NYU-DLSP21

Why we included this project

This repo collects the slides, notebooks, and assignments from NYU's Deep Learning course as taught in spring 2021 by Yann LeCun and Alfredo Canziani. The unusual part is the curriculum: it starts with history, backpropagation, and gradient descent, moves through recurrent and convolutional networks as examples of parameter sharing, then treats latent-variable energy-based models as a core building block rather than an add-on. Because later material builds on that EBM framework, working through the lectures in order gives a different conceptual route into deep learning than most open courseware. It's a good fit for self-study or for instructors who want to foreground energy-based models, and the repo also links earlier editions so you can trace how the material evolved.