Struct GenerateRequest 

pub struct GenerateRequest {

    pub text: Option<String>,
    pub model: Option<String>,
    pub input_ids: Option<InputIds>,
    pub input_embeds: Option<Value>,
    pub image_data: Option<Value>,
    pub video_data: Option<Value>,
    pub audio_data: Option<Value>,
    pub sampling_params: Option<SamplingParams>,
    pub return_logprob: Option<bool>,
    pub logprob_start_len: Option<i32>,
    pub top_logprobs_num: Option<i32>,
    pub token_ids_logprob: Option<Vec<u32>>,
    pub return_text_in_logprobs: bool,
    pub stream: bool,
    pub log_metrics: bool,
    pub return_hidden_states: bool,
    pub modalities: Option<Vec<String>>,
    pub session_params: Option<HashMap<String, Value>>,
    pub lora_path: Option<String>,
    pub lora_id: Option<String>,
    pub custom_logit_processor: Option<String>,
    pub bootstrap_host: Option<String>,
    pub bootstrap_port: Option<i32>,
    pub bootstrap_room: Option<i32>,
    pub bootstrap_pair_key: Option<String>,
    pub data_parallel_rank: Option<i32>,
    pub background: bool,
    pub conversation_id: Option<String>,
    pub priority: Option<i32>,
    pub extra_key: Option<String>,
    pub no_logs: bool,
    pub custom_labels: Option<HashMap<String, String>>,
    pub return_bytes: bool,
    pub return_entropy: bool,
    pub rid: Option<String>,
}

Fields

text: Option<String>

Text input - SGLang native format

model: Option<String>

input_ids: Option<InputIds>

Input IDs for tokenized input

input_embeds: Option<Value>

Input embeddings for direct embedding input Can be a 2D array (single request) or 3D array (batch of requests) Placeholder for future use

image_data: Option<Value>

Image input data Can be an image instance, file name, URL, or base64 encoded string Supports single images, lists of images, or nested lists for batch processing Placeholder for future use

video_data: Option<Value>

Video input data Can be a file name, URL, or base64 encoded string Supports single videos, lists of videos, or nested lists for batch processing Placeholder for future use

audio_data: Option<Value>

Audio input data Can be a file name, URL, or base64 encoded string Supports single audio files, lists of audio, or nested lists for batch processing Placeholder for future use

sampling_params: Option<SamplingParams>

Sampling parameters (sglang style)

return_logprob: Option<bool>

Whether to return logprobs

logprob_start_len: Option<i32>

If return logprobs, the start location in the prompt for returning logprobs.

top_logprobs_num: Option<i32>

If return logprobs, the number of top logprobs to return at each position.

token_ids_logprob: Option<Vec<u32>>

If return logprobs, the token ids to return logprob for.

return_text_in_logprobs: bool

Whether to detokenize tokens in text in the returned logprobs.

stream: bool

Whether to stream the response

log_metrics: bool

Whether to log metrics for this request (e.g. health_generate calls do not log metrics)

return_hidden_states: bool

Return model hidden states

modalities: Option<Vec<String>>

The modalities of the image data [image, multi-images, video]

session_params: Option<HashMap<String, Value>>

Session parameters for continual prompting

lora_path: Option<String>

Path to LoRA adapter(s) for model customization

lora_id: Option<String>

LoRA adapter ID (if pre-loaded)

custom_logit_processor: Option<String>

Custom logit processor for advanced sampling control. Must be a serialized instance of CustomLogitProcessor in python/sglang/srt/sampling/custom_logit_processor.py Use the processor’s to_str() method to generate the serialized string.

bootstrap_host: Option<String>

For disaggregated inference

bootstrap_port: Option<i32>

For disaggregated inference

bootstrap_room: Option<i32>

For disaggregated inference

bootstrap_pair_key: Option<String>

For disaggregated inference

data_parallel_rank: Option<i32>

Data parallel rank routing

background: bool

Background response

conversation_id: Option<String>

Conversation ID for tracking

priority: Option<i32>

Priority for the request

extra_key: Option<String>

Extra key for classifying the request (e.g. cache_salt)

no_logs: bool

Whether to disallow logging for this request (e.g. due to ZDR)

custom_labels: Option<HashMap<String, String>>

Custom metric labels

return_bytes: bool

Whether to return bytes for image generation

return_entropy: bool

Whether to return entropy

rid: Option<String>

Request ID for tracking (inherited from BaseReq in Python)

Trait Implementations

impl Clone for GenerateRequest

fn clone(&self) -> GenerateRequest

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for GenerateRequest

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for GenerateRequest

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl GenerationRequest for GenerateRequest

fn is_stream(&self) -> bool

Check if the request is for streaming

fn get_model(&self) -> Option<&str>

Get the model name if specified

fn extract_text_for_routing(&self) -> String

Extract text content for routing decisions

impl Normalizable for GenerateRequest

fn normalize(&mut self)

Normalize the request by applying defaults and transformations

impl Serialize for GenerateRequest

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl Validate for GenerateRequest

fn validate(&self) -> Result<(), ValidationErrors>

impl<'v_a> ValidateArgs<'v_a> for GenerateRequest

type Args = ()

fn validate_with_args(&self, args: Self::Args) -> Result<(), ValidationErrors>