Heavily inspired by Harrison Chase's LangChain PY 0.1.0 guide repo

These notebooks show off different functionality associated with LangChain.js 0.1.0.

To see the LangChain.js documentation, click here.

To setup, set required environment variables. We will use OpenAI for our language model, and Tavily for our search provider.

export OPENAI_API_KEY=...
export TAVILY_API_KEY=...

We will also use LangSmith for observability:

export LANGCHAIN_TRACING_V2="true"
export LANGCHAIN_API_KEY=...
export LANGCHAIN_ENDPOINT="https://api.smith.langchain.com"

Next, you'll need Deno installed, and their experimental Jupyter notebook support. Follow the instructions here.

The following notebooks are provided:

• Observability
• Composition
• Streaming
• Output Parsing
• Retrieval
• Agents