Closed tejaschumbalkar closed 4 years ago
Didnt understand the intent of this PR. Watcher is only supported for Single node and SM jobs maybe inference (need to check)
in short this commit is still valid
Didnt understand the intent of this PR. Watcher is only supported for Single node and SM jobs maybe inference (need to check)
in short this commit is still valid
Watcher for MPIJob is currently not implemented #959 and hence we can keep #974 until then.
Issue #, if available:
974
Description of changes:
Reverting #974 to support horovod job.
By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.