For anyone not wanting to press the link/want a summary, this is a course "targeting the advanced undergraduate and PhD level students". It says that "students

For anyone not wanting to press the link/want a summary, this is a course "targeting the advanced undergraduate and PhD level students". It says that "students will build a basic deep learning library, comparable to a very minimal version of PyTorch or TensorFlow, scalable to a reasonably-sized system (e.g., with fast GPU implementations of operations)." It requires knowledge both of C++ and Python. The professors behind this course is Tianqi Chen and Zico Kolter, both assistant professors at Carnegie Mellon. Sounds like an interesting course, like that they aim for the students to build their own Deep Learning library. It feels like a logical step for anyone wanting to get a deep understanding of this topic.