Problem while experimenting elasticity

I wish to experiment elasticity and scalability. I have given the following command into my system

kubectl create -f https://raw.githubusercontent.com/robertobruzzese/Project_Cloud/main/simple

Thus I have created 6 worker replicas. The system correctly creates 7 pytorchjob pods that I can visualize while they are running with the command

kubectl get pods -l job-name=pytorch-simple -n kubeflow

The number of replicas matches 7, as it is given in the command (simple). (see figure) Now I want to experiment scalability and with that I means to set a CPU utilization threshold that when it is reached in the pods the system automatically creates more replicas. In order to do this autoscaling I have created and applied the following yaml file

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: pytorch-simple-hpa
  namespace: kubeflow
spec:
  scaleTargetRef:
    apiVersion: kubeflow.org/v1
    kind: PyTorchJob
    name: pytorch-simple
  minReplicas: 1
  maxReplicas: 10
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 50

And after having created this yaml file object I applied it with the following command :

kubectl apply -f pytorch-simple-hpa.yaml  

Unfortunately I did not observe any increase in the number of replicas. My question is , what is wrong with my work ? How I should increase automatically the number of replicas once the CPU utilization threshold has been set ? Thank a lot. Bye