Monthly Archives: March 2018
Deprecated warning: PyTorch 1.0+ uses to Aten as its tensor library. So the article is no longer applicable in PyTorch 1.0+. Python is one of the most popular programming languages today for deep learning applications. However, as an interpreted language, it has been considered too slow for high-performance computing. Therefore, I incorporated CUDA into Python to build my own extension that accelerates the computation. In this tutorial, I give an example of CUDA extension for data augmentation in deep learning. Create Build Scripts First of all, we need to create a script for building FFI extension named build.py. Create CUDA Interface Then, we need to create an interface for CUDA which can be invoked by Python. In this tutorial, we…