It is possible to use python packages and resources of jupyterhub.wwu.de, while using local files and a local installation of VSCode.
Prerequisites
Having VSCode and the official Jupyter Extension installed.
1. Get a token
You can request a new API Token at jupyterhub.wwu.de/hub/token.
Copy your API Token and keep it safe. With your API Token, anyone can access your data on the jupyterhub. You can recreate and revoke tokens at any time.
2. Use the token in VSCode to connect to Jupyterhub
Open an existing or create a new Jupyter Notebook in a locally running VSCode. You then need a running server on jupyterhub.wwu.de. To connect to the Jupyterhub follow these steps within your locally running VSCode:
Jupyter Server (Bottom status bar) or Kernel (Top-Right) -> Existing Jupyter Server (Remote Connection) -> Enter your Connection URL
https://jupyterhub.wwu.de/user/your_jupyterhub_username?token=your_token
where your_jupyterhub_username is your jupyterhub username -> usually the same as your WWU user name
your_token is your token from step 1.
e.g. with the user ttester and token abcd01e0491244efa1f6e71a35dcd80ayour URL would be
https://jupyterhub.wwu.de/user/ttester?token=abcd01e0491244efa1f6e71a35dcd80a
With this your local vscode is now connected to the remote jupyterhub. All the code execution happens on the hub, meaning that you can use the resources of the hub, including GPUs, while working with local notebooks. You can change the kernel at the top right corner.