PyTorch is an open-source machine learning library that provides tensor computation with automatic differentiation and support for GPU acceleration. Developers use it to build and train neural networks in Python, with an interface similar to NumPy but with the ability to run efficiently on GPUs and define dynamic computational graphs. It's widely adopted for research and production deep learning applications because it offers intuitive APIs for both the high-level training workflows and low-level tensor operations.

18,299 open issues · last updated Jul 21, 2026