Code

first init of MDS on Sophia cluster#66

Code

Open

Eugene29 wants to merge 1 commit intoargonne-lcf:microsoft-mainargonne-lcf/Megatron-DeepSpeed:microsoft-mainfrom Eugene29:sophia-clusterEugene29/alcf-MDS:sophia-clusterCopy head branch name to clipboard

+62-11

Commits

Commits on Oct 13, 2024

• first init of MDS on Sophia cluster. Has an issue of using much more memory than polaris.

  

  Eugene Ku

  committed
  9f1038b

  Copy full SHA for 9f1038b