Struct CompletionRequest 

pub struct CompletionRequest {

    pub model: String,
    pub prompt: StringOrArray,
    pub suffix: Option<String>,
    pub max_tokens: Option<u32>,
    pub temperature: Option<f32>,
    pub top_p: Option<f32>,
    pub n: Option<u32>,
    pub stream: bool,
    pub stream_options: Option<StreamOptions>,
    pub logprobs: Option<u32>,
    pub echo: bool,
    pub stop: Option<StringOrArray>,
    pub presence_penalty: Option<f32>,
    pub frequency_penalty: Option<f32>,
    pub best_of: Option<u32>,
    pub logit_bias: Option<HashMap<String, f32>>,
    pub user: Option<String>,
    pub seed: Option<i64>,
    pub top_k: Option<i32>,
    pub min_p: Option<f32>,
    pub min_tokens: Option<u32>,
    pub repetition_penalty: Option<f32>,
    pub regex: Option<String>,
    pub ebnf: Option<String>,
    pub json_schema: Option<String>,
    pub stop_token_ids: Option<Vec<u32>>,
    pub no_stop_trim: bool,
    pub ignore_eos: bool,
    pub skip_special_tokens: bool,
    pub lora_path: Option<String>,
    pub session_params: Option<HashMap<String, Value>>,
    pub return_hidden_states: bool,
    pub sampling_seed: Option<u64>,
    pub other: Map<String, Value>,
}

Fields

model: String

ID of the model to use (required for OpenAI, optional for some implementations, such as SGLang)

prompt: StringOrArray

The prompt(s) to generate completions for

suffix: Option<String>

The suffix that comes after a completion of inserted text

max_tokens: Option<u32>

The maximum number of tokens to generate

temperature: Option<f32>

What sampling temperature to use, between 0 and 2

top_p: Option<f32>

An alternative to sampling with temperature (nucleus sampling)

n: Option<u32>

How many completions to generate for each prompt

stream: bool

Whether to stream back partial progress

stream_options: Option<StreamOptions>

Options for streaming response

logprobs: Option<u32>

Include the log probabilities on the logprobs most likely tokens

echo: bool

Echo back the prompt in addition to the completion

stop: Option<StringOrArray>

Up to 4 sequences where the API will stop generating further tokens

presence_penalty: Option<f32>

Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far

frequency_penalty: Option<f32>

Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far

best_of: Option<u32>

Generates best_of completions server-side and returns the “best”

logit_bias: Option<HashMap<String, f32>>

Modify the likelihood of specified tokens appearing in the completion

user: Option<String>

A unique identifier representing your end-user

seed: Option<i64>

If specified, our system will make a best effort to sample deterministically

top_k: Option<i32>

Top-k sampling parameter (-1 to disable)

min_p: Option<f32>

Min-p nucleus sampling parameter

min_tokens: Option<u32>

Minimum number of tokens to generate

repetition_penalty: Option<f32>

Repetition penalty for reducing repetitive text

regex: Option<String>

Regex constraint for output generation

ebnf: Option<String>

EBNF grammar constraint for structured output

json_schema: Option<String>

JSON schema constraint for structured output

stop_token_ids: Option<Vec<u32>>

Specific token IDs to use as stop conditions

no_stop_trim: bool

Skip trimming stop tokens from output

ignore_eos: bool

Ignore end-of-sequence tokens during generation

skip_special_tokens: bool

Skip special tokens during detokenization

lora_path: Option<String>

Path to LoRA adapter(s) for model customization

session_params: Option<HashMap<String, Value>>

Session parameters for continual prompting

return_hidden_states: bool

Return model hidden states

sampling_seed: Option<u64>

Sampling seed for deterministic outputs

other: Map<String, Value>

Additional fields including bootstrap info for PD routing

Trait Implementations

impl Clone for CompletionRequest

fn clone(&self) -> CompletionRequest

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for CompletionRequest

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for CompletionRequest

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl GenerationRequest for CompletionRequest

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 Serialize for CompletionRequest

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

where __S: Serializer,

Serialize this value into the given Serde serializer.