DISCONTINUATION OF PROJECT. This project will no longer be maintained by Intel. Intel will not provide or guarantee development of or support for this project, including but not limited to, maintenance, bug fixes, new releases or updates. Patches to this project are no longer accepted by Intel. If you have an ongoing need to use this project, are interested in independently developing it, or would like to maintain patches for the community, please create your own fork of the project.

nGraph Python

An Intermediate Representation, Compiler, and Executor for Deep Learning

Updated: March 17, 2018

Welcome to the nGraph Python repo. Now that we've released our C++ rewrite to the community, we are transitioning away from this original Python implementation. You can browse here to learn a bit about the roots of the legacy project.

Why did we build nGraph?

When Deep Learning (DL) frameworks first emerged as the vehicle for training and inference models, they were designed around kernels optimized for a particular platform. As a result, many backend details were being exposed in the model definitions, making the adaptability and portability of DL models to other or more advanced backends inherently complex and expensive.

The traditional approach means that an algorithm developer cannot easily adapt his or her model to different backends. Making a model run on a different framework is also problematic because the developer must separate the essence of the model from the performance adjustments made for the backend, translate to similar ops in the new framework, and finally make the necessary changes for the preferred backend configuration on the new framework.

We designed the Intel nGraph project to substantially reduce these kinds of engineering complexities. While optimized kernels for deep-learning primitives are provided through the project and via libraries like IntelĀ® Math Kernel Library for Deep Neural Networks (IntelĀ® MKL-DNN), there are several compiler-inspired ways in which performance can be further optimized.

How does it work in practice?

Install the nGraph library and write or compile a framework with the library, in order to run training and inference models. Using the command line on any supported system (currently Linux) specify the backend you want to use. Our Intermediate Representation (IR) layer handles all the hardware abstraction details and frees developers to focus on their data science, algorithms and models, rather than on machine code.

At a more granular level of detail:

You can read more about design decisions and what is tentatively in the pipeline for backends and development in our ARXIV abstract and conference paper.