domains / families / devops / tao-run-on-slurm

tao-run-on-slurm

separate

Remote SLURM GPU cluster execution over SSH with sbatch/srun, Pyxis/Enroot containers, and Lustre-backed

When to use

• Use SLURM when the user has access to a managed GPU cluster, shared Lustre storage, and scheduler-owned GPU allocation. Do not use SLURM for local files that exist only on the agent machine; data and outputs must be reachable from the clus…

When not to use

• local files that exist only on the agent machine
• data and outputs not reachable from the cluster

Skill page

Source