Main page

This space contains guides and other resources related to the ML workstation at the Copenhagen campus.

The preferred way of getting managed GPU access at AAU is through CLAAUDIA. For selected projects, you may be granted access to the local workstation with your {AAU_EMAIL_USERNAME}, which is the part of your AAU MAIL before the @ sign (mine is cer).

Then on the campus you can access to the jupyter server by opening a browser with the following address: http://172.30.207.25

The first time you'll connect, your user name is your {AAU_EMAIL_USERNAME} and you'll have to give a password of your choice (and remember it for future logins).

A successful login should bring you to a conda and CUDA enabled jupyter lab instance with 2 GPUs.

You can install your preferred custom packages to your homespace with the --user flag:

pip install --user scikit-learn

If you want to access to the server outside of the campus, you will have to setup a VPN (preferred) or ssh with AAU two factor authorization (not preferred, but included for historical reasons).

At this stage you should have a nice jupyterlab server that can run jupyter notebooks with many extension, a terminal. The following sections will guide you through using the GPUs, exchange data, and many other tasks. Moreover, a list of useful links is provided, including CLAAUDIA GPU access.

For any question about the platform, please contact Cumhur Erkut. For questions, suggestions, and corrections related to the content of this documentation, refer to this page.

Last updated