Azure Machine Learning—Public preview updates for November 2022
Published date: 09 November, 2022
New features now available in Public Preview include the ability to read Delta Lakes in fewer steps, read datastore URIs, debug and monitor your training jobs, and perform a wide variety of data wrangling.
- Enable Delta Support for tabular dataset : You can now read a Delta Lake straight from an Azure Machine Learning MLTable without Spark clusters.
- Read AzureML datastore URIs in pandas/dask (via fsspec integration): You can now use Pandas, Dask, DVC, and other Python libraries (which accept file objects) to read Azure Machine Learning datastore URIs pointing to registered datastores or data assets.
- Debug & monitor your training jobs: You can now quickly reserve your required compute resources, access job containers, monitor your training job, or debug your job remotely – all through Azure Machine Learning’s interactive training experience.
- Configure AzureML Spark to perform data wrangling: You can now perform any type of data wrangling within the Azure Machine Learning ecosystem before training your machine learning model. You can run both Attached and Managed Spark experiences, and this functionality offers the ultimate flexibility in performing your data wrangling.