def deploy_webservice_from_image(amls_config, workspace, image):
"""
Deploy an AMLS docker image in AMLS' ACI
:param amls_config:
:param workspace:
:param image:
:return:
"""
aciconfig = AciWebservice.deploy_configuration(
cpu_cores=1,
memory_gb=1,
tags=amls_config['tags'],
description=amls_config['description'])
try:
Webservice(workspace=workspace, name=amls_config['name']) \
.delete()
logger.info(f"Deleted existing webservice {amls_config['name']}")
except WebserviceException:
# No need to delete
pass
logger.info(f"Creating webservice {amls_config['name']}")
service = Webservice.deploy_from_image(deployment_config=aciconfig,
image=image,
name=amls_config['name'],
workspace=workspace)
service.wait_for_deployment(show_output=True)
return service
def update_service(aml_interface):
inference_config = get_inference_config(aml_interface)
service = Webservice(name=DEPLOYMENT_SERVICE_NAME,
workspace=aml_interface.workspace)
model = aml_interface.workspace.models.get(MODEL_NAME)
service.update(models=[model], inference_config=inference_config)
print(service.state)
print(service.scoring_uri)
def deploy_service(ws, model, inference_config, service_name, compute_target):
tags = {'model': '{}:{}'.format(model.name, model.version)}
try:
service = Webservice(ws, service_name)
print("Service {} exists, update it".format(service_name))
service.update(models=[model],
inference_config=inference_config,
tags=tags)
except Exception:
print('deploy a new service {}'.format(service_name))
deployment_config = AksWebservice.deploy_configuration(
cpu_cores=1,
memory_gb=2,
tags=tags,
collect_model_data=True,
enable_app_insights=True)
service = Model.deploy(ws, service_name, [model], inference_config,
deployment_config, compute_target)
service.wait_for_deployment(show_output=True)
if service.auth_enabled:
token = service.get_keys()[0]
elif service.token_auth_enabled:
token = service.get_token()[0]
return service.scoring_uri, token
def createWebservice(workspace, container_image, service_name, replica_count,
cores_count, compute_target):
'''
TODO: Should allow for the overwrite flag.
Attach a azureml.core.webservice.Webservice for a given container on an AKS cluster.
If a WebService already exists (by name) on the given workspace, return it instead.
PARAMS:
workspace : azureml.core.Workspace : Existing AMLS Workspace
container_image : azureml.core.image.ContainerImage : Name of an existing AKS cluster
service_name : String : Name of the webservice (deployment) in the AMLS workpsace.
replica_count : int : Number of requested instances of container on cluster.
cores_count : int : Number of cores to allocate to each container
compute_target : azureml.core.compute.AksCompute : AKS cluster to create the service on
RETURNS:
azureml.core.webservice.Webservice
'''
web_service = None
services = Webservice.list(workspace=workspace,
image_name=container_image.name)
if len(services) > 0:
for svc in services:
if svc.name == service_name:
print("Returning existing deployed web service ....",
service_name)
web_service = svc
break
if web_service == None:
print("Creating new web service.....", service_name)
aks_config = AksWebservice.deploy_configuration(
num_replicas=replica_count, cpu_cores=cores_count)
web_service = Webservice.deploy_from_image(
workspace=workspace,
name=service_name,
image=container_image,
deployment_config=aks_config,
deployment_target=compute_target,
)
web_service.wait_for_deployment(show_output=True)
return web_service
def main():
# get workspace
ws = load_workspace()
model = Model.register(ws,
model_name='pytorch_mnist',
model_path='model.pth')
# create dep file
myenv = CondaDependencies()
myenv.add_pip_package('numpy')
myenv.add_pip_package('torch')
with open('pytorchmnist.yml', 'w') as f:
print('Writing out {}'.format('pytorchmnist.yml'))
f.write(myenv.serialize_to_string())
print('Done!')
# create image
image_config = ContainerImage.image_configuration(
execution_script="score.py",
runtime="python",
conda_file="pytorchmnist.yml",
dependencies=['./models.py'])
image = Image.create(ws, 'pytorchmnist', [model], image_config)
image.wait_for_creation(show_output=True)
# create service
aciconfig = AciWebservice.deploy_configuration(
cpu_cores=1, memory_gb=1, description='simple MNIST digit detection')
service = Webservice.deploy_from_image(workspace=ws,
image=image,
name='pytorchmnist-svc',
deployment_config=aciconfig)
service.wait_for_deployment(show_output=True)
def __deploy_service(self, image, compute):
service_config = self.__config['deploy']
services = AksWebservice.list(self.__ws)
service = services.find_by_property('name', service_config['name'])
if service:
service.update(auth_enabled=service_config['auth'])
service.wait_for_deployment(show_output=True)
return service
aks_config = AksWebservice.deploy_configuration(
auth_enabled=True,
max_request_wait_time=75000,
replica_max_concurrent_requests=100,
autoscale_enabled=False,
num_replicas=15)
aks_service_name = service_config['name']
aks_service = Webservice.deploy_from_image(
workspace=self.__ws,
name=aks_service_name,
image=image,
deployment_config=aks_config,
deployment_target=compute)
aks_service.wait_for_deployment(show_output=True)
print(aks_service.state)
return aks_service
def teardown_service(subscription_id, resource_group, workspace_name, workspace_region):
yield
# connect to workspace
ws = azureml_utils.get_or_create_workspace(
subscription_id=subscription_id,
resource_group=resource_group,
workspace_name=workspace_name,
workspace_region=workspace_region,
)
# connect to aci_service
aci_service = Webservice(workspace=ws, name="aci-test-service")
# delete aci_service
aci_service.delete()
def deleteWebService(ws, args):
services = Webservice.list(workspace=ws, model_name=args.modelName)
if (len(services) == 0):
print("Webservice is not deployed.")
else:
print("Webservice is deployed")
services[0].delete()
print("Deleted webservice")
def getWebservice(workspace, webserviceName):
webservice = None
services = Webservice.list(workspace)
for svc in services:
if svc.name == webserviceName:
webservice = svc
break
return webservice
def get_result(players):
import os
import azureml
from azureml.core import Workspace
from azureml.core.webservice import Webservice
from azureml.core.authentication import ServicePrincipalAuthentication
print("getting results")
filename = os.path.join(app.static_folder, 'champion.json')
with open(filename) as json_file:
champions = json.load(json_file)['data']
X = [create_feature_row(players, champions)]
# Check core SDK version number
print("SDK version:", azureml.core.VERSION)
workspace = "league-ws-deploy"
subscription_id = "79451499-b2c0-4513-8dea-ef7f37173fbb"
resource_grp = "league"
svc_pr = ServicePrincipalAuthentication(
tenant_id="1f0113ce-bee6-43b0-9e26-61617eced2e4",
service_principal_id="4c9cfeac-dda9-4298-af3c-d51003c7438b",
service_principal_password="******")
ws = Workspace(workspace_name=workspace,
subscription_id=subscription_id,
resource_group=resource_grp,
auth=svc_pr)
ws.get_details()
print('Workspace name: ' + ws.name,
'Azure region: ' + ws.location,
'Subscription id: ' + ws.subscription_id,
'Resource group: ' + ws.resource_group,
sep='\n')
print("Send to server to predict")
sample = json.dumps({"data": X})
sample = bytes(sample, encoding='utf8')
service = Webservice(workspace=ws, name='lrmpredictfinal6')
# predict using the deployed model
result = service.run(input_data=sample)
return result[0][1]
def deploy(workspace,
name,
model,
script,
source_directory,
environment=None,
target='local',
cpu_cores=1,
memory_gb=1,
compute_target_name=None):
inference_config = InferenceConfig(entry_script=script,
source_directory=source_directory,
environment=environment)
if target == 'local':
deployment_config = LocalWebservice.deploy_configuration(port=8890)
elif target == 'aci':
deployment_config = AciWebservice.deploy_configuration(
cpu_cores=cpu_cores, memory_gb=memory_gb)
elif target == 'aks':
if compute_target_name is None:
print("compute_target_name required when target='aks'")
return None
deployment_config = AksWebservice.deploy_configuration(
cpu_cores=cpu_cores,
memory_gb=memory_gb,
compute_target_name=compute_target_name,
auth_enabled=False)
try:
service = Webservice(workspace, name)
except WebserviceException:
service = None
if service is None:
service = Model.deploy(workspace, name, [model], inference_config,
deployment_config)
else:
print(
"Existing service with that name found, updating InferenceConfig\n"
"If you meant to redeploy or change the deployment option, first "
"delete the existing service.")
service.update(models=[model], inference_config=inference_config)
return service
def deploy(self, model_id, locally):
if locally:
self.ctx.log('Local deployment step is not required for Azure..')
return {'model_id': model_id}
ws = AzureProject(self.ctx)._get_ws()
experiment_name = self.ctx.config.get('experiment/name', None)
if experiment_name is None:
raise AzureException('Please specify Experiment name...')
iteration, run_id = self._get_iteration(model_id)
experiment = Experiment(ws, experiment_name)
experiment_run = AutoMLRun(experiment=experiment, run_id=run_id)
model_run = AutoMLRun(experiment=experiment, run_id=model_id)
model_name = model_run.properties['model_name']
self.ctx.log('Regestiring model: %s' % model_name)
description = '%s-%s' % (model_name, iteration)
model = experiment_run.register_model(model_name=model_name,
iteration=iteration,
description=description,
tags=None)
script_file_name = '.azureml/score_script.py'
model_run.download_file('outputs/scoring_file_v_1_0_0.py',
script_file_name)
# Deploying ACI Service
aci_service_name = self._aci_service_name(model_name)
self.ctx.log('Deploying AciWebservice %s ...' % aci_service_name)
inference_config = InferenceConfig(
environment=model_run.get_environment(),
entry_script=script_file_name)
aciconfig = AciWebservice.deploy_configuration(
cpu_cores=1,
memory_gb=2,
tags={'type': "inference-%s" % aci_service_name},
description="inference-%s" % aci_service_name)
# Remove any existing service under the same name.
try:
Webservice(ws, aci_service_name).delete()
self.ctx.log('Remove any existing service under the same name...')
except WebserviceException:
pass
aci_service = Model.deploy(ws, aci_service_name, [model],
inference_config, aciconfig)
aci_service.wait_for_deployment(True)
self.ctx.log('%s state %s' %
(aci_service_name, str(aci_service.state)))
return {'model_id': model_id, 'aci_service_name': aci_service_name}
def deploy_new_web_service(workspace, service_name, aciconfig, image_config,
model):
service = Webservice.deploy_from_model(workspace=workspace,
name=service_name,
deployment_config=aciconfig,
models=[model],
image_config=image_config)
service.wait_for_deployment(show_output=True)
print('The URI to access the web service is: ', service.scoring_uri)
def deployModelAsWebService(
ws,
model_folder_path="models",
model_name="component_compliance",
scoring_script_filename="scoring_service.py",
conda_packages=['numpy', 'pandas'],
pip_packages=['azureml-sdk', 'onnxruntime'],
conda_file="dependencies.yml",
runtime="python",
cpu_cores=1,
memory_gb=1,
tags={'name': 'scoring'},
description='Compliance classification web service.',
service_name="complianceservice"):
# notice for the model_path, we supply the name of the outputs folder without a trailing slash
# this will ensure both the model and the customestimators get uploaded.
print("Registering and uploading model...")
registered_model = Model.register(model_path=model_folder_path,
model_name=model_name,
workspace=ws)
# create a Conda dependencies environment file
print("Creating conda dependencies file locally...")
from azureml.core.conda_dependencies import CondaDependencies
mycondaenv = CondaDependencies.create(conda_packages=conda_packages,
pip_packages=pip_packages)
with open(conda_file, "w") as f:
f.write(mycondaenv.serialize_to_string())
# create container image configuration
print("Creating container image configuration...")
from azureml.core.image import ContainerImage
image_config = ContainerImage.image_configuration(
execution_script=scoring_script_filename,
runtime=runtime,
conda_file=conda_file)
# create ACI configuration
print("Creating ACI configuration...")
from azureml.core.webservice import AciWebservice, Webservice
aci_config = AciWebservice.deploy_configuration(cpu_cores=cpu_cores,
memory_gb=memory_gb,
tags=tags,
description=description)
# deploy the webservice to ACI
print("Deploying webservice to ACI...")
webservice = Webservice.deploy_from_model(workspace=ws,
name=service_name,
deployment_config=aci_config,
models=[registered_model],
image_config=image_config)
webservice.wait_for_deployment(show_output=True)
return webservice
def run(model_path, model_name):
auth_args = {
'tenant_id': os.environ['TENANT_ID'],
'service_principal_id': os.environ['SERVICE_PRINCIPAL_ID'],
'service_principal_password': os.environ['SERVICE_PRINCIPAL_PASSWORD']
}
ws_args = {
'auth': ServicePrincipalAuthentication(**auth_args),
'subscription_id': os.environ['SUBSCRIPTION_ID'],
'resource_group': os.environ['RESOURCE_GROUP']
}
ws = Workspace.get(os.environ['WORKSPACE_NAME'], **ws_args)
print(ws.get_details())
print('\nSaving model {} to {}'.format(model_path, model_name))
model = Model.register(ws, model_name=model_name, model_path=model_path)
print('Done!')
print('Checking for existing service {}'.format(model_name))
service_name = 'simplemnist-svc'
if model_name in ws.webservices:
print('Found it!\nRemoving Existing service...')
ws.webservices[model_name].delete()
print('Done!')
else:
print('Not found, creating new one!')
# image configuration
image_config = ContainerImage.image_configuration(
execution_script="score.py",
runtime="python",
conda_file="environment.yml")
# deployement configuration
aciconfig = AciWebservice.deploy_configuration(cpu_cores=1,
memory_gb=1,
description=model_name)
# deploy
service = Webservice.deploy_from_model(workspace=ws,
name=model_name,
models=[model],
image_config=image_config,
deployment_config=aciconfig)
service.wait_for_deployment(show_output=True)
#print logs
print(service.get_logs())
print('Done!')
def deploy_new_webservice(ws, compute_name, webservice_name, image):
aks_target = ws.compute_targets[compute_name]
aks_config = AksWebservice.deploy_configuration(collect_model_data=True,
enable_app_insights=True)
service = Webservice.deploy_from_image(workspace=ws,
name=webservice_name,
image=image,
deployment_config=aks_config,
deployment_target=aks_target)
service.wait_for_deployment(show_output=True)
print(service.state)
def _deploy_remotly(self, model_id, model_run, ws, experiment):
from azureml.core.model import Model
from azureml.core.model import InferenceConfig
from azureml.core.webservice import Webservice
from azureml.core.webservice import AciWebservice
from azureml.exceptions import WebserviceException
from azureml.train.automl.run import AutoMLRun
# ws, experiment = self._get_experiment()
iteration, run_id = self._get_iteration(model_id)
experiment_run = AutoMLRun(experiment = experiment, run_id = run_id)
model_name = model_run.properties['model_name']
self.ctx.log('Registering model: %s' % model_id)
description = '%s-%s' % (model_name, iteration)
model = experiment_run.register_model(
model_name = model_name, iteration=iteration,
description = description, tags = None)
script_file_name = '.azureml/score_script.py'
model_run.download_file(
'outputs/scoring_file_v_1_0_0.py', script_file_name)
self._edit_score_script(script_file_name)
# Deploying ACI Service
aci_service_name = self._aci_service_name(model_name)
self.ctx.log('Deploying AciWebservice %s ...' % aci_service_name)
inference_config = InferenceConfig(
environment = model_run.get_environment(),
entry_script = script_file_name)
aciconfig = AciWebservice.deploy_configuration(
cpu_cores = 1,
memory_gb = 2,
tags = {'type': "inference-%s" % aci_service_name},
description = "inference-%s" % aci_service_name)
# Remove any existing service under the same name.
try:
Webservice(ws, aci_service_name).delete()
self.ctx.log('Remove any existing service under the same name...')
except WebserviceException:
pass
aci_service = Model.deploy(
ws, aci_service_name, [model], inference_config, aciconfig)
aci_service.wait_for_deployment(True)
self.ctx.log('%s state %s' % (aci_service_name, str(aci_service.state)))
return {'model_id': model_id, 'aci_service_name': aci_service_name}
def deploy_image():
ws = get_workspace()
azure_image = get_image()
aci_config = AciWebservice.deploy_configuration(cpu_cores=1,
memory_gb=1,
tags={'method': 'sklearn'},
description='Worst model',
location=LOCATION)
webservice = Webservice.deploy_from_image(image=azure_image,
workspace=ws,
name=MODEL_NAME,
deployment_config=aci_config)
webservice.wait_for_deployment(show_output=True)
def deploy_to_aci(model_image, workspace, dev_webservice_name):
from azureml.core.webservice import AciWebservice, Webservice
# Deploy a model image to ACI
print("Deploying to ACI...")
# make sure this dev_webservice_name is unique and doesnt already exist, else need to replace
dev_webservice_deployment_config = AciWebservice.deploy_configuration()
dev_webservice = Webservice.deploy_from_image(
name=dev_webservice_name,
image=model_image,
deployment_config=dev_webservice_deployment_config,
workspace=workspace)
dev_webservice.wait_for_deployment()
print("Deployment to ACI successfully complete")
return dev_webservice
def deploy(aciconfig, envfile, name, model):
# configure the image
image_config = ContainerImage.image_configuration(
execution_script="./score.py", runtime="python", conda_file=envfile)
service = Webservice.deploy_from_model(workspace=ws,
name=name,
deployment_config=aciconfig,
models=[model],
image_config=image_config)
service.wait_for_deployment(show_output=True)
print(service.scoring_uri)
def main():
ws = AzureMLUtils.get_workspace()
print("Workspace lookup successful")
# read command line parameters
service_name = getRuntimeArgs()
#look up service
service = Webservice(ws, service_name)
# look up scoring uri
scoring_uri = service.scoring_uri
#Get the first api key
if service.compute_type == "AKS":
api_key = service.get_keys()[0]
elif service.compute_type == "ACI":
api_key = "dummy"
else:
raise Exception("Unknown compute type")
# This line is needed to for Azure Devops to set api key and scoring uri as environment variable.
print("##vso[task.setvariable variable=TMP_SCORING_URI]", scoring_uri)
print("##vso[task.setvariable variable=TMP_API_KEY]", api_key)
class AzureMLService():
def __init__(self, ws:Workspace, service_name: str):
self.__ws = ws
self.__azure_service = Webservice(ws, service_name)
def make_request(self, inference_dataset_name):
inference_dataset = Dataset.get_by_name(self.__ws,inference_dataset_name)
df = inference_dataset.to_pandas_dataframe()
body = json.dumps({'data': json.loads(df.to_json(orient='values'))})
result = self.__azure_service.run(body)
print(result)
def deploy_container_instance(workspace, endpoint_name, inference_config,
model_azure):
# Remove any existing service under the same name.
try:
Webservice(workspace, endpoint_name).delete()
except WebserviceException:
pass
deployment_config = AciWebservice.deploy_configuration(cpu_cores=1,
memory_gb=1)
service = Model.deploy(workspace, endpoint_name, [model_azure],
inference_config, deployment_config)
service.wait_for_deployment(show_output=True)
print('A API {} foi gerada no estado {}'.format(service.scoring_uri,
service.state))
return service.scoring_uri
def deploy_to_aks(workspace, model_image, aks_target, prod_webservice_name="dsswe-mprodm"):
from azureml.core.webservice import Webservice, AksWebservice
# Deploy a model image to AKS
print("Deploying to AKS...")
# Set configuration and service name
prod_webservice_deployment_config = AksWebservice.deploy_configuration()
# Deploy from image
prod_webservice = Webservice.deploy_from_image(workspace=workspace,
name=prod_webservice_name,
image=model_image,
deployment_config=prod_webservice_deployment_config,
deployment_target=aks_target)
# Wait for the deployment to complete
prod_webservice.wait_for_deployment(show_output=True)
print("Deployment to AKS completed sucessfully")
return prod_webservice
def deployWebservice(ws, args, folders):
# this section requries that the processing is done in the directory where the execution script and the conda_file resides
os.chdir(folders.script_folder)
model = Model(ws, args.modelName)
aciconfig = AciWebservice.deploy_configuration(cpu_cores=args.cpuCores,
memory_gb=args.memoryGB)
# configure the image
image_config = ContainerImage.image_configuration(
execution_script=args.scoringScript,
runtime="python",
conda_file=args.environmentFileName)
service = Webservice.deploy_from_model(workspace=ws,
name=args.webserviceName,
deployment_config=aciconfig,
models=[model],
image_config=image_config)
service.wait_for_deployment(show_output=True)
return service.scoring_uri
def deploy_image():
ws = get_workspace()
azure_image = get_image()
# Set the web service configuration (using default here with app insights)
aks_config = AksWebservice.deploy_configuration(enable_app_insights=True)
# Unique service name
service_name = AKS_NAME
aks_target = get_cluster()
# Webservice creation using single command
aks_service = Webservice.deploy_from_image(workspace=ws,
name=service_name,
deployment_config=aks_config,
image=azure_image,
deployment_target=aks_target)
aks_service.wait_for_deployment(show_output=True)
def deploy_service(execution_script,
conda_file,
aciconfig,
service_name,
model,
workspace,
runtime="python"):
image_config = ContainerImage.image_configuration(
execution_script=execution_script,
runtime=runtime,
conda_file=conda_file)
service = Webservice.deploy_from_model(workspace=workspace,
name=service_name,
deployment_config=aciconfig,
models=[model],
image_config=image_config)
service.wait_for_deployment(show_output=True)
print(service.scoring_uri)
return service
def run_deployment() -> None:
"""Train the model to Azure."""
# Create or load an existing Azure ML workspace. You can also load an existing workspace using
azure_workspace = Workspace.get(
name=model_config.WORKSPACE_NAME,
subscription_id=model_config.SUBSCRIPTION_ID,
resource_group=model_config.RESOURCE_GROUP)
# azure_workspace = Workspace.create(name=model_config.WORKSPACE_NAME,
# subscription_id=model_config.SUBSCRIPTION_ID,
# resource_group=model_config.RESOURCE_GROUP,
# location=model_config.LOCATION,
# create_resource_group=False,
# exist_okay=True)
# Build an Azure ML container image for deployment
model_path = config.S3_MODEL_PATH
azure_image, azure_model = mlflow.azureml.build_image(
model_uri=model_path,
workspace=azure_workspace,
description="Loan classification model 1",
synchronous=True)
# If your image build failed, you can access build logs at the following URI:
print("Access the following URI for build logs: {}".format(
azure_image.image_build_log_uri))
# Deploy the container image to ACI
webservice_deployment_config = AciWebservice.deploy_configuration()
webservice = Webservice.deploy_from_image(
image=azure_image,
workspace=azure_workspace,
name=model_config.DEPLOYMENT_NAME)
webservice.wait_for_deployment()
# After the image deployment completes, requests can be posted via HTTP to the new ACI
# webservice's scoring URI.
print("Scoring URI is: %s", webservice.scoring_uri)
def azureml_build_deploy(runid, workspace, model_name, image_name,
deploy_name):
# Build an Azure ML Container Image for an MLflow
azure_image, azure_model = mlflow.azureml.build_image(
model_uri='runs:/{}/{}'.format(runid, MODEL_SAVE_PATH),
workspace=workspace,
model_name=model_name,
image_name=image_name,
synchronous=True)
# Deploy the image to Azure Container Instances (ACI) for real-time serving
aci_config = AciWebservice.deploy_configuration()
deployment_stub = ''.join(
[random.choice(string.ascii_lowercase) for i in range(5)])
print("Deploying as " + deploy_name + "-" + deployment_stub)
webservice = Webservice.deploy_from_image(image=azure_image,
workspace=workspace,
name=deploy_name + "-" +
deployment_stub,
deployment_config=aci_config)
webservice.wait_for_deployment()
return webservice
from azureml.core import Workspace
from azureml.core.webservice import Webservice
# Requires the config to be downloaded first to the current working directory
ws = Workspace.from_config()
# Set with the deployment name
name = "automl-voting-ensemble"
# load existing web service
service = Webservice(name=name, workspace=ws)
service.update(enable_app_insights=True)
logs = service.get_logs()
for line in logs.split('\n'):
print(line)
