Collective Knowledge repository for MXNet

All CK components can be found at cKnowledge.io and in one GitHub repository!

Introduction

This repository provides high-level, portable and customizable Collective Knowledge workflows for MXNet. It is a part of our long-term community initiative to unify and automate AI, ML and systems R&D using Collective Knowledge Framework (CK), and to collaboratively co-design efficient SW/HW stack for AI/ML during open ACM ReQuEST competitions as described in the ACM ReQuEST report. All benchmarking and optimization results are available in the public CK repository. See CK getting started guide for more details about CK.

Minimal CK installation

The minimal installation requires:

• Python 2.7 or 3.3+ (limitation is mainly due to unitests)
• Git command line client.

Linux/MacOS

You can install latest CK via PIP (with sudo on Linux) as follows:

$ sudo pip install ck

You can also install CK in your local user space without sudo as follows:

$ git clone http://github.com/ctuning/ck
$ export PATH=$PWD/ck/bin:$PATH
$ export PYTHONPATH=$PWD/ck:$PYTHONPATH

Windows

First you need to download and install a few dependencies from the following sites:

• Git: https://git-for-windows.github.io
• Minimal Python: https://www.python.org/downloads/windows

You can then install CK as follows:

 $ pip install ck

or

 $ git clone https://github.com/ctuning/ck.git ck-master
 $ set PATH={CURRENT PATH}\ck-master\bin;%PATH%
 $ set PYTHONPATH={CURRENT PATH}\ck-master;%PYTHONPATH%

CK workflow installation for MXNet

CPU

$ ck pull repo:ck-mxnet
$ ck install package --tags=lib,mxnet,vcpu

GPU

$ ck pull repo:ck-mxnet
$ ck install package --tags=lib,mxnet,vcuda

Checking classification example (and automatically installing available MXNet model(s) via CK)

$ ck run program:mxnet

• Select 'classify-cpu' or 'classify-gpu' command line
• Select image to classify
• Observe result

Using CK virtual environment

CK support lightweight virtual environment for all packages (automatically setting all necessary environment variables for different versions of different tools natively installed on a user machine).

You can start a virtual environment for a given MXNet package as follows:

$ ck virtual env --tags=lib,mxnet
> export | grep "CK_"

Building from sources on ARM-based system (FireFly, RPi)

$ ck install package:lib-mxnet-master-cpu --env.USE_F16C=0

Trying CK MXNet via Docker

See available Docker images with different python version:

$ ck ls docker:ck-mxnet*

Build the one you need, for example ck-mxnet-py35:

$ ck build docker:ck-mxnet-py35 --sudo

You can now run this Docker image and check classification:

$ ck run docker:ck-mxnet-py35 --sudo
$ ck run program:mxnet

Skip --sudo if you have a local Docker installation.

Related Publications

@article{DBLP:journals/corr/ChenLLLWWXXZZ15,
  author    = {Tianqi Chen and Mu Li and Yutian Li and Min Lin and Naiyan Wang and Minjie Wang and Tianjun Xiao and Bing Xu and Chiyuan Zhang and Zheng Zhang},
  title     = {MXNet: {A} Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems},
  journal   = {CoRR},
  volume    = {abs/1512.01274},
  year      = {2015},
  url       = {http://arxiv.org/abs/1512.01274},
  archivePrefix = {arXiv},
  eprint    = {1512.01274},
  timestamp = {Wed, 07 Jun 2017 14:40:48 +0200},
  biburl    = {http://dblp.org/rec/bib/journals/corr/ChenLLLWWXXZZ15},
  bibsource = {dblp computer science bibliography, http://dblp.org}
}

@inproceedings{ck-date16,
    title = {{Collective Knowledge}: towards {R\&D} sustainability},
    author = {Fursin, Grigori and Lokhmotov, Anton and Plowman, Ed},
    booktitle = {Proceedings of the Conference on Design, Automation and Test in Europe (DATE'16)},
    year = {2016},
    month = {March},
    url = {https://www.researchgate.net/publication/304010295_Collective_Knowledge_Towards_RD_Sustainability}
}

Feedback

• CK community.
• MXNet community