A hands-on course for building AI agents from scratch using Python and Llama.
This course teaches you how to build various types of AI agents using large language models (LLMs), with a focus on practical implementation and no external frameworks. All code is written from scratch to ensure deep understanding.
- Python programming knowledge
- Local setup of Ollama with Llama 3.2
- No external AI/ML frameworks required
computer_use_coldplay.mp4
- Deploying LLMs: Getting started with local LLM deployment and interaction
- React-based Agents: Building agents that can reason and act
- JSON-based Agents: Creating structured agents with JSON outputs
- Code Execution Agents: Developing agents that can write and run code
- Math Agent: An agent that can solve mathematical problems
- API Use Agent: Agent for interacting with external APIs
- Browser Use Agent: Agent capable of web browsing tasks
- Computer Use Agent: Agent for computer automation tasks
- Snack & Stream Planner (ReAct): An agent that plans movie nights with snack pairings
- Snack & Stream Planner (JSON): JSON-based implementation of the movie night planner
- Password Strength Checker - Code Execution Agent: Agent for evaluating password security
- Course Whisperer - Browser Use Agent: Agent for course recommendations and learning paths
- Computer use Agent: Agent for visual recognition and computer control
- Install Ollama and download Llama 3.2
- Clone this repository
- Follow along with the course modules
- Build your own agents!
- Each module contains hands-on exercises
- No external frameworks - you'll write everything from scratch
- Focus on practical implementation
- Live streamed lessons available online
This is a learning repository. Feel free to submit PRs for improvements or bug fixes.
Created by James Murdza (@jamesmurdza)