Struct ExtractorBuilder 

pub struct ExtractorBuilder<M, T>
where
    M: CompletionModel,
    T: JsonSchema + for<'a> Deserialize<'a> + Serialize + WasmCompatSend + WasmCompatSync + 'static,
{ /* private fields */ }

Builder for the Extractor

Implementations

impl<M, T> ExtractorBuilder<M, T>

where M: CompletionModel, T: JsonSchema + for<'a> Deserialize<'a> + Serialize + WasmCompatSend + WasmCompatSync + 'static,

pub fn new(model: M) -> Self

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

Add additional preamble to the extractor

pub fn context(self, doc: &str) -> Self

Add a context document to the extractor

pub fn additional_params(self, params: Value) -> Self

pub fn max_tokens(self, max_tokens: u64) -> Self

Set the maximum number of tokens for the completion

pub fn retries(self, retries: u64) -> Self

Set the maximum number of retries for the extractor.

pub fn tool_choice(self, choice: ToolChoice) -> Self

Set the tool_choice option for the inner Agent.

pub fn build(self) -> Extractor<M, T>

Build the Extractor

pub fn dynamic_context( self, sample: usize, dynamic_context: impl VectorStoreIndexDyn + Send + Sync + 'static, ) -> Self

Add dynamic context (RAG) to the extractor.

On each prompt, sample documents will be retrieved from the index based on the RAG text and inserted in the request.

Trait Implementations

impl<M, T> From<ExtractorBuilder<M, T>> for LlmJudgeBuilder<M, T>

where M: CompletionModel, T: Send + Sync + JsonSchema + Serialize + for<'a> Deserialize<'a>,

Available on crate feature experimental only.

fn from(ext: ExtractorBuilder<M, T>) -> Self

Converts to this type from the input type.