Flow Matching and Diffusion

This repository contains my personal assignment work for the MIT Computer Science Class 6.S184: Generative AI with Stochastic Differential Equations course on Flow Matching and Diffusion Models.

Course Overview

The course covers fundamental concepts and practical implementations of:

• Ordinary Differential Equations (ODEs)
• Stochastic Differential Equations (SDEs)
• Flow Matching
• Diffusion Models

Setup

This project uses Python 3.11 and is managed with uv.

Prerequisites

• Python 3.11 or higher
• uv package manager
• CUDA-compatible GPU (optional, for GPU acceleration)

Installation

1. Clone the repository:

git clone https://github.com/whats2000/FlowMatchingAndDiffusion.git
cd FlowMatchingAndDiffusion

2. Create a virtual environment and install dependencies:

uv venv --python 3.11
uv sync

The environment includes PyTorch with CUDA 12.4 support for GPU acceleration.

Labs

• Lab 1: Simulating ODEs and SDEs - An intuitive and hands-on walkthrough of ordinary and stochastic differential equations

Acknowledgments

Course materials from MIT 6.S898: Deep Learning taught by erives@mit.edu and phold@mit.edu