Generative Models
How AI learned to create
VAEs, GANs, and diffusion — the model families behind AI-generated images, audio, and text.
Intermediate3 modules · 10 lessons~88 min875 XPAligned with Hugging Face Diffusion CourseStart with: Deep Learning
Progress0/10
Module
Generation as a Goal
The shift from recognizing the world to imagining it
Generative vs Discriminative9 min · 85 XPGenerative vs Discriminative
Latent Space8 min · 80 XPLatent Space
Sampling as Creativity8 min · 80 XPSampling as Creativity
Module
The Architectures
Three families of generators — compress, compete, predict
Autoencoders & VAEs9 min · 85 XPAutoencoders & VAEs
GANs: The Adversarial Game9 min · 90 XPGANs: The Adversarial Game
Autoregressive Models9 min · 90 XPAutoregressive Models
Module
Diffusion
Destroy the data with noise, then learn the way back
Noise In, Noise Out9 min · 90 XPNoise In, Noise Out
DDPM & DDIM8 min · 85 XPDDPM & DDIM
Latent Diffusion9 min · 90 XPLatent Diffusion
Guidance & Control10 min · 100 XPGuidance & Control
Capstone project
Generator study
Generate images, text, or audio with two different models or prompt strategies. Curate the outputs, then analyze: what does each approach reliably get right and wrong?
Deliverable: A small gallery plus a comparison write-up of strengths, failure modes, and use cases.
Milestones
Finish every lesson to earn the Generative Models Certificate.