Atlas CE Quick Start Guide

Installation

Estimated time: 3 minutes

Prerequisites

  1. Docker version >18.09 (Docker installation instructions)
  2. Python >=3.6
  3. >5GB of free machine storage
  4. The atlas_ce_installer.py file

Atlas CE currently supports Mac OSX, Linux & Windows 10.

Tip

Windows 10 Home edition users can set up Atlas CE by installing Docker Toolbox. Windows 10 Enterprise edition users should use Docker for Windows.

Common

  1. Create a new, empty directory where you will install Atlas CE.

  2. Copy the atlas_ce_installer.py file into this directory.

  3. Create and activate a Python >=3.6 virtual environment using Conda or venv to minimize dependency issues.

  4. Follow you OS-specific instructions below.

Linux/OSX

  1. Run the install script with python atlas_ce_installer.py.

Windows 10

  1. Install the following python package into the environment created in step 3: pip install pypiwin32==223

  2. Run the install script with python atlas_ce_installer.py

  3. [Windows 10 Home only]:

    • Open Oracle VirtualBox.
    • Right-click on the docker VM and click on Settings.
    • Select 'Network' in the configuration pane, expand the Advanced section and click on Port Forwarding.
    • Add the following port forwarding rules:
Name Protocol Host IP Host Port Guest IP Guest Port
Scheduler TCP 127.0.0.1 5000 5000
GUI TCP 127.0.0.1 5555 5555
Tracker TCP 127.0.0.1 5556 5556
Archive TCP 127.0.0.1 5557 5557
Tboard TCP 127.0.0.1 5959 5959

Python SDK Only

  1. Run the install script with python atlas_ce_installer.py -ia

Tip

Running python atlas_ce_installer.py --help will give you troubleshooting advice if the script isn't working as expected.

Tip

The longest part of the script is pulling the Atlas CE docker images, if the script fails at this point, you can re-run it using python atlas_ce_installer.py -dp to skip over the download and unpacking and go directly to the image pull.


Start-up

After completing the installation section, you can do the following to start Atlas CE:

  1. Validate that you are in the same Python environment that was used to run the installation script.
  2. Run atlas-server start.

Tip

You can also start Atlas Server with GPU support by running atlas-server start -g. This will allow Atlas to use all CUDA-enabled GPUs on your system.

Success

Validate that the GUI is running by going to the GUI. This is your centralized location to track all of your experiments.


Hello Atlas

After completing the start-up section, follow the next few steps to launch your first Atlas CE job:

  1. Navigate to where you'd like to create your Atlas project directory.
  2. Ensure that you are in the environment that was used during installation.
  3. Run foundations init hello-atlas to create an example project.
  4. Navigate into the newly created hello-atlas directory.
  5. Run the sample code provided by running python main.py.
  6. Head to the GUI to see your experiment!

Note

When you run a job for the first time, it will download the appropriate worker image needed.
This will take roughly 1.5 minutes.