Viewing a single comment thread. View all comments

Kinwwizl t1_j4slam2 wrote

That's one of the reasons GCP is nice for ML training workloads - you can kill VM after training is finished calling poweroff at the end of bash script for training.

1