The maximum number of CPUs that can run concurrently across all active runs in the run group.
The maximum time for each run (in minutes). If a run exceeds the maximum run time, the run fails automatically.
The maximum number of GPUs that can run concurrently across all active runs in the run group.
The maximum number of runs that can be running at the same time.
A name for the group.
To ensure that requests don't run multiple times, specify a unique ID for each request.
Tags for the group.