question

SamMcKinleyHinson-6915 avatar image
0 Votes"
SamMcKinleyHinson-6915 asked ramr-msft answered

Azure Automated ML Model Deployment to Online Endpoint Stuck in Transitioning

I have a model produced by Azure's Automated ML service. I am following this tutorial (https://docs.microsoft.com/en-us/azure/machine-learning/how-to-deploy-automl-endpoint?tabs=Studio) to deploy to an online endpoint via the portal. Currently, the endpoint is successfully created but the actual deployment fails. It hangs for around 2.5 hours before crashing on a timeout error. The deployment logs are empty.

Other posts suggest that this might be due to some kind of issue in the configuration file or specification of dependencies. However, those files are being generated and provided by azure. I am at a loss for how to proceed. Thought about trying to deploy from the CLI but the ml extension fails to install citing version conflicts in Docker. Any advice on what might be going wrong or ways to investigate this would be greatly appreciated!

azure-machine-learning
· 2
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.

@SamMcKinleyHinson-6915 Thanks for the question. Can you please confirm on the CLI version that you have installed.

Please share details of your experiment and issue from the ml.azure.com portal for a service engineer to lookup the issue from the back-end? This option is available from the top right hand corner of the portal by clicking the smiley face, Please select the option Microsoft can email you about the feedback along with a screen shot so our service team can lookup and advise through email.

0 Votes 0 ·

@ramr-msft thank you for the reply. My CLI version is 2.35. When I run "az extension install -n ml" I get the following output:

193917-image.png


0 Votes 0 ·
image.png (102.5 KiB)

1 Answer

ramr-msft avatar image
0 Votes"
ramr-msft answered

@SamMcKinleyHinson-6915 Thanks, Can you try this notebook for deployment and if that works for you (it should), compare with your code?
https://github.com/CESARDELATORRE/Easy-AutoML-MLOps/blob/master/notebooks/5-automl-model-service-deployment-and-inference/automl-model-service-deployment-and-inference-safe-driver-classifier.ipynb

You can also use the notebook with a simple AutoML remote run, but you might need to change the name of the model when registering it in the Workspace since it’s a different name to what the deployment notebook is using:
https://github.com/CESARDELATORRE/Easy-AutoML-MLOps/blob/master/notebooks/3-automl-remote-compute-run/automl-remote-compute-run-safe-driver-classifier.ipynb

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.