Trait ComputeLanguageModelRequests

pub trait ComputeLanguageModelRequests {
    type Seed: HasAssociatedOutputName + Send + Sync;

    // Required method
    fn compute_language_model_requests(
        &self,
        model: &LanguageModelType,
        input_tokens: &[Self::Seed],
    ) -> Vec<LanguageModelBatchAPIRequest>;
}

This is the trait we will typically need to implement manually

Required Associated Types

type Seed: HasAssociatedOutputName + Send + Sync

Required Methods

fn compute_language_model_requests( &self, model: &LanguageModelType, input_tokens: &[Self::Seed], ) -> Vec<LanguageModelBatchAPIRequest>

Identify which new items need to be processed and build the requests.

Implementors

impl<E> ComputeLanguageModelRequests for LanguageModelTokenExpander<E>

where E: ExpandedToken<Error = SaveLoadError> + DeserializeOwned + Named + AiJsonTemplate + GetTargetPathForAIExpansion + LoadFromFile + 'static,

type Seed = <LanguageModelTokenExpander<E> as ComputeLanguageModelCoreQuery>::Seed