Interface RAGPipelineConfig

Configuration for RAG pipeline

interface RAGPipelineConfig {
    llmProvider: LLMProvider;
    vectorStore: VectorStore;
    chunker: DocumentChunker;
    embeddingModel: string;
    generationModel?: string;
    defaultSearchLimit?: number;
    defaultMinScore?: number;
}

Index

Properties

llmProvider vectorStore chunker embeddingModel generationModel? defaultSearchLimit? defaultMinScore?

Properties

`Readonly`llmProvider

llmProvider: LLMProvider

The LLM provider for generating embeddings and completions

`Readonly`vectorStore

vectorStore: VectorStore

The vector store for document persistence

`Readonly`chunker

chunker: DocumentChunker

The document chunker for splitting documents

`Readonly`embeddingModel

embeddingModel: string

Model to use for embeddings

`Optional` `Readonly`generationModel

generationModel?: string

Model to use for text generation

`Optional` `Readonly`defaultSearchLimit

defaultSearchLimit?: number

Default search parameters

`Optional` `Readonly`defaultMinScore

defaultMinScore?: number

Default minimum similarity score