scheduling. This may help reduce the CPU overheads, leading to better latency and throughput. However, async scheduling is currently not supported with some features such as structured outputs, speculative decoding, and pipeline parallelism.

Default: False

Speculative decoding configuration.

Should either be a valid JSON string or JSON keys passed individually.

Default: None

The configurations for distributed KV cache transfer.

Should either be a valid JSON string or JSON keys passed individually.

Default: None

The configurations for event publishing.

Should either be a valid JSON string or JSON keys passed individually.

Default: None

torch.compile and cudagraph capture configuration for the model.

As a shorthand, -O<n> can be used to directly specify the compilation level n: -O3 is equivalent to -O.level=3 (same as -O='{"level":3}'). Currently, -O <n>and -O=<n>are supported as well but this will likely be removed in favor of clearer -O<n>syntax in the future.</n></n></n>

You can specify the full compilation config like so: {"level": 3, "cudagraph_capture_sizes": [1, 2, 4, 8]}

Should either be a valid JSON string or JSON keys passed individually.

Default:

{
   "level":null,
   "debug_dump_path":"",
   "cache_dir":"",
   "backend":"",
   "custom_ops":[

   ],
   "splitting_ops":null,
   "use_inductor":true,
   "compile_sizes":null,
   "inductor_compile_config":{
      "enable_auto_functionalized_v2":false
   },
   "inductor_passes":{

   },
   "cudagraph_mode":null,
   "use_cudagraph":true,
   "cudagraph_num_of_warmups":0,
   "cudagraph_capture_sizes":null,
   "cudagraph_copy_inputs":false,
   "full_cuda_graph":false,
   "use_inductor_graph_partition":false,
   "pass_config":{

   },
   "max_capture_size":null,
   "local_cache_dir":null
}

{
   "level":null,
   "debug_dump_path":"",
   "cache_dir":"",
   "backend":"",
   "custom_ops":[

   ],
   "splitting_ops":null,
   "use_inductor":true,
   "compile_sizes":null,
   "inductor_compile_config":{
      "enable_auto_functionalized_v2":false
   },
   "inductor_passes":{

   },
   "cudagraph_mode":null,
   "use_cudagraph":true,
   "cudagraph_num_of_warmups":0,
   "cudagraph_capture_sizes":null,
   "cudagraph_copy_inputs":false,
   "full_cuda_graph":false,
   "use_inductor_graph_partition":false,
   "pass_config":{

   },
   "max_capture_size":null,
   "local_cache_dir":null
}

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.

Default: {}

Structured outputs configuration.

Should either be a valid JSON string or JSON keys passed individually.

Default:

config = StructuredOutputsConfig(
    backend='auto',
    disable_fallback=False,
    disable_any_whitespace=False,
    disable_additional_properties=False,
    reasoning_parser=''
)

config = StructuredOutputsConfig(
    backend='auto',
    disable_fallback=False,
    disable_any_whitespace=False,
    disable_additional_properties=False,
    reasoning_parser=''
)