Skip to content

Instantly share code, notes, and snippets.

@aaronmarkham
Last active October 16, 2018 16:57
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save aaronmarkham/132c5bf044bc179b9eb25e866b6ba102 to your computer and use it in GitHub Desktop.
Save aaronmarkham/132c5bf044bc179b9eb25e866b6ba102 to your computer and use it in GitHub Desktop.

Installation Instructions for MXNet-$binding on $OS

The following guide will help you install MXNet-$binding on $OS.

Prerequisites

The following are required prerequisites:

  1. Install $prerequisite:
brew/apt/yum/conda/pip install $package

{repeat}

The following are recommended prerequisites:

  1. Install a math library. MKL is recommended.

Option A: Install MKL

brew/apt/yum install mkl

Option X: Install Y

brew/apt/yum install Y
  1. Install some other recommended prerequisite.
brew/apt/yum install $other
  1. Follow-up steps (optional)

Install MXNet

  1. Install $mxnet-package:
pip install $mxnet-package
  1. Follow-up steps (optional)

Install MXNet-$binding

  1. Install mxnet-$binding:
# custom instructions
  1. Follow-up steps (optional)

Next Steps

  • Tutorials

    1. x
    2. y
    3. z
  • Examples

    1. x
    2. y
    3. z

Default Installation Script

The following script contains the default recommended installation options.

# Example to install MXNet-R on macOS:

# Install prerequisites
brew install opencv

# Install recommended math library
brew install openblas@0.3.1
ln -sf /usr/local/opt/openblas/lib/libopenblasp-r0.3.* /usr/local/opt/openblas/lib/libopenblasp-r0.3.1.dylib

# Install MXNet-R
echo "cran <- getOption("repos")" | R
echo "cran["dmlc"] <- "https://apache-mxnet.s3-accelerate.dualstack.amazonaws.com/R/CRAN/" | R
echo "options(repos = cran)" | R
echo "install.packages("mxnet")" | R

Notes

When updating this installation guide, be sure to update the default installation script. It will be tested through CI.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment