Worker template should be configurable.

[MartinForReal]

The QoS of worker pod created by operator is Burstable due to the resource config here:

pytorch-operator/pkg/common/config/config.go

Lines 9 to 20 in 4aeb650

	var initContainerTemplate = `
	- name: init-pytorch
	image: {{.InitContainerImage}}
	imagePullPolicy: IfNotPresent
	resources:
	limits:
	cpu: 100m
	memory: 20Mi
	requests:
	cpu: 50m
	memory: 10Mi
	command: ['sh', '-c', 'until nslookup {{.MasterAddr}}; do echo waiting for master; sleep 2; done;']`

This is a vital issue because only the pods with Guaranteed class can be processed by cpumanager.

[johnugeorge]

We could make them guaranteed or take these values as args