Skip to main content

OllamaEmbeddings

This will help you get started with Ollama embedding models using LangChain. For detailed documentation on OllamaEmbeddings features and configuration options, please refer to the API reference.

Overview​

Integration details​

ClassPackageLocalPy supportPackage downloadsPackage latest
OllamaEmbeddings@langchain/ollamaβœ…βœ…NPM - DownloadsNPM - Version

Setup​

To access Ollama embedding models you’ll need to follow these instructions to install Ollama, and install the @langchain/ollama integration package.

Credentials​

If you want to get automated tracing of your model calls you can also set your LangSmith API key by uncommenting below:

# export LANGCHAIN_TRACING_V2="true"
# export LANGCHAIN_API_KEY="your-api-key"

Installation​

The LangChain OllamaEmbeddings integration lives in the @langchain/ollama package:

yarn add @langchain/ollama @langchain/core

Instantiation​

Now we can instantiate our model object and embed text:

import { OllamaEmbeddings } from "@langchain/ollama";

const embeddings = new OllamaEmbeddings({
model: "mxbai-embed-large", // Default value
baseUrl: "http://localhost:11434", // Default value
});

Indexing and Retrieval​

Embedding models are often used in retrieval-augmented generation (RAG) flows, both as part of indexing data as well as later retrieving it. For more detailed instructions, please see our RAG tutorials under the working with external knowledge tutorials.

Below, see how to index and retrieve data using the embeddings object we initialized above. In this example, we will index and retrieve a sample document using the demo MemoryVectorStore.

// Create a vector store with a sample text
import { MemoryVectorStore } from "langchain/vectorstores/memory";

const text =
"LangChain is the framework for building context-aware reasoning applications";

const vectorstore = await MemoryVectorStore.fromDocuments(
[{ pageContent: text, metadata: {} }],
embeddings
);

// Use the vector store as a retriever that returns a single document
const retriever = vectorstore.asRetriever(1);

// Retrieve the most similar text
const retrievedDocuments = await retriever.invoke("What is LangChain?");

retrievedDocuments[0].pageContent;
LangChain is the framework for building context-aware reasoning applications

Direct Usage​

Under the hood, the vectorstore and retriever implementations are calling embeddings.embedDocument(...) and embeddings.embedQuery(...) to create embeddings for the text(s) used in fromDocuments and the retriever’s invoke operations, respectively.

You can directly call these methods to get embeddings for your own use cases.

Embed single texts​

You can embed queries for search with embedQuery. This generates a vector representation specific to the query:

const singleVector = await embeddings.embedQuery(text);

console.log(singleVector.slice(0, 100));
[
0.026051683, 0.029081265, -0.040726297, -0.015116953, -0.010691089,
0.030181013, -0.0065084146, -0.02079503, 0.013575795, 0.03452527,
0.009578291, 0.007026421, -0.030110886, 0.013489622, -0.04294787,
0.011141899, -0.043768786, -0.00362867, -0.0081198225, -0.03426076,
0.010075142, 0.027787417, -0.09052663, -0.06039698, -0.009462592,
0.06232288, 0.051121354, 0.011977532, 0.089046724, 0.059000008,
0.031860664, -0.034242127, 0.020339863, 0.011483523, -0.05429335,
-0.04963588, 0.03263794, -0.05581542, 0.013908403, -0.012356067,
-0.007802118, -0.010027855, 0.00281217, -0.101886116, -0.079341754,
0.011269771, 0.0035983133, -0.027667878, 0.032092705, -0.052843474,
-0.045283325, 0.0382421, 0.0193055, 0.011050924, 0.021132186,
-0.037696265, 0.0006107435, 0.0043520257, -0.028798066, 0.049155913,
0.03590549, -0.0040995986, 0.019772101, -0.076119535, 0.0031298609,
0.03368174, 0.039398745, -0.011813277, -0.019313531, -0.013108803,
-0.044905286, -0.022326004, -0.01656178, -0.06658457, 0.016789088,
0.049952697, 0.006615693, -0.01694402, -0.018105473, 0.0049101883,
-0.004966945, 0.049762275, -0.03556957, -0.015986584, -0.03190983,
-0.05336687, -0.0020468342, -0.0016106658, -0.035291273, -0.029783724,
-0.010153295, 0.052100364, 0.05528949, 0.01379487, -0.024542747,
0.028773975, 0.010087022, 0.030448131, -0.042391222, 0.016596776
]

Embed multiple texts​

You can embed multiple texts for indexing with embedDocuments. The internals used for this method may (but do not have to) differ from embedding queries:

const text2 =
"LangGraph is a library for building stateful, multi-actor applications with LLMs";

const vectors = await embeddings.embedDocuments([text, text2]);

console.log(vectors[0].slice(0, 100));
console.log(vectors[1].slice(0, 100));
[
0.026051683, 0.029081265, -0.040726297, -0.015116953, -0.010691089,
0.030181013, -0.0065084146, -0.02079503, 0.013575795, 0.03452527,
0.009578291, 0.007026421, -0.030110886, 0.013489622, -0.04294787,
0.011141899, -0.043768786, -0.00362867, -0.0081198225, -0.03426076,
0.010075142, 0.027787417, -0.09052663, -0.06039698, -0.009462592,
0.06232288, 0.051121354, 0.011977532, 0.089046724, 0.059000008,
0.031860664, -0.034242127, 0.020339863, 0.011483523, -0.05429335,
-0.04963588, 0.03263794, -0.05581542, 0.013908403, -0.012356067,
-0.007802118, -0.010027855, 0.00281217, -0.101886116, -0.079341754,
0.011269771, 0.0035983133, -0.027667878, 0.032092705, -0.052843474,
-0.045283325, 0.0382421, 0.0193055, 0.011050924, 0.021132186,
-0.037696265, 0.0006107435, 0.0043520257, -0.028798066, 0.049155913,
0.03590549, -0.0040995986, 0.019772101, -0.076119535, 0.0031298609,
0.03368174, 0.039398745, -0.011813277, -0.019313531, -0.013108803,
-0.044905286, -0.022326004, -0.01656178, -0.06658457, 0.016789088,
0.049952697, 0.006615693, -0.01694402, -0.018105473, 0.0049101883,
-0.004966945, 0.049762275, -0.03556957, -0.015986584, -0.03190983,
-0.05336687, -0.0020468342, -0.0016106658, -0.035291273, -0.029783724,
-0.010153295, 0.052100364, 0.05528949, 0.01379487, -0.024542747,
0.028773975, 0.010087022, 0.030448131, -0.042391222, 0.016596776
]
[
0.0558515, 0.028698817, -0.037476595, 0.0048659276, -0.019229038,
-0.04713716, -0.020947812, -0.017550547, 0.01205507, 0.027693441,
-0.011791304, 0.009862203, 0.019662278, -0.037511427, -0.022662448,
0.036224432, -0.051760387, -0.030165697, -0.008899774, -0.024518963,
0.010077767, 0.032209765, -0.0854303, -0.038666975, -0.036021013,
0.060899545, 0.045867186, 0.003365381, 0.09387081, 0.038216405,
0.011449426, -0.016495887, 0.020602569, -0.02368503, -0.014733645,
-0.065408126, -0.0065152845, -0.027103946, 0.00038956117, -0.08648814,
0.029316466, -0.054449145, 0.034129277, -0.055225655, -0.043182302,
0.0011148591, 0.044116337, -0.046552557, 0.032423045, -0.03269365,
-0.05062933, 0.021473562, -0.011019348, -0.019621233, -0.0003149565,
-0.0046085776, 0.0052610254, -0.0029293327, -0.035793293, 0.034469575,
0.037724957, 0.009572597, 0.014198464, -0.0878237, 0.0056973165,
0.023563445, 0.030928325, 0.025520306, 0.01836824, -0.016456697,
-0.061934732, 0.009764942, -0.035812028, -0.04429064, 0.031323086,
0.056027107, -0.0019782048, -0.015204176, -0.008684945, -0.0010460864,
0.054642987, 0.044149086, -0.032964867, -0.012044753, -0.019075096,
-0.027932597, 0.018542245, -0.02602878, -0.04645578, -0.020976603,
0.018999187, 0.050663687, 0.016725155, 0.0076955976, 0.011448177,
0.053931057, -0.03234989, 0.024429373, -0.023123834, 0.02197912
]

Ollama model parameters are also supported:

import { OllamaEmbeddings } from "@langchain/ollama";

const embeddingsCustomParams = new OllamaEmbeddings({
requestOptions: {
useMmap: true, // use_mmap 1
numThread: 6, // num_thread 6
numGpu: 1, // num_gpu 1
},
});

API reference​

For detailed documentation of all OllamaEmbeddings features and configurations head to the API reference


Was this page helpful?


You can also leave detailed feedback on GitHub.