YamlRagSection

Models & Types · AiDotNet.Configuration

Composite YAML section for the full retrieval-augmented generation (RAG) pipeline. Each sub-component names a concrete implementation (via YamlTypeSection) that the applier resolves from the type registry and passes to the matching builder facade method.

For Beginners

A RAG pipeline is made of several parts working together. This section lets you configure them all at once from YAML:

Constructors

Constructor Summary
YamlRagSection()

Properties

Property Summary
List<YamlTypeSection> QueryProcessors { get; set; } Ordered query processors applied before retrieval (resolved from the “QueryProcessor” registry).
YamlTypeSection DocumentStore { get; set; } The document store for vector/hybrid retrieval (resolved from the “DocumentStore” registry).
YamlTypeSection EmbeddingModel { get; set; } The embedding model used to vectorize text (resolved from the “EmbeddingModel” registry).
YamlTypeSection Generator { get; set; } The generator that produces grounded answers (resolved from the “Generator” registry).
YamlTypeSection GraphStore { get; set; } The graph storage backend for Graph RAG (resolved from the “GraphStore” registry).
YamlTypeSection KnowledgeGraph { get; set; } Advanced knowledge-graph options.
YamlTypeSection Reranker { get; set; } The reranker that reorders retrieved documents (resolved from the “Reranker” registry).
YamlTypeSection Retriever { get; set; } The retriever that finds relevant documents (resolved from the “Retriever” registry).
YamlTypeSection SimilarityMetric { get; set; } The similarity metric used for vector search (resolved from the “SimilarityMetric” registry).