question

XinfenHan-0907 avatar image
1 Vote"
XinfenHan-0907 asked lauragutierrrez commented

azure cosmos db as a datastore in ml

Hi, I'm wondering if I can register azure cosmos db as a datastore in azure machine learning?
From your documentation, it seems not https://docs.microsoft.com/en-us/python/api/azureml-core/azureml.core.datastore%28class%29?view=azure-ml-py

  • Do you have a plan to implement the feature in near future?

  • Any recommended alternative solutions for now?

Thanks.



azure-machine-learning
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

GiftA-MSFT avatar image
0 Votes"
GiftA-MSFT answered lauragutierrrez commented

Hi, thanks for reaching out. Currently, Cosmos DB isn't a supported datasource when using Azure ML datastores. However, the product team are aware of this request and will provide updates accordingly. An alternative for now will be to use Azure ML Studio (Classic) which supports Cosmos DB as data source. You can also try a heuristic approach via Execute Python Script module in Designer to import data using python. Hope this helps.


· 1
5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

Any updates about this integration?

0 Votes 0 ·
TopAnkoYIANDextadmin-7282 avatar image
2 Votes"
TopAnkoYIANDextadmin-7282 answered

Hi, Is there an update on this as a year has passed? For us, Cosmos DB is an important datasource (with no workable workarounds ), and not being able to use it as a datastore in Azure ML is forcing us to make a choice to continue with Azure ML or not.

5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.

danielhiggs-2186 avatar image
0 Votes"
danielhiggs-2186 answered

I'm also interested in an update. We're in the process of migrating to azure and have lots of metadata associated with our training set stored in json files. It would be very nice to store it as a Table under our storage account and connect to it from AzureML via a datastore.

5 |1600 characters needed characters left characters exceeded

Up to 10 attachments (including images) can be used with a maximum of 3.0 MiB each and 30.0 MiB total.