You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Here are my instructions for the cluster I built: https://github.com/nateGeorge/slurm_gpu_ubuntu. Some things still seem a little off (doesn't seem robust upon restarts for some reason, some weird debug-level-3/4 messages about cgroups in the logs), but it works. I'm able to schedule GPU jobs which run on all the GPUs in the cluster which was the goal.
alexmyczko
changed the title
update for Ubuntu 18.04?
update for Ubuntu 20.04?
May 8, 2020
alexmyczko
changed the title
update for Ubuntu 20.04?
update for Ubuntu 22.04?
Aug 19, 2022
will you update the docs for the new 18.04 LTS?
also the ubuntu 18.04 slurm version is not the latest and a simple backport from sid builds fine.
cuda-10 and nvidia-410 with self built tensorflow works great too (needs bazel)
The text was updated successfully, but these errors were encountered: