MLOps Engineering at Scale

As a female data scientist, I had trouble working with the rest of the team when having to deploy my deep learning and other machine learning models in the cloud. I felt that I didn't have enough of the understanding of the PyTorch workings to train the models and get them ready for production. Also, most of my data science coursework was on a laptop so I didn't really understand or get what to do with the cloud. Fortunately, this book walked me through an entire project using a single dataset, from data clean up to model training. Along the way, I learned how to setup the project in the cloud, how to use PyTorch properly, and how to take advantage of distributed training in Amazon Web Services. The book helped me fill in the blanks in the MLOps that I didn't realize I missed from studying machine learning and data science.

I spent about 2-3 months reading through this book. There is just so much good stuff. I used to think that I understood automatic differentiation but I didn't. I actually did the project from the book to implement a small scale autodiff and then I realized how important it is for distributed (scalable) deep learning. Before I used to be a TensorFlow guy, now I'm PyTorch all the way. Do yourself a favor and study this book! This will help you in interviews and your deep learning projects. Highly recommended!