Deep Learning @ TAU
Course home across academic iterations
Deep Learning @ TAU
I teach Deep Learning at Tel Aviv University as a hands-on course on modern neural networks, spanning optimization, CNNs, sequence models, transformers, and generative modeling in PyTorch.
Course Iterations
Recent iterations have treated deep learning as both a conceptual and engineering subject. Students build models in PyTorch, reason about optimization and regularization, and work through image, text, and sequence problems before reaching transformer-based and generative methods.
This page collects shared materials from multiple iterations of the course together with links to the concrete semester pages.
Course Outline
- The course starts with neural networks, backpropagation, optimization, and practical training workflows.
- It continues through deep feedforward models, convolutional networks, and sequence models such as RNNs, LSTMs, and GRUs.
- Later sessions cover attention, transformers, transfer learning, and large-scale model training.
- Recent iterations also introduce generative modeling, including autoencoders, VAEs, GANs, and diffusion-style ideas.
- Assessment typically mixes class assignments, home assignments, and a final pair project.