NYU-DLSP20

Why we included this project

This is the actual material from NYU's graduate deep learning course, taught in spring 2020 by Yann LeCun and Alfredo Canziani, preserved as a full run of Jupyter notebooks, slides, and the recorded lectures themselves. Because it follows a real semester's schedule, it has a coherent arc that scattered tutorial collections rarely manage: each notebook pairs a lecture with a concrete PyTorch exercise, covering supervised and unsupervised learning, convolutional and recurrent networks, and embedding and metric learning, with applications in vision, language, and speech. You can run the notebooks in Binder or a local conda environment, so the material works as a self-paced curriculum or as ready-made teaching material for someone running their own course. A companion website mirrors the content in video and text, and the README has been translated into more than a dozen languages, which makes the material considerably easier to approach for non-English-speaking readers.