Installation

Here is the step by step guide to run simple MLCube cubes.

Create a python environment

# Create Python Virtual Environment
virtualenv -p python3 ./env && source ./env/bin/activate

Install MLCube Runners

# Install MLCube Docker runner 
pip install mlcube-docker

# Optionally, setup host environment by providing the correct `http_proxy` and `https_proxy` environmental variables.
# export http_proxy=...
# export https_proxy=..

# Optionally, install other runners
# pip install mlcube-singularity
# pip install mlcube-ssh
# pip install mlcube-k8s

Explore with examples

git clone https://github.com/mlcommons/mlcube_examples.git && cd ./mlcube_examples

A great way to learn about MLCube is try out the example MLCube cubes located here.