apiVersion: "kubeflow.org/v1"
kind: PyTorchJob
metadata:
  name: mnist
spec:
  pytorchReplicaSpecs:
    Master:
      replicas: 1
      restartPolicy: OnFailure
      template:
        spec:
          containers:
            - name: pytorch
              image: docker.io/kubeflowkatib/pytorch-mnist:v1beta1-45c5727
              imagePullPolicy: Always
              command:
                - "python3"
                - "/opt/pytorch-mnist/mnist.py"
                - "--epochs=10"
    Worker:
      replicas: 2
      restartPolicy: OnFailure
      template:
        metadata:
          labels:
            app: mnist
        spec:
          #nodeSelector:
            #node.kubernetes.io/instance-type: "g4dn.metal"
          affinity:
            podAffinity:
              requiredDuringSchedulingIgnoredDuringExecution:
              - labelSelector:
                  matchExpressions:
                  - key: app
                    operator: In
                    values:
                    - mnist
                topologyKey: "topology.kubernetes.io/zone" 
          containers:
            - name: pytorch
              image: docker.io/kubeflowkatib/pytorch-mnist:v1beta1-45c5727
              imagePullPolicy: Always
              resources:
                requests:
                  cpu: 4
                  #nvidia.com/gpu: 8
                limits:
                  cpu: 4
                  #nvidia.com/gpu: 8
              command:
                - "python3"
                - "/opt/pytorch-mnist/mnist.py"
                - "--epochs=10"