Building with LLMs Made Simple

A tutorial on how to build Python programs with LLMs enhanced in the mix.

Made with ❤️ by Eric J. Ma (@ericmjl).

Installation Instructions

This tutorial requires several components to be installed:

0. Install Pixi

Pixi is a package management tool that we'll use to manage dependencies. Install it using one of these methods:

Linux & macOS (using curl):

curl -fsSL https://pixi.sh/install.sh | bash

Windows (using PowerShell):

iwr -useb https://pixi.sh/install.ps1 | iex

After installation, you may need to restart your terminal for the changes to take effect.

1. Install Ollama

Ollama is required to run the local LLM models used in this tutorial.

• Linux: Visit https://ollama.com/download/linux
• macOS: Visit https://ollama.com/download/mac
• Windows: Visit https://ollama.com/download/windows

Follow the installation instructions for your platform.

2. Run the Installation Script

After installing Ollama, run the following command to:

• Pull the required LLM models
• Install uv (Python package manager)
• Set up a virtual environment
• Install all required dependencies

pixi run start

The command will:

1. Pull the following Ollama models:
   • llama3.2
   • phi4
   • gemma2:2b
2. Install uv if not already installed
3. Create a Python virtual environment
4. Install all required dependencies

Please do this before arrival to the tutorial session, as they may take some time to download!

3. Running the Notebooks

After installation is complete, you can run the notebooks using:

# Run the first notebook
uvx marimo edit --sandbox notebooks/01_simple_bot.py

# Or run the second notebook
uvx marimo edit --sandbox notebooks/02_structured_bot.py

Manual Installation (Alternative)

If you prefer to install components manually:

1. Install Ollama from https://ollama.com/download

2. Pull the required models:

   ollama pull llama3.2
   ollama pull phi4
   ollama pull gemma2:2b

3. Install uv:

   curl -fsSL https://astral.sh/uv/install.sh | bash

   The installation guide is here, and includes Windows installation instructions too: https://docs.astral.sh/uv/getting-started/installation/

4. Run a notebook:

   cd notebooks/ # super duper important!
   uvx marimo edit --sandbox 01_simple_bot.py # or any other notebook

Troubleshooting

• Ollama Model Download Issues: If you encounter issues downloading models, ensure you have a stable internet connection and sufficient disk space.
• uv Installation Problems: If uv installation fails, you can try installing it using pip: pip install uv
• Notebook Errors: Ensure all dependencies are correctly installed and that Ollama is running in the background.