YOUR COURSE

Deep learning and neural networks

Work through each chapter, practice the ideas, and finish its challenge.

— XPLoading progress…
← All courses

8 chapters · 48 lessons · 0 completed

D1The neuron and the MLP0/6 lessons complete

Compute a neural-network forward pass and count the parameters of a small multilayer perceptron.

Start chapter
  1. D1.1Weighted sums
  2. D1.2Activations
  3. D1.3Layers and shapes
  4. D1.4A forward pass
  5. D1.5Capacity and parameters
  6. D1.6Classification by a network
D2Backpropagation0/6 lessons complete

Apply the chain rule through a graph and implement the core of a scalar reverse-mode autodiff engine.

Start chapter
  1. D2.1Computation graphs
  2. D2.2Local derivatives
  3. D2.3Reverse accumulation
  4. D2.4Topological order
  5. D2.5Gradient checks
  6. D2.6A small autograd engine
D3Training well0/6 lessons complete

Diagnose an unstable training run using losses, gradient magnitudes, initialization, normalization and regularization.

Start chapter
  1. D3.1Loss and update contracts
  2. D3.2Initialization
  3. D3.3SGD and Adam
  4. D3.4Normalization
  5. D3.5Dropout and schedules
  6. D3.6Gradient health
D4PyTorch0/6 lessons complete

Translate an inspected NumPy model into tensors, modules, automatic differentiation and a clear training loop.

Start chapter
  1. D4.1Tensors and dimensions
  2. D4.2Autograd
  3. D4.3Modules and parameters
  4. D4.4Data and minibatches
  5. D4.5The training loop
  6. D4.6Devices and debugging
D5Convolutional networks0/6 lessons complete

Calculate a convolutional output and explain filters, stride, padding, pooling and residual connections.

Start chapter
  1. D5.1Local filters
  2. D5.2Stride and padding
  3. D5.3Channels
  4. D5.4Feature maps
  5. D5.5Pooling
  6. D5.6Residual connections
D6Sequence models0/6 lessons complete

Trace a recurrent state and explain memory loss, gating, sequence-to-sequence models and attention.

Start chapter
  1. D6.1Recurrent state
  2. D6.2Unrolling and backpropagation
  3. D6.3Vanishing and exploding memory
  4. D6.4LSTM and GRU gates
  5. D6.5Sequence-to-sequence
  6. D6.6Attention as a bridge
D7Generative models0/6 lessons complete

Explain the different training objectives of autoencoders, VAEs, GANs and diffusion models.

Start chapter
  1. D7.1Autoencoders
  2. D7.2Variational autoencoders
  3. D7.3Reparameterization
  4. D7.4GANs
  5. D7.5Diffusion
  6. D7.6Latent spaces and evaluation
D8Representation learning0/6 lessons complete

Compare learned embeddings, transfer learning, fine-tuning and contrastive training objectives.

Start chapter
  1. D8.1Embeddings
  2. D8.2Transfer learning
  3. D8.3Fine-tuning
  4. D8.4Self-supervision
  5. D8.5Contrastive learning
  6. D8.6Adaptation and evaluation