Integrating Lang Chain with Azure OpenAI

Table of Contents

1. Introduction
2. Setting Up Azure Open AI
3. Integrating Lang Chain
4. Importing Required Packages
5. Reading Text Files
6. Splitting Text using Text Splitter
7. Working on Embeddings
8. Creating Vector Store
9. Constructing the Chain for Retrieval qa
10. Querying the Model
11. Conclusion

Introduction

In this article, we will explore how to use Lang chain with Azure Open AI. We will start with a simple example where we read a text file and query the contents using Azure Open AI. We will go step by step setting up the necessary properties, importing the required packages, reading the text file, splitting the text using a text splitter, working on embeddings, creating a vector store, constructing the chain for retrieval QA, and finally querying the model. By the end of this article, you will have a clear understanding of how to integrate Lang chain with Azure Open AI.

1. Setting Up Azure Open AI

Before we begin, we need to set up Azure Open AI. We will need the API key and the base URL, which can be obtained from the Azure portal. If you haven't created an instance on Azure, I recommend you to watch the Relevant video where I explain the process in detail.

2. Integrating Lang Chain

Once we have the necessary properties set, we can proceed with integrating Lang chain. We will be using the text splitter from Lang chain to split our input text. Let's import the required Package and get started.

3. Importing Required Packages

To work with Lang chain and Azure Open AI, we need to import the necessary packages. We will import the character text splitter from Lang chain, the open AI embeddings, and Azure Open AI.

4. Reading Text Files

In this step, we will read the text file that we want to query. We will use the 'read' function to read the contents of the file. Make sure to replace the file name with the name of your text file.

5. Splitting Text using Text Splitter

To split the text into chunks, we will use the text splitter from Lang chain. We will set the chunk size to 5, but you can adjust it based on your input data. The text splitter will split the text and store it in the 'text' variable.

6. Working on Embeddings

Next, we will work on generating embeddings for the text. We will instantiate the embeddings variable using the open AI embeddings. We will also create a vector store using Chroma to store the embeddings.

7. Creating Vector Store

To create a vector store, we will use Chroma and pass the text and embeddings as parameters. This will generate the required embeddings for our text.

8. Constructing the Chain for Retrieval QA

In this step, we will construct the chain for retrieval QA. We will pass the Azure open AI engine and the retriever to the chain. Make sure to replace the engine name with the appropriate engine for your model.

9. Querying the Model

Finally, we can query the model by running our question through the chain. We will pass the question "Where does a homeless person stay?" to the chain and retrieve the answer.

Conclusion

In this article, we have seen how to integrate Lang chain with Azure Open AI. We have set up the necessary properties, imported the required packages, read a text file, split the text using a text splitter, worked on embeddings, created a vector store, constructed the chain for retrieval QA, and queried the model. This is just a basic example, but it should give you an idea of how to integrate Lang chain with Azure Open AI. Feel free to explore and extend this example further.