Struct LlamaRequest

pub struct LlamaRequest {
    pub model: Option<Model>,
    pub messages: Vec<Message>,
    pub stream: bool,
    pub options: Options,
    pub url: Option<Url>,
}

Request builder for Ollama LLM interactions.

Provides a fluent interface for constructing requests to the Ollama service, handling URL management, message construction, and model configuration.

Examples

let request =
  LlamaRequest::new().with_model(Model::Llama3p2c3b).with_message("Explain how a computer works");

let response = request.send().await?;

Fields

model: Option<Model>

The LLM model to use for generation. If not specified, will result in an error when sending the request.

messages: Vec<Message>

Vector of conversation messages. Must contain at least one message before sending the request. Messages are processed in order to maintain conversation context.

stream: bool

Whether to stream the response. Currently not implemented, but when true will enable token-by-token streaming of the model’s response.

options: Options

Generation parameters including temperature, top-k, top-p, and maximum token count. Uses sensible defaults if not explicitly configured.

url: Option<Url>

The target URL for the request. If not specified, defaults to localhost:11434 with a warning. Skipped during serialization.

Implementations

impl LlamaRequest

pub fn new() -> Self

Creates a new request with builder-style API with default settings.

pub fn with_host(self, host: &str) -> Self

Sets the host URL for the request.

Arguments

• host - Base URL for the Ollama service

pub fn with_endpoint(self, endpoint: OllamaEndpoint) -> Self

Sets the API endpoint for the request.

Arguments

• endpoint - The API endpoint to use

Note: Currently only the Chat endpoint is fully supported. Other endpoints will generate warnings about potential limitations.

pub fn with_model(self, model: Model) -> Self

Sets the model to use for the request.

Arguments

• model - The LLM model to use

pub fn with_message(self, content: &str) -> Self

Adds a message to the conversation.

Arguments

• content - The message content

pub async fn send(&self) -> Result<LlamaResponse>

Sends the request to the Ollama service.

Returns

Returns a Result containing either:

• A LlamaResponse with the model’s response
• A LearnerError if the request fails

Errors

This function will return an error if:

• No model is specified
• No messages are provided
• The network request fails
• The response cannot be parsed

Trait Implementations

impl Default for LlamaRequest

fn default() -> LlamaRequest

Returns the “default value” for a type.

impl Serialize for LlamaRequest

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

where __S: Serializer,

Serialize this value into the given Serde serializer.