Generative Models
VAEs, GANs, and Diffusion Models — learning to generate data from scratch.
7 lessons
01What Are Generative Models?
The two paradigms of machine learning — discriminative vs generative.
Autoencoders
Learning compressed representations — and why they're not generative yet.
Variational Autoencoders
Combining deep learning with variational inference to generate new data.
Generative Adversarial Networks
Two networks competing — a generator that creates and a discriminator that judges.
Diffusion Models
State-of-the-art generation by learning to reverse a gradual noising process.
ViT & Modern Generative AI
Vision Transformers, Classifier-Free Guidance, StyleGAN, CycleGAN, and the Stable Diffusion pipeline — how generative models scaled to photorealistic synthesis.
Generative Models Quiz
Test your understanding of the generative-model families — VAEs, GANs, diffusion, and the modern latent-diffusion pipeline.