What it does
Dive into Deep Learning combines mathematical explanations, code and discussion with implementations in several learning frameworks.
Read an explanation beside its implementation rather than separating the mathematics from the program. The framework options let you choose a code version appropriate to your work. Follow one implementation consistently before comparing how another library expresses the same idea. The listed implementations include PyTorch, NumPy/MXNet, JAX and TensorFlow. Pick one branch and check that branch’s environment instructions; those names do not imply that every example uses the latest release of its framework.