AzureSpotDSVM.sh

#!/bin/bash
# Script to create Azure DSVM Spot instance with NVidia P100 GPU  
read -p "Azure VM Name (default: dsvm): " vminput
vmname=${vminput:=dsvm}

while [ $password != $password2 ] ; do 
	read -s -p "Choose your Password: " password
	echo
	read -s -p "Re-enter Password: " password2
	echo
done

echo "The following locations are available: "
az account list-locations --query "[].{DisplayName:displayName, Name:name}" -o table
read -p "Azure VM Zone (default:Australia East): " vmzoneinput
vmzone=${vmzoneinput:=australiaeast}

az group create --name $vmname -l $vmzone

echo "Creating Azure Data Science VM $vmname..."
# You can change the size parameter if you want something other than NVidia K80 GPU instance. 
# NC series VMs provide GPU (including Tesla K80, V100) for deep learning. 
# Here is the link for all available GPU: https://azure.microsoft.com/en-au/pricing/details/virtual-machines/linux/
# You can find Azure size labels by running Azure CLI command in Cloud Shell "az vm list-sizes -l westus2 -o table"  
az vm create --name $vmname -g $vmname --image microsoft-dsvm:ubuntu-1804:1804:latest  --priority Spot --size Standard_NC6s_v3 --eviction-policy Deallocate --storage-sku StandardSSD_LRS --os-disk-size-gb 150 --public-ip-address 13.75.213.23 --admin-user xinzhel --admin-password $password 
az vm open-port --name $vmname -g $vmname --port 8000

#echo "Installing  fastai v2 and notebooks..."
#az vm extension set --resource-group $vmname --vm-name $vmname --name customScript --publisher Microsoft.Azure.Extensions --protected-settings '{"fileUris": ["https://raw.githubusercontent.com/Azure/DataScienceVM/master/Samples/fastai2/installfastai2.sh"],"commandToExecute": "./installfastai2.sh"}'
IP=$(az vm show -d -g ${vmname} --name ${vmname} --query publicIps -o tsv)

echo "You can now login to VM with SSH or use Jupyterhub by visiting https://${IP}:8000/ (Ignore self signed cert warnings)"

echo "Login userid is 'xinzhel' with password you entered above"