PyTorch is an open source DL library having tensors as its core data structure providing an extensive set of operations on them via the torch module.
With most of it written in C++ and CUDA, PyTorch provides high performance scientific computing with massive parallelism support.
This repository demonstrates how PyTorch could be easily used to develop large deep learning systems to solve important problems. Project for educational purpose only.
Since I like technical writing, I also write blogs about PyTorch as I learn it. Find links to them -
Dataset, DataLoader And the New Datapipe - PyTorch keeps making life easier for Deep Learners!
Coming Soon: PyTorch Autograd and backward hooks - Let's see under the hood!