Overview

This course features Coursera Coach! A smarter way to learn with interactive, real-time conversations that help you test your knowledge, challenge assumptions, and deepen your understanding as you progress through the course. Delve into the world of generative AI and large language models (LLMs) with hands-on applications using Python. You'll explore the power of Variational Auto-Encoders (VAEs) and Generative Adversarial Networks (GANs) to create synthetic data, including images and music. Alongside, you'll get to grips with Transformers and self-attention mechanisms, which are foundational to models like GPT and ChatGPT, unlocking advanced AI applications. Learn the intricacies of GPT architecture, including tokenization and fine-tuning, and apply these concepts using tools like Hugging Face and Google Colab. The course also covers cutting-edge topics such as Retrieval Augmented Generation (RAG) and advanced LLM agents. Through interactive activities, you’ll create powerful AI applications like chatbots and personalized systems. This course is designed for learners aiming to advance their knowledge of AI, machine learning, and Python, with a focus on generative models and LLMs. If you want to build your own AI-driven applications and deepen your understanding of state-of-the-art AI technologies, this course is for you.

Syllabus

Generative Models

In this module, we will explore the inner workings of Variational Auto-Encoders (VAEs) and Generative Adversarial Networks (GANs), two powerful generative models. You'll gain both theoretical knowledge and practical skills through hands-on exercises and demonstrations, specifically with the Fashion MNIST dataset. By the end, you'll be able to effectively understand and implement these models for generative tasks in deep learning.

Generative AI: GPT, ChatGPT, Transformers, Self-Attention Based Neural Networks

In this module, we will dive deep into the workings of Transformer-based architectures, exploring essential concepts like self-attention, masked attention, and multi-headed attention. You'll learn how models like GPT function, focusing on tokenization, positional encoding, and fine-tuning. Additionally, hands-on activities will guide you through real-world applications, such as fine-tuning GPT models and exploring the transition from GPT to ChatGPT with reinforcement learning techniques.

The OpenAI API (Developing with GPT and ChatGPT)

In this module, we will explore a range of OpenAI APIs, guiding you through the process of integrating and utilizing APIs such as chat completions, image generation, embeddings, and audio processing. You will also learn how to fine-tune GPT models for custom tasks and use moderation tools. Practical activities will give you hands-on experience in developing with OpenAI’s APIs, providing the skills necessary to build advanced applications using GPT and ChatGPT technologies.

Retrieval Augmented Generation (RAG), Advanced RAG, and LLM Agents

In this module, we will dive into Retrieval Augmented Generation (RAG) and its advanced methods, focusing on improving generative outputs through retrieval-based techniques and fine-tuning strategies. You will explore key metrics like precision, recall, and relevancy, while engaging in hands-on activities using RAG and langchain to simulate data and create practical solutions. Additionally, we will explore the concept of LLM agents, building integrated systems such as chatbots with web search and math tools.

Final Project

In this module, we will focus on your final project, where you'll apply your learning to classify mammogram images. The assignment will guide you through the process, from designing the model to evaluating the performance. In the final review, you'll assess your project and receive feedback to further refine your skills.

You Made It!

In this module, we will look at the exciting opportunities ahead of you. You'll get recommendations on valuable resources, including books and websites, to continue expanding your knowledge in data science. Additionally, we’ll provide career advice to help you confidently step into the data science field and apply what you've learned to real-world projects and job prospects.