Azure Machine Learning general availability announcements //Build, May 2021

We are making the following features generally available in Azure Machine Learning:   

• Data labelling – image instance segmentation in Azure Machine Learning gives you a central place to create, manage and monitor labelling projects. Use it to coordinate data, labels and team members to efficiently manage labelling tasks. It supports image classification, either multi-label or multi-class, object identification with bounded boxes and image instance segmentation (polygon). 
• MLflow support: Azure Machine Learning supports managing the end-to-end machine learning cycle using open MLflow standards, enabling existing workloads to move from local execution to execution in the intelligent cloud and edge.  

  • Use MLflow tracking and logging APIs 

  • Track experiments and metrics with both services 

  • Track models in the MLflow and Azure Machine Learning model registry 

  • Deploy MLflow tracked models using the azureml-mlflow package to ACI and AKS

Learn more on how to label images.

Read the how-to doc on MLflow.