Skip to main content

Llama CPP

Compatibility

Only available on Node.js.

This module is based on the node-llama-cpp Node.js bindings for llama.cpp, allowing you to work with a locally running LLM. This allows you to work with a much smaller quantized model capable of running on a laptop environment, ideal for testing and scratch padding ideas without running up a bill!

Setup

You'll need to install the node-llama-cpp module to communicate with your local model.

npm install -S node-llama-cpp
npm install @langchain/community

You will also need a local Llama 2 model (or a model supported by node-llama-cpp). You will need to pass the path to this model to the LlamaCpp module as a part of the parameters (see example).

Out-of-the-box node-llama-cpp is tuned for running on a MacOS platform with support for the Metal GPU of Apple M-series of processors. If you need to turn this off or need support for the CUDA architecture then refer to the documentation at node-llama-cpp.

For advice on getting and preparing llama2 see the documentation for the LLM version of this module.

A note to LangChain.js contributors: if you want to run the tests associated with this module you will need to put the path to your local model in the environment variable LLAMA_PATH.

Usage

Basic use

We need to provide a path to our local Llama2 model, also the embeddings property is always set to true in this module.

import { LlamaCppEmbeddings } from "@langchain/community/embeddings/llama_cpp";

const llamaPath = "/Replace/with/path/to/your/model/gguf-llama2-q4_0.bin";

const embeddings = new LlamaCppEmbeddings({
modelPath: llamaPath,
});

const res = embeddings.embedQuery("Hello Llama!");

console.log(res);

/*
[ 15043, 365, 29880, 3304, 29991 ]
*/

API Reference:

Document embedding

import { LlamaCppEmbeddings } from "@langchain/community/embeddings/llama_cpp";

const llamaPath = "/Replace/with/path/to/your/model/gguf-llama2-q4_0.bin";

const documents = ["Hello World!", "Bye Bye!"];

const embeddings = new LlamaCppEmbeddings({
modelPath: llamaPath,
});

const res = await embeddings.embedDocuments(documents);

console.log(res);

/*
[ [ 15043, 2787, 29991 ], [ 2648, 29872, 2648, 29872, 29991 ] ]
*/

API Reference:


Was this page helpful?


You can also leave detailed feedback on GitHub.