Jupyter Notebooks Execution

The course includes a set of Jupyter notebooks for hands-on labs and homework exercises.

Prerequisites

The labs and exercises use APIs from OpenAI, Anthropic, and Cohere. You will need to get API keys from each to execute them. A secure and simple way to use the API keys is to create a local .env file and put the keys, for example:

OPENAI_API_KEY=sk-proj-###
ANTHROPIC_API_KEY=sk-ant-###
CO_API_KEY=3lljis###

The notebooks get access to the API keys using:

from dotenv import load_dotenv
load_dotenv()

Notebook Execution Options

You have a few options to execute the Jupyter notebooks:

1. Local Jupyter Notebook Server

Set up your local environment to run the Jupyter notebooks. Follow these steps:

git clone https://github.com/guyernest/advanced-rag.git
cd advanced-rag
pip install uv
uv venv --python cpython-3.12.3-macos-aarch64-none # Create a virtual environment with python 3.12 for MacOS with Apple Silicon (adapt as needed to your machine)
source .venv/bin/activate
uv pip compile requirements.in --output-file requirements.txt # Create the requirements.txt file for the environment 
uv pip install requirements.txt # Install the required dependencies in the environment
uv pip install jupyterlab ipykernel # Install Jupyter Lab and ipykernel to manager kernels for Jupyter
python -m ipykernel install --user --name=.venv --display-name="Python (.venv)" # Create kernel from the virtual environment
jupyter lab 

Choose the recently created kernel from the kernel list on the top right of Jupyter lab:
Jupyter Lab Kernel Selection

Please note that the instructions above are using , which speeds up dependency management. It supports Windows, MacOS, and Linux. Check the detailed instructions for its installation and usage here: https://github.com/astral-sh/uv.

2. Jupyter Notebook Support within Visual Studio Code

Visual Studio Code (VSC) has built-in support for Jupyter notebooks. Follow these steps:

git clone https://github.com/guyernest/advanced-rag.git
cd advanced-rag
code .

Once VSC is opened in the course directory, open a terminal and execute the following commands:

pip install uv
uv venv --python 3.11.0
source .venv/bin/activate
uv pip compile requirements.in --output-file requirements.txt
uv pip install requirements.txt

Alternatively, you can use the traditional pip route by:

python -m venv myenv
source myenv/bin/activate
pip install -r requirements.txt

If your VSC doesn’t have Jupyter support yet, you will need to install ipykernel:

curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py # get the pip package
python get-pip.py # Install the pip package
python -m pip install ipykernel # Install ipykernel to add the support for Jupyter notebooks to VSC