Bertha
Documentation for the shared workstation of the Emmy Noether group at BIPS
Bertha is the shared ML workstation of the Emmy Noether group at BIPS (Leibniz Institute for Prevention Research and Epidemiology). It provides GPU-accelerated computing for statistical analysis, machine learning, and deep learning workflows using R and Python.
TipNew to Bertha?
- Set up SSH access following the Remote Access guide
- Check the Resource Dashboard at http://bertha.bips.de:443 to see current usage
- Then set up Python or R depending on your workflow
Quick links
| I want to… | Go to… |
|---|---|
| Connect from my laptop | Remote Access |
| Check who’s using what | Resource Dashboard |
| Set up Python | Python with uv |
| Use R | R & RStudio |
| Run PyTorch / TensorFlow on the GPU | GPU & Deep Learning |
| Use containers | Apptainer |
For administrators
| Topic | Page |
|---|---|
| Initial system setup | Initial Setup |
| System packages | System Packages |
| Network config | Network |
| Backup system | Backups |
| Other tools | Miscellaneous |
Scripts
Admin scripts live in the repository and are organized by frequency of use:
bin/— regularly used tools (backup, restore, R modules, RStudio management)setup-scripts/— one-time installation scriptsnvidia/— GPU management and ML setup scripts