I was recently pulled into Discord thanks to a friend (TS), and started looking around for some other communities to join and learn from. A couple of google searches and medium articles pointed me to several communities, and several links in came across Yann LeCun’s course from NYU with recordings from 2020 (Deep Learning with Pytorch):
Course website: https://atcold.github.io/pytorch-Deep-Learning/
Links to videos, notebooks and exercises (in Pytorch)
As well as notes from all the videos
Reddit: https://www.reddit.com/r/NYU_DeepLearning/
Even though I’ve only covered 1 lecture so far, I’ve been enjoying myself: this course seems perfectly suited for those of us who’ve messed around with DL in the past but don’t quite have everything down yet. Both the lectures and the practicum are fascinating, and fun to watch.
Another book I picked up — and found really promising by the table contents, but haven’t had time to dig into — is Probabilistic Machine Learning: An Introduction by Kevin P Murphy. It’s available for download here.
Finally, I’ve ordered Prof. Strang’s Linear Algebra and Learning From Data — only available as a hard copy — that’s slowly making it’s way in the mail. I’m really excited to get my hands on this book. In the mean time, 3Blue1Brown’s Essence of Linear Algebra was really good overlapped with mml-book to stop feeling completely lost with Vectors.
PS. I’ve been using deepnote.com to run the notebooks from the course; it works fairly elegantly with a github repository, which made it really convenient to run the tutorial notebooks.