PyTorch

PyTorch is composed of the following, Tensor library like NumPy with strong Graphics processing unit (GPU) support , a tape based automatic differentiation library that supports all differentiable Tensor operations in Torch, a neural networks library deeply integrated with Autograd which is designed for maximum flexibility , an optimization package to be used with neural networks with standard optimization methods such as SGD, RMSProp, LBFGS, Adam and other methods, Python multiprocessing with memory sharing of Torch Tensors across processes for data loading and Hogwild training, DataLoader and other utility functions and a legacy code ported from Torch for backward compatibility.