Fix Azure example-gpu.yml (#9242)

`imageSku` should be string ... otherwise this is the error it shows up:
```
msrestazure.azure_exceptions.CloudError: Azure Error: InvalidTemplate
Message: Deployment template validation failed: 'Template parameter JToken type is not valid. Expected 'String, Uri'. Actual 'Integer'. Please see https://aka.ms/resource-manager-parameter-files for usage details.'.
Additional Information:
	Type: TemplateViolation
	Info: {
	    "lineNumber": 1,
	    "linePosition": 673,
	    "path": "properties.template.parameters.imageSku"
	}

```

python/ray/autoscaler/azure/example-gpu.yaml

@@ -73,7 +73,7 @@ head_node:
         # List images https://docs.microsoft.com/en-us/azure/virtual-machines/linux/cli-ps-findimage
         imagePublisher: microsoft-dsvm
         imageOffer: ubuntu-1804
-        imageSku: 1804
+        imageSku: "1804"
         imageVersion: 20.02.01
 
 # Provider-specific config for worker nodes, e.g. instance type. By default
@@ -84,7 +84,7 @@ worker_nodes:
         # List images https://docs.microsoft.com/en-us/azure/virtual-machines/linux/cli-ps-findimage
         imagePublisher: microsoft-dsvm
         imageOffer: ubuntu-1804
-        imageSku: 1804
+        imageSku: "1804"
         imageVersion: 20.02.01
 
 # Files or directories to copy to the head and worker nodes. The format is a
@@ -129,4 +129,4 @@ head_start_ray_commands:
 # Command to start ray on worker nodes. You don't need to change this.
 worker_start_ray_commands:
     - ray stop
-    - ulimit -n 65536; ray start --address=$RAY_HEAD_IP:6379 --object-manager-port=8076
+    - ulimit -n 65536; ray start --address=$RAY_HEAD_IP:6379 --object-manager-port=8076