Deep Learning
A free, open graduate course from UVA's School of Data Science — from linear models and backpropagation to transformers and generative AI, taught through video lectures, the open course book, selected alternative readings, and hands-on PyTorch notebooks.
Taught by Heman Shakeri · School of Data Science, University of Virginia
1
Introduction to Deep Learning
FoundationsNeural Networks Basics
Perceptrons
History of AI
2
Backpropagation
FoundationsBackpropagation
Gradient Descent
Activations
3
Optimization Foundations & Ablation Methodology
FoundationsOptimizers
LR Schedules
Regularization
4
Convolutional Neural Networks (CNNs)
CNN ArchitecturesConvolution
Pooling
Vision Models
5
Advanced CNN Architectures
CNN ArchitecturesResNet
DenseNet
EfficientNet
6
Encoder Decoder Architectures
Encoder • DecoderAutoencoders
PCA
Encoder–Decoder
7
Recurrent Neural Networks
Sequence ModelsRNN Basics
LSTM
GRU
8
Attention Mechanism
AttentionAttention Basics
Q-K-V
Scaled Dot-Product
9
Transformer
TransformersSelf-Attention
Positional Encoding
Transformer
10
Transformer Models in Vision and Text
TransformersViT
BERT/T5/GPT
Scaling
Test-Time Memory & Control
11
Prompting, PEFT, and Quantization (Gemma)
Modern LLM StackPrompting/RAG
PEFT
QLoRA
12
Multimodal Learning & GenAI
Modern LLM StackMultimodal
Diffusion
VAE
GAN
All materials stay freely available — no account needed. Course code and original materials are open under CC BY 4.0 on GitHub.