GTCOARLab

an environment for interactive exploration of reinforcement learning

Build	Interactive Demo

Downloading GTCOARLab

• View the releases on GitHub
• Choose the right artifacts for your operating system and hardware
  • CPU: does not require a CUDA-capable graphics card and drivers
  • GPU: some features require a CUDA-capable graphics card and drivers
• Download the artifacts (they are big, it will take a while)

  • If your platform contains multiple .zip and .z01, .z02 files, you need to download them all
    • a GUI or the zip command line tool will know how to decompress them correctly by using the .zip file, which should be smaller than the others

• Follow the instructions for e.g. miniconda
  • on Windows, prefer installing to a short path on a fast device (e.g. SSD) such as C:/gtcl
  • if you've already installed Anaconda, Miniconda, or Miniforge, it is recommended to not enable shell integration, environment variables, or registering as the default Python
• From the command line, activate the environment

  • on Windows, you should see a Start Menu entry

  • run jupyter lab, and you should see the JupyterLab interface open in your default browser with
    • to choose your browser, start jupyter lab --no-browser and copy/paste the URL shown into your browser of choice
• If you run into problems, create an issue on GitHub

Copyright (c) 2021 University System of Georgia and GTCOARLab Contributors

Distributed under the terms of the BSD-3-Clause License