added installation instructions for paula/SC/ULEI

docs/00_setup/install_paula.md

@@ -0,0 +1,53 @@
+# Installation instructions for Scientific Computing Uni Leipzig (paula)
+
+Request a [Scientific Computing Account at ULei](https://www.urz.uni-leipzig.de/servicedesk-und-hilfe/hilfe-zu-unseren-services/forschung/hilfe-webbasiertes-data-science-und-machine-learning-mit-jupyter).
+
+Login to VPN and [Jupyter Hub](https://lab.sc.uni-leipzig.de/), select the configuration paula/A30:
+
+![](config-paula.png)
+
+Open a terminal and install [ana]conda
+```
+module load Anaconda3
+```
+
+```
+conda init bash
+```
+
+At this point, we need to reopen the terminal. Afterwards, create a conda environment with one specific python version only.
+
+```
+conda env create -f https://raw.githubusercontent.com/ScaDS/generative-ai-notebooks/main/docs/00_setup/environment-gpu.yml
+```
+
+```
+conda activate genai-gpu
+```
+
+Make this conda environment available to Jupyter hub like this:
+```
+python -m ipykernel install --user --name 'genai-gpu' --display-name "ScaDS.AI genai-gpu"
+```
+
+To make stackview work, install it outside the conda environment:
+```
+conda deactivate
+```
+
+```
+pip install stackview --user
+```
+... and reload the page (in browser)
+
+
+Just in case you want to install more packages later, reopen the terminal, activate the environment and install them:
+```
+conda activate genai-gpu
+pip install other-package 
+```
+
+## Read more
+
+These instructions were derived/modified from [this page](https://www.sc.uni-leipzig.de/05_Instructions/Jupyter/#how-to-use-your-own-environments-and-kernels).
+

docs/_toc.yml

@@ -8,6 +8,7 @@ parts:
   - caption: Installation
     chapters:
     - file: 00_setup/readme.md
+    - file: 00_setup/install_paula.md
 
   - caption: Prompting basics
     chapters: