This repository accompanies a comprehensive 8-week LLM Engineering course, designed for individuals seeking to master AI and LLMs through hands-on project building. It offers a structured learning path, progressing from foundational concepts to advanced agentic AI solutions, with a strong emphasis on practical application and skill development.

How It Works

The course leverages Ollama for immediate local LLM interaction, enabling users to run models like Llama 3.2 directly on their machines. For more intensive tasks, it integrates Google Colab, providing access to GPUs and facilitating the use of various LLM APIs and Hugging Face libraries. The curriculum emphasizes a "learning by doing" approach, encouraging users to run, inspect, and modify code examples to deepen their understanding.

Quick Start & Requirements

• Install Ollama: Download and install from https://ollama.com.
• Run Llama 3.2: Execute ollama run llama3.2 or ollama run llama3.2:1b for smaller machines.
• Prerequisites: Python, Ollama. Google account for Colab.
• Optional: API keys for services like OpenAI, Anthropic, Google Gemini.
• Resources: Course slides and links available at https://edwarddonner.com/2024/11/13/llm-engineering-resources/.
• Colab Notebooks: Provided within week-specific folders and linked in resources.

Highlighted Details

• Focuses on practical LLM engineering skills applicable to business.
• Includes a final project building an autonomous agentic AI solution.
• Provides guidance on managing API costs and offers free alternatives.
• Encourages community contributions via Pull Requests.

Maintenance & Community

The instructor, Ed Donner, is accessible via email (ed@edwarddonner.com) and LinkedIn (https://www.linkedin.com/in/eddonner/) for support and feedback.

Licensing & Compatibility

The repository's licensing is not explicitly stated in the provided README.

Limitations & Caveats

The README strongly advises against using Llama 3.3 due to its large parameter size (70B), recommending Llama 3.2 for home computer compatibility. Some cloud services used in the course may incur minimal costs, though free alternatives are provided.