Take a deep look into Azure Machine Learning, a cloud service that helps you build, train, deploy, and manage models. Walk through the data science process and then have some fun creating a ML recognition model based on the Simpsons cartoon with PyTorch. You'll leave this session with a better grasp of the technological components of Azure Machine Learning services.

1. Henk Boelman
Cloud Advocate @ Microsoft
Build and Deploy
PyTorch Models with
Azure Machine
Learning
HenkBoelman.com @hboelman

2. Azure Machine Learning studio
A fully-managed cloud service that enables you to easily build, deploy, and share
predictive analytics solutions.

3. What is Azure Machine Learning?
Set of Azure
Cloud Services
Python SDK
Azure CLI
Web interface
✓Prepare Data
✓Build Models
✓Train Models
✓Manage Models
✓Track
Experiments
✓Deploy Models
That enables
you to:

4. Is it Marge or Homer or .. or ..?

5. Simpson Lego dataset
https://github.com/hnky/dataset-lego-figures

6. Sophisticated pretrained models
To simplify solution development
Azure
Databricks
Machine Learning
VMs
Popular frameworks
To build advanced deep learning solutions TensorFlow KerasPytorch Onnx
Azure
Machine Learning
LanguageSpeech
…
Azure
Search
Vision
On-premises Cloud Edge
Productive services
To empower data science and development teams
Powerful infrastructure
To accelerate deep learning
Flexible deployment
To deploy and manage models on intelligent cloud and edge
Machine Learning on Azure
Cognitive Services

8. Prepare
your environment
Experiment
with your model & data
Deploy
Your model into production

9. Step 1: Prepare your environment

10. Setup your environment
VS Code Python SDKAzure Portal

11. Create a workspace
ws = Workspace.create(
name='<NAME>',
subscription_id='<SUBSCRIPTION ID>',
resource_group='<RESOURCE GROUP>',
location='westeurope')
ws.write_config()
ws = Workspace.from_config()
Create a workspace

13. Datasets – registered, known data sets
Experiments – Training runs
Models – Registered, versioned models
Endpoints:
Real-time Endpoints – Deployed model endpoints
Pipeline Endpoints – Training workflows
Compute – Managed compute
Datastores – Connections to data
Azure Machine Learning Service

14. Create Compute
cfg = AmlCompute.provisioning_configuration(
vm_size='STANDARD_NC6',
min_nodes=1,
max_nodes=6)
cc = ComputeTarget.create(ws, '<NAME>', cfg)
Create a workspace Create compute

16. Demo: Setup your workspace
Create a workspace Create compute Setup storage

17. Step 1
Prepare your environment
Create a workspace Create compute Setup storage

18. Step 2: Experiment with your model & data

19. Create an experiment
exp = Experiment(workspace=ws, name=“<ExperimentName>”)
Create an
Experiment

20. Create a training file
Create an
Experiment
Create a
training file

21. Create an estimator
params = {'--data-folder': ws.get_default_datastore().as_mount()}
estimator = TensorFlow(
source_directory = script_folder,
script_params = params,
compute_target = computeCluster,
entry_script = 'train.py’,
use_gpu = True,
conda_packages = ['scikit-learn','keras','opencv’],
framework_version='1.10')
Create an
Experiment
Create a
training file
Create an
estimator

22. Submit the experiment to the cluster
run = exp.submit(estimator)
RunDetails(run).show()
Create an
Experiment
Create a
training file
Submit to the
AI cluster
Create an
estimator

23. Create an
Experiment
Create a
training file
Submit to the
AI cluster
Create an
estimator
Demo: Creating and run an experiment

24. Azure Notebook
Compute Target
Experiment Docker Image
Data store
1. Snapshot folder and
send to experiment
2. create docker image
3. Deploy docker
and snapshot to
compute
4. Mount datastore
to compute
6. Stream
stdout,
logs,
metrics 5. Launch the script
7. Copy over
outputs

25. Register the model
model = run.register_model(
model_name='SimpsonsAI',
model_path='outputs')
Create an
Experiment
Create a
training file
Submit to the
AI cluster
Create an
estimator
Register the
model

26. Create an
Experiment
Create a
training file
Submit to the
AI cluster
Create an
estimator
Register the
model
Demo: Register and test the model

27. Step 2
Experiment with your model & data

28. Step 3: Deploy your model into production

29. AMLS to deploy
The Model Score.py Environment file
Docker Image

30. Score.py
%%writefile score.py
from azureml.core.model import Model
def init():
model_root = Model.get_model_path('MyModel’)
loaded_model = model_from_json(loaded_model_json)
loaded_model.load_weights(model_file_h5)
def run(raw_data):
url = json.loads(raw_data)['url’]
image_data = cv2.resize(image_data,(96,96))
predicted_labels = loaded_model.predict(data1)
return json.dumps(predicted_labels)

31. Environment File
from azureml.core.runconfig import CondaDependencies
cd = CondaDependencies.create()
cd.add_conda_package('keras==2.2.2')
cd.add_conda_package('opencv')
cd.add_tensorflow_conda_package()
cd.save_to_file(base_directory='./', conda_file_path='myenv.yml')

32. Inference config
inference_config = InferenceConfig(
runtime= "python",
entry_script="score.py",
conda_file="myenv.yml"
)

33. Deployment using AMLS

34. Deploy to ACI
aciconfig = AciWebservice.deploy_configuration(
cpu_cores = 1,
memory_gb = 2)
service = Model.deploy(workspace=ws,
name='simpsons-aci',
models=[model],
inference_config=inference_config,
deployment_config=aciconfig)

35. Demo: Deploy to ACI

37. @hboelman
github.com/hnky
henkboelman.com
Thank you!
Read more on: henkboelman.com