PR edits

Author: ssalgadodev

articles/machine-learning/how-to-interactive-jobs.md

@@ -1,5 +1,5 @@
 ---
-title: Interact with your jobs (debug & monitor)
+title: Interact with your jobs (debug and monitor)
 titleSuffix: Azure Machine Learning
 description: Debug or monitor your Machine Learning job as it runs on AzureML compute with your training application of choice. 
 services: machine-learning
@@ -16,39 +16,40 @@ ms.date: 03/15/2022
 
 # Debug jobs and monitor training progress
 
-Machine learning model training is usually an iterative process and requires significant experimentation. With the AzureML interactive job experience, data scientists can use the AzureML Python SDKv2, AzureML CLIv2 or AzureML Studio to access the container where their job is running.  Once the job container is accessed, users can iterate on training scripts, monitor training progress or debug the job remotely like they typically do on their local machines. Jobs can be interacted with via different training applications including **JupyterLab, TensorBoard, VS Code** or by connecting to the job container directly via **SSH**.  
+Machine learning model training is usually an iterative process and requires significant experimentation. With the Azure Machine Learning interactive job experience, data scientists can use the Azure Machine Learning Python SDKv2, Azure Machine Learning CLIv2 or the Azure Studio to access the container where their job is running.  Once the job container is accessed, users can iterate on training scripts, monitor training progress or debug the job remotely like they typically do on their local machines. Jobs can be interacted with via different training applications including **JupyterLab, TensorBoard, VS Code** or by connecting to the job container directly via **SSH**.  
 
-Interactive training is supported on **AzureML Compute Cluster** and **Azure Arc-enabled Kubernetes Cluster**.
+Interactive training is supported on **Azure Machine Learning Compute Clusters** and **Azure Arc-enabled Kubernetes Cluster**.
 
-## Pre-requisites
-- Review [getting started with training on AzureML](./how-to-train-model.md).
+## Prerequisites
+- Review [getting started with training on Azure Machine Learning](./how-to-train-model.md).
 - To use **VS Code**, [follow this guide](how-to-setup-vs-code.md) to set up the Azure Machine Learning extension.
-- Make sure your job environment has the `openssh-server` and `ipykernel ~=6.0` packages installed (all AzureML curated training environments have these packages installed by default).
-- Interactive applications can't be enabled on distributed training runs where the distribution type is anything other than Pytorch, Tensorflow or MPI. Custom distributed training setup (configuring multi-node training without using the above distribution frameworks) is not currently supported. 
+- Make sure your job environment has the `openssh-server` and `ipykernel ~=6.0` packages installed (all Azure Machine Learning curated training environments have these packages installed by default).
+- Interactive applications can't be enabled on distributed training runs where the distribution type is anything other than Pytorch, Tensorflow or MPI. Custom distributed training setup (configuring multi-node training without using the above distribution frameworks) is not currently supported.
 
-![screenshot supported-distribution-types](media/interactive-jobs/supported-distribution-types.png)
+The supported distribution types are:
+  * PyTorch
+  * Mpi
+  * TensorFlow
 
 ## Interact with your job container
 
 By specifying interactive applications at job creation, you can connect directly to the container on the compute node where your job is running. Once you have access to the job container, you can test or debug your job in the exact same environment where it would run. You can also use VS Code to attach to the running process and debug as you would locally. 
 
 ### Enable during job submission
-# [AzureML Studio](#tab/ui)
+# [Azure Machine Learning Studio](#tab/ui)
 1. Create a new job from the left navigation pane in the studio portal.
 
-  ![screenshot select-job-ui](media/interactive-jobs/create-job.png) 
 
 2. Choose `Compute cluster` or `Attached compute` (Kubernetes) as the compute type, choose the compute target, and specify how many nodes you need in `Instance count`. 
 
-  ![screenshot select-compute-ui](media/interactive-jobs/select-compute.png) 
+  :::image type="content" source="./media/interactive-jobs/select-compute.png" alt-text="Screenshot of selecting a compute location for a job.":::
 
 3. Follow the wizard to choose the environment you want to start the job.
 
-  ![screenshot select-environment-ui](media/interactive-jobs/select-environment.png) 
 
 4. In `Job settings` step, add your training code (and input/output data) and reference it in your command to make sure it's mounted to your job.
 
-  ![screenshot set-command](media/interactive-jobs/sleep-command.png) 
+  :::image type="content" source="./media/interactive-jobs/sleep-command.png" alt-text="Screenshot of reviewing a drafted job and completing the creation.":::
 
   You can put `sleep <specific time>` at the end of your command to specify the amount of time you want to reserve the compute resource. The format follows: 
       * sleep 1s
@@ -63,7 +64,7 @@ By specifying interactive applications at job creation, you can connect directly
 
 5. Select the training applications you want to use to interact with the job.
 
-  ![screenshot select-apps](media/interactive-jobs/select-training-apps.png) 
+  :::image type="content" source="./media/interactive-jobs/select-training-apps.png" alt-text="Screenshot of selecting a training application for the user to use for a job.":::
 
 6. Review and create the job.
 
@@ -171,21 +172,22 @@ By specifying interactive applications at job creation, you can connect directly
 ---
 
 ### Connect to endpoints
-# [AzureML Studio](#tab/ui)
+# [Azure Machine Learning Studio](#tab/ui)
 To interact with your running job, click the button **Debug and monitor** on the job details page. 
 
-![screenshot debug-and-monitor](media/interactive-jobs/debug-and-monitor.png)
+:::image type="content" source="media/interactive-jobs/debug-and-monitor.png" alt-text="Screenshot of interactive jobs debug and monitor panel location.":::
+
 
 Clicking the applications in the panel opens a new tab for the applications. You can access the applications only when they are in **Running** status and only the **job owner** is authorized to access the applications. If you're training on multiple nodes, you can pick the specific node you would like to interact with.
 
-![screenshot ij-right-panel](media/interactive-jobs/ij-right-panel.png)
+:::image type="content" source="media/interactive-jobs/interactive-jobs-right-panel.png" alt-text="Screenshot of interactive jobs right panel information. Information content will vary depending on the users data":::
 
 It might take a few minutes to start the job and the training applications specified during job creation.
 
 # [Python SDK](#tab/python)
 - Once the job is submitted, you can use `ml_client.jobs.show_services("<job name>", <compute node index>)` to view the interactive service endpoints.
 
-- To connect via SSH to the container where the job is running, run the command `az ml job connect-ssh --name <job-name> --node-index <compute node index> --private-key-file-path <path to private key>`. To set up the AzureML CLIv2, follow this [guide](./how-to-configure-cli.md). 
+- To connect via SSH to the container where the job is running, run the command `az ml job connect-ssh --name <job-name> --node-index <compute node index> --private-key-file-path <path to private key>`. To set up the Azure Machine Learning CLIv2, follow this [guide](./how-to-configure-cli.md). 
 
 You can find the reference documentation for the SDKv2 [here](/sdk/azure/ml).
 
@@ -205,45 +207,45 @@ You can access the applications only when they are in **Running** status and onl
 ### Interact with the applications
 When you click on the endpoints to interact when your job, you're taken to the user container under your working directory, where you can access your code, inputs, outputs, and logs. If you run into any issues while connecting to the applications, the interactive capability and applications logs can be found from **system_logs->interactive_capability** under **Outputs + logs** tab.
 
-![screenshot check-logs](./media/interactive-jobs/ij-logs.png)
+:::image type="content" source="./media/interactive-jobs/interactive-jobs-logs.png" alt-text="Screenshot of interactive jobs interactive logs panel location.":::
 
 - You can open a terminal from Jupyter Lab and start interacting within the job container. You can also directly iterate on your training script with Jupyter Lab. 
 
-  ![screenshot jupyter-lab](./media/interactive-jobs/jupyter-lab.png)
+  :::image type="content" source="./media/interactive-jobs/jupyter-lab.png" alt-text="Screenshot of interactive jobs Jupyter lab content panel.":::
 
 - You can also interact with the job container within VS Code. To attach a debugger to a job during job submission and pause execution, [navigate here](./how-to-interactive-jobs.md#attach-a-debugger-to-a-job).
 
-  ![screenshot vs-code-open](./media/interactive-jobs/vs-code-open.png)
+  :::image type="content" source="./media/interactive-jobs/vs-code-open.png" alt-text="Screenshot of interactive jobs VS Code panel when first opened. This shows the sample python file that was created to print two lines.":::
 
 - If you have logged tensorflow events for your job, you can use TensorBoard to monitor the metrics when your job is running.
 
-  ![screenshot tensorboard-open](./media/interactive-jobs/tensorboard-open.png)
+  :::image type="content" source="./media/interactive-jobs/tensorboard-open.png" alt-text="Screenshot of interactive jobs tensorboard panel when first opened. This information will vary depending upon customer data":::
 
 ### End job
 Once you're done with the interactive training, you can also go to the job details page to cancel the job which will release the compute resource. Alternatively, use `az ml job cancel -n <your job name>` in the CLI or `ml_client.job.cancel("<job name>")` in the SDK. 
 
-![screenshot cancel-job](./media/interactive-jobs/cancel-job.png)
+:::image type="content" source="./media/interactive-jobs/cancel-job.png" alt-text="Screenshot of interactive jobs cancel job option and its location for user selection":::
 
 ## Attach a debugger to a job
-To submit a job with a debugger attached and the execution paused, you can use debugpy & VS Code (`debugpy` must be installed in your job environment). 
+To submit a job with a debugger attached and the execution paused, you can use debugpy and VS Code (`debugpy` must be installed in your job environment). 
 
 1. During job submission (either through the UI, the CLIv2 or the SDKv2) use the debugpy command to run your python script. For example, the below screenshot shows a sample command that uses debugpy to attach the debugger for a tensorflow script (`tfevents.py` can be replaced with the name of your training script).
 
-![screenshot use-debugpy](./media/interactive-jobs/use-debugpy.png)
+:::image type="content" source="./media/interactive-jobs/use-debugpy.png" alt-text="Screenshot of interactive jobs configuration of debugpy":::
 
 2. Once the job has been submitted, [connect to the VS Code](./how-to-interactive-jobs.md#connect-to-endpoints), and click on the in-built debugger.
 
-   ![screenshot open-debugger](./media/interactive-jobs/open-debugger.png)
+   :::image type="content" source="./media/interactive-jobs/open-debugger.png" alt-text="Screenshot of interactive jobs location of open debugger on the left side panel":::
 
 3. Use the "Remote Attach" debug configuration to attach to the submitted job and pass in the path and port you configured in your job submission command. You can also find this information on the job details page.
 
-   ![screenshot debug-path-and-port](./media/interactive-jobs/debug-path-and-port.png)
-   
-   ![screenshot remote-attach](./media/interactive-jobs/remote-attach.png)
+   :::image type="content" source="./media/interactive-jobs/debug-path-and-port.png" alt-text="Screenshot of interactive jobs completed jobs":::
+
+   :::image type="content" source="./media/interactive-jobs/remote-attach.png" alt-text="Screenshot of interactive jobs add a remote attach button":::
 
 4. Set breakpoints and walk through your job execution as you would in your local debugging workflow. 
 
-   ![screenshot set-breakpoints](./media/interactive-jobs/set-breakpoints.png)
+    :::image type="content" source="./media/interactive-jobs/set-breakpoints.png" alt-text="Screenshot of location of an example breakpoint that is set in the Visual Studio Code editor":::
 
 
 > [!NOTE]

articles/machine-learning/media/interactive-jobs/ij-logs.png renamed to articles/machine-learning/media/interactive-jobs/interactive-jobs-logs.png

@@ -480,6 +480,9 @@
             - name: Use automated ML with Databricks
               displayName: automl
               href: how-to-configure-databricks-automl-environment.md
+            - name: Debug jobs and monitor training progress
+              displayName: automl
+              href: how-to-interactive-jobs.md
             - name: Prep image data for computer vision models (Python)
               displayName: SDK, automl, image, datasets, conversion scripts, schema, image model
               href: how-to-prepare-datasets-for-automl-images.md