Skip to main content

IBM watsonx.ai

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

Overview​

Integration details​

ClassPackageLocalPy supportPackage downloadsPackage latest
IBM watsonx.ai@langchain/community❌✅NPM - DownloadsNPM - Version

Setup​

To access IBM WatsonxAI embeddings you’ll need to create an IBM watsonx.ai account, get an API key or any other type of credentials, and install the @langchain/community integration package.

Credentials​

Head to IBM Cloud to sign up to IBM watsonx.ai and generate an API key or provide any other authentication form as presented below.

IAM authentication​

export WATSONX_AI_AUTH_TYPE=iam
export WATSONX_AI_APIKEY=<YOUR-APIKEY>

Bearer token authentication​

export WATSONX_AI_AUTH_TYPE=bearertoken
export WATSONX_AI_BEARER_TOKEN=<YOUR-BEARER-TOKEN>

CP4D authentication​

export WATSONX_AI_AUTH_TYPE=cp4d
export WATSONX_AI_USERNAME=<YOUR_USERNAME>
export WATSONX_AI_PASSWORD=<YOUR_PASSWORD>
export WATSONX_AI_URL=<URL>

Once these are placed in your environment variables and object is initialized authentication will proceed automatically.

Authentication can also be accomplished by passing these values as parameters to a new instance.

IAM authentication​

import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";

const props = {
  version: "YYYY-MM-DD",
  serviceUrl: "<SERVICE_URL>",
  projectId: "<PROJECT_ID>",
  watsonxAIAuthType: "iam",
  watsonxAIApikey: "<YOUR-APIKEY>",
};
const instance = new WatsonxEmbeddings(props);

Bearer token authentication​

import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";

const props = {
  version: "YYYY-MM-DD",
  serviceUrl: "<SERVICE_URL>",
  projectId: "<PROJECT_ID>",
  watsonxAIAuthType: "bearertoken",
  watsonxAIBearerToken: "<YOUR-BEARERTOKEN>",
};
const instance = new WatsonxEmbeddings(props);

CP4D authentication​

import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";

const props = {
  version: "YYYY-MM-DD",
  serviceUrl: "<SERVICE_URL>",
  projectId: "<PROJECT_ID>",
  watsonxAIAuthType: "cp4d",
  watsonxAIUsername: "<YOUR-USERNAME>",
  watsonxAIPassword: "<YOUR-PASSWORD>",
  watsonxAIUrl: "<url>",
};
const instance = new WatsonxEmbeddings(props);

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 IBM watsonx.ai integration lives in the @langchain/community package:

tip

See this section for general instructions on installing integration packages.

yarn add @langchain/community @langchain/core

Instantiation​

Now we can instantiate our model object and embed text:

import { WatsonxEmbeddings } from "@langchain/community/embeddings/ibm";

const embeddings = new WatsonxEmbeddings({
  version: "YYYY-MM-DD",
  serviceUrl: process.env.API_URL,
  projectId: "<PROJECT_ID>",
  spaceId: "<SPACE_ID>",
  idOrName: "<DEPLOYMENT_ID>",
  model: "<MODEL_ID>",
});

Note:

  • You must provide spaceId, projectId or idOrName(deployment id) in order to proceed.
  • Depending on the region of your provisioned service instance, use correct serviceUrl.

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);
singleVector.slice(0, 100);
[
   -0.017436018,   -0.01469498,   -0.015685871,  -0.013543149, -0.0011519607,
   -0.008123747,   0.015286108,   -0.023845721,   -0.02454774,    0.07235078,
   -0.032333843, -0.0035843418,   -0.015389036,     0.0455373,  -0.021119863,
   -0.022039745,   0.021746712,   -0.017774817,  -0.008232582,  -0.036727764,
   -0.015734928,    0.03606811,   -0.005108186,  -0.036052454,   0.024462992,
     0.02359307,    0.03273164,    0.009195497, -0.0077208397,    -0.0127943,
   -0.023869334,  -0.029473905,  -0.0080457395, -0.0021337876,    0.04949132,
    0.013950589,  -0.010046689,    0.021029025,  -0.031725302,   0.004251065,
   -0.034171984,   -0.03696642,   -0.014253629,  -0.017757406,  -0.007531065,
     0.07187789,   0.009661725,    0.041889492,   -0.04660478,   0.028036641,
    0.059334517,   -0.04561291,    0.056029715,   -0.00676024,   0.026493236,
      0.0116374,   0.050126843,   -0.018036349,  -0.013711887,   0.042252757,
    -0.04453391,    0.04705777, -0.00044598224,  -0.030227259,   0.029286578,
      0.0252211,   0.011694125,   -0.031404093,    0.02951232,    0.08812359,
    0.023539362,  -0.011082862,    0.008024676, 0.00084492035,  -0.007984158,
  -0.0005008702,  -0.025189219,    0.021000557, -0.0065513053,   0.036524914,
   0.0015150858, -0.0042383806,    0.049065087,   0.000941666,    0.04447001,
    0.012942205,  -0.078316726,    -0.03004237,  -0.025807172,   -0.03446275,
    -0.00932942,  -0.044925686,     0.03190307,   0.010136769,  -0.048854534,
    0.025738232,  -0.017840309,    0.023738133,   0.014214792,   0.030452395
]

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));
[
   -0.017436024,  -0.014695002,    -0.01568589,  -0.013543164, -0.001151976,
   -0.008123703,   0.015286064,   -0.023845702,  -0.024547677,   0.07235076,
   -0.032333862, -0.0035843418,   -0.015389038,   0.045537304, -0.021119865,
    -0.02203975,   0.021746716,    -0.01777481,  -0.008232588,  -0.03672781,
   -0.015734889,   0.036068108,     -0.0051082,  -0.036052432,  0.024462998,
    0.023593083,    0.03273162,    0.009195521,  -0.007720828, -0.012794304,
   -0.023869323,  -0.029473891,   -0.008045726,  -0.002133793,  0.049491342,
    0.013950573,  -0.010046691,     0.02102898,   -0.03172528, 0.0042510596,
   -0.034171965,  -0.036966413,   -0.014253668,  -0.017757434, -0.007531062,
     0.07187787,   0.009661732,    0.041889492,   -0.04660476,  0.028036654,
    0.059334517,  -0.045612894,    0.056029722,   -0.00676024,  0.026493296,
   0.0116374055,   0.050126873,   -0.018036384,  -0.013711868,    0.0422528,
   -0.044533912,   0.047057763, -0.00044596897,  -0.030227251,  0.029286569,
    0.025221113,   0.011694138,    -0.03140413,   0.029512335,   0.08812357,
    0.023539348,  -0.011082865,    0.008024677, 0.00084490055, -0.007984145,
  -0.0005008745,  -0.025189226,    0.021000564, -0.0065513197,  0.036524955,
   0.0015150585, -0.0042383634,    0.049065102,   0.000941638,  0.044469994,
    0.012942193,  -0.078316696,     -0.0300424,  -0.025807157,   -0.0344627,
   -0.009329439,   -0.04492573,    0.031903077,   0.010136808, -0.048854522,
    0.025738247,   -0.01784033,    0.023738142,   0.014214801,  0.030452369
]
[
     0.03278884, -0.017893745, -0.0027520044,   0.016506646,   0.028271576,
    -0.01284331,  0.014344065,  -0.007968607,   -0.03899479,   0.039327156,
   -0.047726233,  0.009559004,   -0.05302522,   0.011498492, -0.0055542476,
  -0.0020940166, -0.029262392,  -0.025919685,   0.024261741, -0.0010863725,
   0.0074619935,  0.014191284,  -0.009054746,  -0.038633537,   0.039744128,
    0.012625762,  0.030490868,   0.013526139,  -0.024638629,  -0.011268263,
   -0.012759613,  -0.04693565,  -0.013087251,   -0.01971696,  0.0125782555,
    0.024156926, -0.011638484,   0.017364893,    -0.0405832, -0.0032466082,
    -0.01611277, -0.022583133,   0.019492855,   -0.03664484,  -0.022627067,
    0.011026938, -0.014631298,   0.043255687,  -0.029447634,   0.017212389,
    0.029366229, -0.041978795,   0.005347565, -0.0106230285,  -0.008334342,
   -0.008841154,  0.045096103,    0.03996879,  -0.002039457, -0.0051824683,
   -0.019464444,  0.092018366,  -0.009283633,  -0.020052811,  0.0043408144,
   -0.029403884,   0.02587689,  -0.027253918,     0.0159064,     0.0421537,
     0.05078811, -0.012380686,  -0.018032575,    0.01711449,    0.03636163,
   -0.014590949, -0.015076142, 0.00018201554,   0.002490666,   0.044776678,
     0.05301749, -0.007891316,   0.028668318, -0.0016632816,    0.04487743,
   -0.032529455, -0.040372133,  -0.020566158,  -0.011109745,   -0.01724949,
  -0.0047519016, -0.041635286,  0.0068111843,   0.039498538,   -0.02491227,
    0.016853934, -0.017926402,  -0.006154979,   0.025893573,   0.015262395
]

API reference​

For detailed documentation of all module_name features and configurations head to the API reference: api_ref_module

Was this page helpful?


You can also leave detailed feedback on GitHub.