Short, task-shaped guides to installing ClusterPilot, submitting jobs, and getting the most out of your cluster. Pick one from the list, or search.
Install ClusterPilot, write a config file for your cluster, and submit your first SLURM job from the terminal UI. Roughly ten minutes of setup.
ClusterPilot puts each job under your cluster's scratch path and syncs results home. Why that copy is temporary, and how to set the path correctly.
seff tells you how much CPU and memory a finished job actually used, so your next request is closer to reality. It says nothing at all about GPUs.
Most Alliance clusters let you request part of a GPU instead of a whole one. The syntax per cluster, why slices queue faster, and when not to use them.