Constructor RagEngine

Namespace

LMKit.Retrieval

Assembly

LM-Kit.NET.dll

RagEngine(LM, IVectorStore)

Initializes a new instance of the RagEngine class using a specified model for embedding generation.

public RagEngine(LM embeddingModel, IVectorStore vectorStore = null)

Parameters

embeddingModel LM

An LM object specifying the embeddings model.

vectorStore IVectorStore

An optional IVectorStore instance that provides support for vector storage operations. If provided, it will be used for storing and retrieving embedding vectors; otherwise, vector storage functionality may be disabled or handled using default mechanisms.

Examples

using LMKit.Model;
using LMKit.Retrieval;
using System;

class Example
{
    static void Main()
    {
        // Create an embedding model
        LM embeddingModel = LM.LoadFromModelID("nomic-embed-text");

        RagEngine ragEngine = new RagEngine(embeddingModel);
        Console.WriteLine("RagEngine created successfully.");
    }
}

Exceptions

ArgumentNullException

Thrown if embeddingModel is null.

RagEngine(IEnumerable<LM>, IVectorStore)

Initializes a new instance of the RagEngine class using one or more embedding models. Supports multimodal retrieval-augmented generation when you supply, for example, both text-only and vision-enabled models.

All provided LM instances must share the same embedding space (same dimension + coordinate system), otherwise similarity lookups will be invalid.

public RagEngine(IEnumerable<LM> embeddingModels, IVectorStore vectorStore = null)

Parameters

embeddingModels IEnumerable<LM>

A non-null, non-empty collection of LM objects specifying the embedding models to register.

vectorStore IVectorStore

An optional IVectorStore instance for persisting embeddings; if omitted, an in-memory store is used.

Exceptions

ArgumentNullException

Thrown if embeddingModels is null or contains no elements.