Merge pull request #89346 from MayMSFT/patch-20

Update how-to-access-data.md

Author: PRMerger20

articles/machine-learning/service/how-to-access-data.md

@@ -145,7 +145,8 @@ The `target_path` parameter is the location of the local directory to download t
 <a name="train"></a>
 ## Access your data during training
 
-To access data during training, you can either download or mount your data from your Azure storage services to the compute target via datastores.
+> [!IMPORTANT]
+> Using [Azure Machine Learning datasets (preview)](how-to-create-register-datasets.md) is the new recommended way to access your data in training. Datasets provide functions that load tabular data into pandas or spark DataFrame, and the ability to download or mount files of any format from Azure Blob, Azure File, Azure Data Lake Gen 1, Azure Data Lake Gen 2, Azure SQL, Azure PostgreSQL. Learn more about [how to train with datasets](how-to-train-with-datasets.md).
 
 The following table lists the methods that tell the compute target how to use the datastores during runs.