Struct LanguageModelCallOptions 

pub struct LanguageModelCallOptions {

    pub prompt: LanguageModelPrompt,
    pub stream: Option<bool>,
    pub max_output_tokens: Option<u32>,
    pub temperature: Option<f32>,
    pub top_p: Option<f32>,
    pub top_k: Option<u32>,
    pub stop_sequences: Option<Vec<String>>,
    pub presence_penalty: Option<f32>,
    pub frequency_penalty: Option<f32>,
    pub response_format: Option<LanguageModelResponseFormat>,
    pub seed: Option<u64>,
    pub tools: Option<Vec<LanguageModelTool>>,
    pub tool_choice: Option<LanguageModelToolChoice>,
    pub include_raw_chunks: Option<bool>,
    pub abort_signal: Option<CancellationToken>,
    pub headers: Option<HeaderMap>,
    pub provider_options: Option<ProviderOptions>,
}

Options for calling a language model

Fields

prompt: LanguageModelPrompt

The prompt to send to the language model, which is a sequence of messages from the system, user, assistant, and tools

stream: Option<bool>

Whether to stream the response as it’s generated, or return it all at once when complete

max_output_tokens: Option<u32>

The maximum number of tokens to generate in the response

temperature: Option<f32>

Sampling temperature to use, between 0 and 1

top_p: Option<f32>

Top-p (nucleus) sampling probability

top_k: Option<u32>

Top-k sampling

stop_sequences: Option<Vec<String>>

Stop sequences to end generation when encountered

presence_penalty: Option<f32>

Presence penalty to penalize new tokens based on whether they appear in the prompt

frequency_penalty: Option<f32>

Frequency penalty to penalize new tokens based on their frequency in the prompt

response_format: Option<LanguageModelResponseFormat>

The format of the response

seed: Option<u64>

Seed for random number generation

tools: Option<Vec<LanguageModelTool>>

Tools available to the language model

tool_choice: Option<LanguageModelToolChoice>

The tool choice strategy to use when the model calls a tool

include_raw_chunks: Option<bool>

Whether to include raw chunks in the response

abort_signal: Option<CancellationToken>

Signal to abort the request

headers: Option<HeaderMap>

Custom headers to include in the request

provider_options: Option<ProviderOptions>

Provider-specific options that can be used to pass additional information to the provider or control provider-specific behavior

Trait Implementations

impl Clone for LanguageModelCallOptions

fn clone(&self) -> LanguageModelCallOptions

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LanguageModelCallOptions

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

Formats the value using the given formatter.