Merge pull request #115985 from shivp950/patch-14

updated deployment section

Author: PRMerger20

articles/machine-learning/how-to-use-mlflow.md

@@ -261,26 +261,9 @@ runid = runs[0].id
 model_save_path = 'model'
 ```
 
-### Create Docker image
+### Deploy the model
 
-The `mlflow.azureml.build_image()` function builds a Docker image from the saved model in a framework-aware manner. It automatically creates the framework-specific inferencing wrapper code and specifies package dependencies for you. Specify the model path, your workspace, run ID and other parameters.
-
-The following code builds a docker image using *runs:/<run.id>/model* as the model_uri path for a Scikit-learn experiment.
-
-```python
-import mlflow.azureml
-
-azure_image, azure_model = mlflow.azureml.build_image(model_uri='runs:/{}/{}'.format(runid, model_save_path),
-                                                      workspace=ws,
-                                                      model_name='sklearn-model',
-                                                      image_name='sklearn-image',
-                                                      synchronous=True)
-```
-The creation of the Docker image can take several minutes. 
-
-### Deploy the Docker image 
-
-After the image is created, use the Azure Machine Learning SDK to deploy the image as a web service.
+Use the Azure Machine Learning SDK to deploy the model as a web service.
 
 First, specify the deployment configuration. Azure Container Instance (ACI) is a suitable choice for a quick dev-test deployment, while Azure Kubernetes Service (AKS) is suitable for scalable production deployments.
 
@@ -299,30 +282,21 @@ aci_config = AciWebservice.deploy_configuration(cpu_cores=1,
                                                 location='eastus2')
 ```
 
-Then, deploy the image by using the Azure Machine Learning SDK [deploy_from_image()](/python/api/azureml-core/azureml.core.webservice.webservice(class)?view=azure-ml-py#deploy-from-image-workspace--name--image--deployment-config-none--deployment-target-none--overwrite-false-) method. 
+Then, register and deploy the model by using the Azure Machine Learning SDK [deploy](/python/api/azureml-core/azureml.core.model.model?view=azure-ml-py#deploy-workspace--name--models--inference-config-none--deployment-config-none--deployment-target-none--overwrite-false-) method. 
 
 ```python
-webservice = Webservice.deploy_from_image( image=azure_image, 
-                                           workspace=ws, 
-                                           name='diabetes-model-1', 
-                                           deployment_config=aci_config)
+(webservice,model) = mlflow.azureml.deploy( model_uri='runs:/{}/{}'.format(run.id, model_path),
+                      workspace=ws,
+                      model_name='sklearn-model', 
+                      service_name='diabetes-model-1', 
+                      deployment_config=aci_config, 
+                      tags=None, mlflow_home=None, synchronous=True)
 
 webservice.wait_for_deployment(show_output=True)
 ```
 #### Deploy to AKS
 
-To deploy to AKS, first create an AKS cluster and bring over the Docker image you want to deploy. For this example, bring over the previously created image from the ACI deployment.
-
-To get the image from the previous ACI deployment use the [Image](https://docs.microsoft.com/python/api/azureml-core/azureml.core.image.image.image?view=azure-ml-py) class. 
-
-```python
-from azureml.core.image import Image
-
-# Get the image by name, you can change this based on the image you want to deploy
-myimage = Image(workspace=ws, name='sklearn-image') 
-```
-
-Create an AKS cluster using the [ComputeTarget.create()](https://docs.microsoft.com/python/api/azureml-core/azureml.core.computetarget?view=azure-ml-py#create-workspace--name--provisioning-configuration-) method. It may take 20-25 minutes to create a new cluster.
+To deploy to AKS, first create an AKS cluster. Create an AKS cluster using the [ComputeTarget.create()](https://docs.microsoft.com/python/api/azureml-core/azureml.core.computetarget?view=azure-ml-py#create-workspace--name--provisioning-configuration-) method. It may take 20-25 minutes to create a new cluster.
 
 ```python
 from azureml.core.compute import AksCompute, ComputeTarget
@@ -346,26 +320,25 @@ Set up your deployment configuration with the [deploy_configuration()](https://d
 
 ```python
 from azureml.core.webservice import Webservice, AksWebservice
-from azureml.core.image import ContainerImage
 
 # Set the web service configuration (using default here with app insights)
-aks_config = AksWebservice.deploy_configuration(enable_app_insights=True)
+aks_config = AksWebservice.deploy_configuration(enable_app_insights=True, compute_target_name='aks-mlflow')
 
-# Unique service name
-service_name ='aks-service'
-```
 
-Then, deploy the image by using the Azure Machine Learning SDK [deploy_from_image()](/python/api/azureml-core/azureml.core.webservice.webservice(class)?view=azure-ml-py#deploy-from-image-workspace--name--image--deployment-config-none--deployment-target-none--overwrite-false-) method. 
+Then, deploy the image by using the Azure Machine Learning SDK [deploy()](Then, register and deploy the model by using the Azure Machine Learning SDK [deploy](/python/api/azureml-core/azureml.core.model.model?view=azure-ml-py#deploy-workspace--name--models--inference-config-none--deployment-config-none--deployment-target-none--overwrite-false-) method. 
 
 ```python
 # Webservice creation using single command
-aks_service = Webservice.deploy_from_image( workspace=ws, 
-                                            name=service_name,
-                                            deployment_config = aks_config
-                                            image = myimage,
-                                            deployment_target = aks_target)
+from azureml.core.webservice import AksWebservice, Webservice
+(webservice, model) = mlflow.azureml.deploy( model_uri='runs:/{}/{}'.format(run.id, model_path),
+                      workspace=ws,
+                      model_name='sklearn-model', 
+                      service_name='my-aks', 
+                      deployment_config=aks_config, 
+                      tags=None, mlflow_home=None, synchronous=True)
+
 
-aks_service.wait_for_deployment(show_output=True)
+webservice.wait_for_deployment()
 ```
 
 The service deployment can take several minutes.