Class 8: LangServe

NOTE

The folowing instructions are for the demo directory meant to showcase live examples in class. For the main deployment example on Render, see the DEPLOYMENT README.

Slides

Slides

Prerequisites

Before you begin, ensure you have met the following requirements:

• Docker Desktop (recommended for local setup)
• Python 3.11.0 or greater (local setup with virtual environment)

Set up environment variables:

• Copy the sample environment file:

  cp .env.sample .env

• Edit the .env file:

  OPENAI_API_KEY=your-openai-key
  LANGCHAIN_API_KEY=your-langchain-key
  LANGCHAIN_TRACING_V2=true
  LANGCHAIN_PROJECT=class_8
  

Docker (recommended)

1. In the CLI, cd into the demo directory. You MUST follow this step before proceeding to access the correct Dockerfile/docker-compose.yml.
2. Look at the docker-compose.yml and Dockerfile file to understand the different services they are managing/running.
3. Start the Python servers in in_class_examples in their respective directories (each command runs its separate server container, and is run from the demo directory):

   docker compose up agent
   docker compose up chat
   

4. Start Jupyter to run the corresponding .ipynb files you may create:

   docker compose up jupyter
   

5. Run any .py files you may create in the demo directory in this manner:

   docker compose run --rm main python <the_py_file>
   

Running Different Scripts

You can use the provided run.sh script for easier execution. Make sure you are in the demo directory when running the script. Make sure to make the script executable with chmod +x run.sh in the CLI before using:

./run.sh agent #(starts the agent server)
./run.sh chat #(starts the chat server)
./run.sh jupyter #(starts the jupyter notebook server)
./run.sh <your_new_py_file> #(runs other .py files)

Local Setup (Alternative to Docker)

If you prefer to run the examples locally:

1. Ensure you have Python 3.11.0 or higher installed.
2. Clone the repository:

   git clone [repository-url]
   cd [repository-name]

3. Set up the virtual environment:

   python3 -m venv .venv
   source .venv/bin/activate  # On Windows use `.venv\Scripts\activate`
   pip install -r requirements.txt

4. Configure environment variables as described in the Setup section.
5. Export your .env variables to the system (LangChain will handle this for you "under the hood" in the .ipynb files, but this is included for reference): Linux / Mac / Bash

   export $(grep -v '^#' .env | xargs)

6. Run the notebooks:

   run the `.ipynb` files in VSCode (it will prompt you to allow the installation of ipykernel: do so) or another IDE that supports notebooks.
   

Need Help?

Reach out to the learning assistants in the course Slack channel.