1. Use VSCode to Login or

   ssh [email protected]
   

2. load the right python

   module switch python3/3.10.12
   
   

   Optional**:**

   Create virtualenv in directory

   		python3 -m venv .venv
   

     Install packages or load from requirements
   

   source .venv/bin/activate
   python3 -m pip install <your_packages>
   or 
   python3 -m pip install -r requirements.txt
   

3. Switch to a HPC interactive node with GPUs (in increasing order of power, but decreasing order of memory capacity)

 voltash | sxm2sh | a100sh

1. Check the GPUs availability

nvidia-smi

1. Load the venv

source .venv/bin/source

1. Execute the script by choosing the least utilized GPU or press play to use default GPU

CUDA_VISIBLE_DEVICES=1 python3 src/XRaySegmentation.py

Further Information: https://github.com/YoungChulDK/DTU-HPC-Setup

Jupyter Notebook on HPC [GUIDE]:

Access remote code in a breeze with JupyterLab via SSH

Create requirement.txt file

1. python3 -m pip install pipreqs (once)
2. pipreqs (in directory with python files)