deployment on azure using a ACI

azure_deployment.py

## deploy the model to the ACI (azure container instances)

from azureml.core.webservice import AciWebservice
from azureml.core.model import InferenceConfig
from azureml.core.webservice import Webservice
from azureml.core.model import Model

inference_config = InferenceConfig(runtime= "python",  ## runtime for the image, either python or spark-py
                                   entry_script="score.py",
                                   conda_file="myenv.yml") ##dependencies

aciconfig = AciWebservice.deploy_configuration(cpu_cores=1,
                                               auth_enabled=True, #True generates API keys to secure access
                                               memory_gb=1,
                                               tags={'name': 'founder-classifier', 'framework': 'tensorflow'},
                                               description='classify founders using CNN')

service = Model.deploy(workspace=ws, 
                           name='founder-classifier-webservice', 
                           models=[model],  #registered model defined earlier
                           inference_config=inference_config, 
                           deployment_config=aciconfig)

service.wait_for_deployment(True)
print(service.state)
print(service.scoring_uri)