Azure ML Studio AutoML Model Explanations Not Loading

JJ 5 Reputation points
2024-07-10T03:48:22.6766667+00:00

When using the AutoML feature in Azure Machine Learning studio, I am not able to load model explanations where it tells you feature importances, even on old models with relatively small dataset sizes where they were definitely loading for me in the past. All that shows is a loading screen and the data never renders.

I have tried clearing out memory, starting up new workspaces, and running specific explanation child jobs to no avail. Any other suggestions for trouble shooting?

Azure Machine Learning
Azure Machine Learning
An Azure machine learning service for building and deploying models.
2,978 questions
{count} votes

1 answer

Sort by: Most helpful
  1. dupammi 8,535 Reputation points Microsoft Vendor
    2024-07-10T06:30:19.2166667+00:00

    Hi @JJ

    Thank you for your question.

    Please try clearing your browser cache and cookies or try accessing the studio from an incognito window or a different browser. Ensure your Azure subscription and workspace are not hitting any quotas and check the logs for any errors by navigating to the "Jobs" section in the Azure ML Studio and inspecting the logs of the specific AutoML run and explanation child jobs. Try loading model explanations using the Azure ML SDK with the ExplanationClient and ensure your dataset and model sizes are within reasonable limits. Make sure your compute target has sufficient resources to handle the explanations and try re-running the model explanations manually. If you need to view the explanations in the "Models" tab and the model's explanation is not visible, ensure that you complete the remote interpretability steps (uploading generated explanations to Azure Machine Learning Run History) to view the visualizations on the explanation's dashboard in Azure Machine Learning studio.

    Please also try to see the explanations by going into for AutoML job and clicking the explanations.

    I hope this helps. Thank you.

    0 comments No comments

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.