Geek Answers Handbook

Find an Azure ML experiment with deployed web service

I have a list of ML experiments that I created in Azure Machine Learning Studio. I deployed them as web services (new version, not classic).

How can I go into Azure Machine Learning Web Services, click on a web service (which was deployed from an experiment), and then return to the experimental / predictive model that feeds it?

The only link I can find between the two is to update the web service from a predictive experiment, which then confirms what the web service is. I see that "ExperimentId" is the GUID in the url when it is in the experiment and the web service, so hopefully this is possible.

My reasoning is that relying on appropriate naming conventions, etc., a human error depends on which model to update.

+7
web-services azure-machine-learning azure
source share
1 answer

The new web service does not store any information about the experimental or workspace that has been deployed (not all new web services are deployed from the experiment).

Here are the options available to track the relationship between the experiment and the new web service.

last deployment

However, the experiment tracks the new last web service that was deployed from the experiment. each deployment to a new web service overwrites this value.

The value is saved on the experiment graph. One way to get a graph is to use the powershell amlps module

Export-AmlExperimentGraph -ExperimentId <Experiment Id> -OutputFile e.json

e.json

 { "ExperimentId":"<Experiment Id>", // . . . "WebService":{ // . . . "ArmWebServiceId":"<Arm Id>" }, // . . . }

azure resource tags

The tag feature for Azure resources is supported by new web services. Installing tag software on a web service programmatically, using powershell, or through the azure UX portal can be used to store links to an experiment in a new web service.

+1
source share

More articles:


All Articles