Merge branch 'master' of github.com:MicrosoftDocs/azure-docs-pr into release-cli-rebrand

Author: SyntaxC4

articles/machine-learning/service/concept-azure-machine-learning-architecture.md

@@ -163,7 +163,7 @@ A compute target is the compute resource used to run your training script or hos
 * Azure Container Instance
 * Azure Kubernetes Service
 
-Compute targets are attached to a workspace. Computer targets other than the local machine are shared by users of the workspace.
+Compute targets are attached to a workspace. Compute targets other than the local machine are shared by users of the workspace.
 
 Most compute targets can be created directly through the workspace by using the Azure portal, Azure Machine Learning SDK, or Azure CLI. If you have compute targets that were created by another process (for example, the Azure portal or Azure CLI), you can add (attach) them to your workspace. Some compute targets must be created outside the workspace, and then attached.
 

articles/machine-learning/service/concept-ml-pipelines.md

@@ -57,7 +57,7 @@ Use Python to create your ML pipelines. The Azure Machine Learning SDK offers im
 
 Using declarative data dependencies, you can optimize your tasks. The SDK includes a framework of pre-built modules for common tasks such as data transfer, compute target creation, and model publishing. The framework can be extended to model your own conventions by implementing custom steps that are reusable across pipelines.
 
-Pipelines can be saved as templates so you can schedule batch-scoring or retraining jobs.
+Pipelines can be saved as templates and can be deployed to a REST endpoint so you can schedule batch-scoring or retraining jobs.
 
 Check out the [Python SDK reference docs for pipelines](http://aka.ms/aml-sdk).
 

articles/machine-learning/service/concept-model-management-and-deployment.md

@@ -49,12 +49,14 @@ Images allow for reliable model deployment, along with all components needed to
 
 The image can also include SDK components for logging and monitoring. The SDK logs data can be used to fine-tune or retrain your model, including the input and output of the model.
 
+Azure Machine Learning supports the most popular frameworks, but in general any framework that can be pip installed can work.
+
 When your workspace was created, so were other several other Azure resources used by that workspace.
 All the objects used to create the image are stored in the Azure storage account in your workspace. The image is created and stored in the Azure Container Registry. You can provide additional metadata tags when creating the image, which are also stored by the image registry and can be queried to find your image.
 
 ## Step 3: Deployment
 
-You can deploy registered images into the cloud or to edge devices. The deployment process creates all the resources needed to monitor, load-balance, and auto-scale your model. You can also upgrade an existing deployment to use a newer image.
+You can deploy registered images into the cloud or to edge devices. The deployment process creates all the resources needed to monitor, load-balance, and auto-scale your model. Access to the deployed services can be secured with certificate based authentication by providing the security assets during deployment. You can also upgrade an existing deployment to use a newer image.
 
 Web service deployments are also searchable. For example, you can search for all deployments of a specific model or image.
 

articles/machine-learning/service/how-to-migrate.md

@@ -17,7 +17,7 @@ ms.date: 09/24/2018
 **If you have installed the Workbench (preview) application and/or have experimentation and model management preview accounts, use this article to migrate to the latest version.**  If you don't have preview Workbench installed, or an experimentation and/or model management account, you don't need to migrate anything.
 
 ## What can I migrate?
-Most artifacts created in the first preview of Azure Machine Learning service are stored in your own local or cloud storage. These artifacts won't disappear. To migrate, register the artifacts again with the updated Azure Machine Learning offering. 
+Most artifacts created in the first preview of Azure Machine Learning service are stored in your own local or cloud storage. These artifacts won't disappear. To migrate, register the artifacts again with the updated Azure Machine Learning service. 
 
 The following table and article explain what you can do with your existing assets and resources before or after moving over to the latest version of Azure Machine Learning service. You can also continue to use the previous version and your assets for some time ([see transition support timeline](overview-what-happened-to-workbench.md#timeline)).
 
@@ -76,7 +76,7 @@ In the latest version, models are deployed as web services to [Azure Container I
 Learn more in these articles:
 + [Deploy to ACI](how-to-deploy-to-aci.md)
 + [Deploy to AKS](how-to-deploy-to-aks.md)
-+ [Tutorial:Deploy models with Azure Machine Learning service](tutorial-deploy-models-with-aml.md)
++ [Tutorial: Deploy models with Azure Machine Learning service](tutorial-deploy-models-with-aml.md)
 
 When [support for the previous CLI ends](overview-what-happened-to-workbench.md#timeline), you won't be able to manage the web services you originally deployed with your Model Management account. However, those web services will continue to work for as long as Azure Container Service (ACS) is still supported.
 

articles/machine-learning/service/how-to-set-up-training-targets.md

@@ -32,7 +32,7 @@ Azure Machine Learning supports the following compute targets:
 
 __[Azure Container Instances (ACI)](#aci)__ can also be used to train models. It is a serverless cloud offering that is inexpensive and easy to create and work with. ACI does not support GPU acceleration, automated hyper parameter tuning, or automated model selection. Also, it cannot be used in a pipeline.
 
-The key differentiators between the computer targets are:
+The key differentiators between the compute targets are:
 * __GPU acceleration__: GPUs are available with the Data Science Virtual Machine and Azure Batch AI. You may have access to a GPU on your local computer, depending on the hardware, drivers, and frameworks that are installed.
 * __Automated hyperparameter tuning__: Azure Machine Learning automated hyperparameter optimization helps you find the best hyperparameters for your model.
 * __Automated model selection__: Azure Machine Learning can intelligently recommend algorithm and hyperparameter selection when building a model. Automated model selection helps you converge to a high-quality model faster than manually trying different combinations. For more information, see the [Tutorial: Automatically train a classification model with Azure Automated Machine Learning](tutorial-auto-train-models.md) document.

articles/machine-learning/service/overview-what-happened-to-workbench.md

@@ -31,7 +31,7 @@ While there are new improved CLI and SDK clients in the current release, the des
 
 ## How do I migrate?
 
-Most of the artifacts created in the earlier version of the Azure Machine Learning service are stored in your own local or cloud storage. These artifacts won't ever disappear. To migrate, you need to register the artifacts again with the updated Azure Machine Learning offering. Learn what you can migrate and how in this [migration article](how-to-migrate.md).
+Most of the artifacts created in the earlier version of the Azure Machine Learning service are stored in your own local or cloud storage. These artifacts won't ever disappear. To migrate, you need to register the artifacts again with the updated Azure Machine Learning service. Learn what you can migrate and how in this [migration article](how-to-migrate.md).
 
 <a name="timeline"></a>
 
@@ -71,21 +71,27 @@ You can use this same SDK for larger data sets or use Azure Databricks to prepar
 
 You won't lose any code or work. In the older version, projects are cloud entities with a local directory. In the latest version, you attach local directories to the Azure Machine Learning Workspace using a local config file. [See a diagram of the latest architecture](concept-azure-machine-learning-architecture.md).
 
-Since much of the project contents was already on your local machine, you just need to create a config file in that directory and reference it in your code to connect to your workspace. [Learn how migrate your existing projects.](how-to-migrate.md#projects)
+Since much of the project content was already on your local machine, you just need to create a config file in that directory and reference it in your code to connect to your workspace. [Learn how migrate your existing projects.](how-to-migrate.md#projects)
 
 Learn how to get started [in Python with the SDK](quickstart-get-started.md).
 
+## What about my registers models and images?
+ 
+The models that you registered in your old model registry must migrated to your new workspace if you want to continue to use them. You can do this by [downloading the models and re-registering them](how-to-migrate.md) in your new workspace. 
+
+The images that you created in your old image registry must be re-created in the new workspace to continue to use them. You can do this by following the [create docker image](how-to-deploy-to-aci.md) section. 
+
 ## What about deployed web services?
 
 The models you deployed as web services using your Model Management account will continue to work for as long as Azure Container Service (ACS) is supported. Those web services will even work after support has ended for Model Management accounts. However, when support for the old CLI ends, so does your ability to manage those web services.
 
 In the newer version, models are deployed as web services to [Azure Container Instances](how-to-deploy-to-aci.md) (ACI) or [Azure Kubernetes Service](how-to-deploy-to-aks.md) (AKS) clusters. You can also [deploy to FPGAs and to the IoT edge](how-to-deploy-and-where.md). Without having to change any of your scoring files, dependencies, and schemas, you can redeploy your models using the new SDK or CLI. 
 
-## What about the SDK & CLI?
+## What about the old SDK & CLI?
 
 Yes, they will continue to work for a while (see the [timeline](#timeline) above). We recommend that you start creating your new experiments and models with the latest SDK and/or CLI.
 
-In the latest release, the new Python SDK allows you to interact with the Azure Machine Learning service in any Python environment. Learn how to install the <a href="http://aka.ms/aml-sdk" target="_blank">SDK</a>.  You can also use the [updated Azure CLI machine learning extension](reference-azure-machine-learning-cli.md) with the rich set of `az ml` commands to interact the service in any command-line environment, including Azure portal cloud shell.
+In the latest release, the new Python SDK allows you to interact with the Azure Machine Learning service in any Python environment. Learn how to install the latest <a href="http://aka.ms/aml-sdk" target="_blank">SDK</a>.  You can also use the [updated Azure CLI machine learning extension](reference-azure-machine-learning-cli.md) with the rich set of `az ml` commands to interact the service in any command-line environment, including Azure portal cloud shell.
 
 ## What about VS Code Tools for AI?