Welcome to MLSpace!

MLSpace is a no-hassle tool for data science, machine learning and deep learning.

MLSpace has pre-made environments for pytorch, tensorflow and everything else you might need. All environments come with VSCode (code-server) and JupyterLab. You no longer need to care about CUDA/cuDNN versions!

Setting up MLSpace is a three step process:

• installation

• set up

• create and run environments

$ mlspace --help

usage: mlspace <command> [<args>]

positional arguments:
{create,setup,start,stop}
                        commands
   create              Create a new MLSpace
   setup               Setup MLSpace and install all dependencies. Run with `sudo`
   start               Start a new space
   stop                Stop a running MLSpace instance

optional arguments:
-h, --help            show this help message and exit
--version, -v         Display MLSpace version

For more information about a command, run: `mlspace <command> --help`

Installation

There are multiple ways to install MLSpace. Easiest is if you have python and pip installed.

If you already have python & pip installed on your system, you can just do:

$ pip install -U mlspace

If you do not have python and pip installed on your system, the first step would be to install them.

$ sudo apt-get update
$ sudo apt-get install -y python3 python3-pip

If you have multiple versions of python installed, you might want to update alternatives and point python command to a particular version.

NOTE: MLSpace will work with any python >= 3.5!

Once python & pip are installed, you can now install mlspace using:

$ pip install -U mlspace

Setup

To start the setup process run:

$ mlspace setup

Sit back, relax and let it install everything you will need :)

Creating an environment

An environment can be created using the mlspace create command.

For example, you can create a torch environment without GPU using:

$ mlspace create --name name_of_your_env --backend torch

and if you want to create an environment with GPU support, just add –gpu to the create command.

$ mlspace create --name name_of_your_env --backend torch --gpu

At any point, you can get help for a command using –help. E.g.

$ mlspace create --help

usage: mlspace <command> [<args>] create [-h] --name NAME --backend {torch} [--gpu]

optional arguments:
-h, --help         show this help message and exit
--name NAME        Name of MLSpace
--backend {torch}  MLSpace backend
--gpu              Whether to use GPU

Running an environment