Generative & Advanced Deep Learning
A 14-module path through generative & advanced deep learning, in study order from foundations to the frontier.
Modules
- Generative Modeling: Learning Data Distributions
- Latent Variable Models and the Evidence Lower Bound
- Variational Autoencoders
- The Reparameterization Trick
- Generative Adversarial Networks
- GAN Training Dynamics and Failure Modes
- Diffusion Models: Forward and Reverse Processes
- Denoising and Score-Based Models
- Self-Supervised Learning
- Contrastive Learning
- Model Compression and Knowledge Distillation
- Quantization of Neural Networks
- Neural Architecture Search
- Interpretability and Adversarial Examples