Subslicing workload annotations

[scaliby]

Fixes / Features

Adds subslicing workload annotations necessary for the feature to work.

Testing / Documentation

Added a temporary print and this is the generated yaml:

template:
  metadata:
    labels:
      xpk.google.com/workload: workload
    annotations:
      
      kueue.x-k8s.io/podset-required-topology: "google.com/gke-tpu-slice-10x10-id"
      cloud.google.com/gke-tpu-slice-topology: 10x10
  spec:
    schedulerName: default-scheduler
    imagePullSecrets:
    - name: None
    restartPolicy: Never
    
    nodeSelector:
      cloud.google.com/gke-tpu-accelerator: tpu-v6e-slice
      cloud.google.com/gke-tpu-topology: 2x4
      
    priorityClassName: medium
    hostNetwork: true
    dnsPolicy: ClusterFirstWithHostNet

Experiment disabled:

          template:
            metadata:
              labels:
                xpk.google.com/workload: workload
              annotations:
                
                
            spec:
              schedulerName: default-scheduler
              imagePullSecrets:
              - name: None
              restartPolicy: Never
              
              nodeSelector:
                cloud.google.com/gke-tpu-accelerator: tpu-v6e-slice
                cloud.google.com/gke-tpu-topology: 2x4
                
              priorityClassName: medium
              hostNetwork: true
              dnsPolicy: ClusterFirstWithHostNet
              terminationGracePeriodSeconds: 30

the golden hash has changed because I added a new line in annotations section to yaml.

• Tests pass
• Appropriate changes to documentation are included in the PR