Before starting with PyTorch, I spent several years working with TensorFlow versions 1 and 2. Since I switched to PyTorch, I became more productive as a machine learning practitioner, and I found my experience learning and using PyTorch delightful. I wanted to share this experience with the readers of this book. In the process, I aim to help you pick up the core elements of PyTorch, guide you though the levels of abstraction available in the framework, and prepare you to transition from using PyTorch in isolation to using machine learning models implemented in PyTorch and integrated into a broader machine learning pipeline.