Welcome to the Generative AI Course at Northeastern University! This repository contains the labs and reading materials designed to help you grasp the concepts and applications of generative AI. Throughout this course, you'll explore foundational and advanced techniques, gain practical experience through hands-on labs, and delve into various generative models. You can learn more on Machine learning topics by watching my videos on Youtube or visit my Website.
Generative AI, or GenAI, refers to a class of artificial intelligence technologies that can generate new content, ranging from text and images to music and code, based on patterns it learns from vast amounts of data. This technology uses advanced machine learning models, such as neural networks, to understand the underlying structure and elements of the input data and then uses this understanding to create original, plausible outputs that are similar in nature but novel in content. GenAI holds significant potential across various fields, including creative arts, where it can assist in designing unique artworks, in software development, where it can write code, and in business, where it can generate reports or innovative product ideas. Its ability to automate and enhance creative processes makes it a valuable tool for boosting productivity and fostering innovation.
The Deep Generative Models course at Northeastern University is designed to provide students with a comprehensive understanding of the Gen-AI field. Throughout the course, students will learn how to:
- Introductory Materials: Neural Networks (NNs), Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs) — foundational concepts and applications.
- Transformers and AutoRegressive Models: Learn advanced sequence modeling and generation techniques.
- Energy-Based Models and Score-Based Models: Explore models that learn the probability distribution of data.
- Flow Normalization: Understand techniques for improving model training stability and performance.
- Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs): Explore foundational techniques to understand their mechanisms and applications.
- Diffusion Models: Dive into cutting-edge advancements for high-quality image synthesis through hands-on projects.
- Reinforcement Learning with Human Feedback (RLHF): Learn to implement RLHF techniques to optimize generative processes with human input.
- Graph Neural Networks (GNNs): Harness GNNs to generate structured outputs from graph-structured data.
- Theoretical Insights: Receive in-depth theoretical knowledge of these models.
- Practical Exercises: Reinforce your understanding through practical exercises.
- Hands-On Projects: Tackle real-world challenges in computer vision, natural language processing, social network analysis, and recommendation systems by mastering these advanced techniques.
- IE 7300, or any other machine learning course with minimum grades of C+
- Proficiency with Python programming
- Experience with TensorFlow
- Knowledge in Linear Algebra, Probability, and Statistics
The labs in this repository are designed to provide hands-on experience with the concepts covered in the course. Each lab includes detailed instructions, code samples, and exercises to help you apply what you've learned in a practical setting. Topics include:
- Neural Networks (FeedForward NNs, Optimization, Regularization, Dropout, Batch-normalization, CNNs, RNNs, LSTM)
- Generative Models (GANs, VAEs, Diffusion Models)
- Transformers (Transformers, BERT, GPT, Performer)
- Deep Reinforcement Learning (RLHF, Chat-GPT, Alignment)
- Graph Neural Networks
To get started with the labs and exercises in this repository, please follow these steps:
- Clone this repository to your local machine.
- Navigate to the specific lab you are interested in.
- Read the lab instructions and review any accompanying documentation.
- Follow the provided code samples and examples to complete the lab exercises.
- Feel free to explore, modify, and experiment with the code to deepen your understanding.
For more detailed information on each lab and prerequisites, please refer to the lab's README or documentation.
We welcome contributions from students and instructors to improve and expand the materials in this repository. If you have suggestions, bug reports, or would like to add new content, please submit a pull request or open an issue. Make sure to follow the contribution guidelines outlined in the CONTRIBUTING.md file.
The reading materials of this repository were collected from the internet under the Creative Commons License. These materials are intended for educational purposes and to enhance your learning experience.
This repository is licensed under the MIT License. For more details, please refer to the LICENSE file.
We hope you find these resources helpful and enjoy your journey into the world of Generative AI! For any questions or support, please contact the course instructor or teaching assistant.