langchainjs-example-2024

NOTE: A presentation(pdf) which can be read along with example here can be found in the root of the repo.

This repository includes some examples that cover different usages of langchainJS (and open AI's LLM):

• 1-simple-example-and-chatbot A quick and simple tutorial on how to get started with Langchain JS. How to build a chatbot with nodeJS and langchain. Introducing the concept of chain and adding memory to the chatbot

• 2-simple-rag.ts How to build a simple RAG system by reading a pdf (in the example, senior developer salary in Finland). Try to ask question like "what is the average salary of senior developer in Finland in 2024 ?"

• 4-agent.ts How to enable LLM to do more with agents and tools. This example give the LLM a simple tool call "mimmitKoodaaInfo" and "nordcloudInfo" which allows the LLM to get more information about Mimmit Koodaa event and the speaker information (employee of Nordcloud). Try to ask question like "what happen in mimitkooda on 19th of April ? Also tell me a bit about the speaker".

How to run an example:

• Run npm i
• Run npm i -g ts-node --> remove the -g if you don't want to install ts-node in global level
• Create .env file, check out Environment variables section below
• Run ts-node src/<example_file_name.ts>

Environment variables

The example use dotenv to load environment variable. Remember to create a .env in the root and add a following keys NOTE: Only use either Open AI or Azure setup

To use Open AI hosted in Azure

AZURE_OPENAI_API_VERSION=2023-07-01-preview
AZURE_OPENAI_BASE_PATH=https://oai-int-azg-we-001.openai.azure.com/openai/deployments
AZURE_OPENAI_API_KEY=<fill-in>
AZURE_OPENAI_API_DEPLOYMENT_NAME=dep-gpt-35-turbo
AZURE_OPENAI_API_EMBEDDINGS_DEPLOYMENT_NAME=text-embedding-ada-002

To use Open AI hosted in Open AI

OPENAI_API_KEY=<fill-in>

For Pinecone Vector DB

PINECONE_KEY=<fill in>
PINECONE_ENVIRONMENT=gcp-starter
PINECONE_INDEX=internal-demo