section distinction

Author: nibaccam

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

@@ -190,6 +190,7 @@ On the [Azure portal](https://ms.portal.azure.com), you can link your Azure Data
 
 After you instantiate your workspace, set the MLflow tracking URI. By doing so, you link the MLflow tracking to Azure Machine Learning workspace. After this, all your experiments will land in the managed Azure Machine Learning tracking service.
 
+#### Directly set MLflow Tracking in you notebook
 
 ```python
 uri = ws.get_mlflow_tracking_uri()
@@ -203,6 +204,8 @@ import mlflow
 mlflow.log_metric('epoch_loss', loss.item()) 
 ```
 
+#### Automate setting MLflow Tracking
+
 Instead of manually setting the tracking URI in every subsequent experiment notebook session on your clusters, do so automatically using this [Azure Machine Learning Tracking Cluster Init script](https://github.com/Azure/MachineLearningNotebooks/blob/3ce779063b000e0670bdd1acc6bc3a4ee707ec13/how-to-use-azureml/azure-databricks/linking/README.md).
 
 When configured correctly, you are able to see your MLflow tracking data in Azure Machine Learning's REST API and all clients, and in Azure Databricks via the MLflow user interface or by using the MLflow client.