ML Skill Error - file runs all the way through but no result

Hi, I am trying to use MLSkill for Document Understanding. I have created an extractor in the AICenter that I have deployed and made public. I want to connect to it using the Endpoint. In Studio, when using the using “test skill” I get an “InternalServerError, Prediction failed” error if I enter a string or “Invalid document” when entering a file. If I just try to run the program straight through, I don’t get any errors, but the program appears to stop before finishing the MLSkill actitvity. I created log messages after the activity that it doesn’t run through.

Hi @megalodon09

The “InternalServerError, Prediction failed” error and “Invalid document” error when testing the ML Skill in UiPath Document Understanding could be due to various reasons. Here are some troubleshooting steps you can follow to resolve these issues:

  1. Check Skill Configuration:

    • Ensure that the skill configuration in the AI Center is correctly set up, including the input and output parameters, model version, and model endpoint. Double-check that the model has been trained and published successfully.
  2. Document Format:

    • Confirm that the document you are testing with (whether a string or a file) is in a format that the ML Skill is trained to handle. Different ML models are trained for specific document types, and providing an unsupported document type can result in errors.
  3. Endpoint Configuration:

    • Double-check the endpoint configuration in UiPath Studio. Ensure that you are using the correct endpoint URL, and that there are no extra spaces or characters in the URL.
  4. Authorization:

    • Verify that the ML Skill’s endpoint requires the appropriate authentication or API key. If authentication is needed, ensure that you are providing it correctly in the activity.
  5. Log Details:

    • Check the logs for more detailed error messages. Enable detailed logging for the activity to see if it provides additional information about the issue.
  6. Testing with Sample Data:

    • Try testing the ML Skill with known and well-structured data that it has been trained for. This can help determine if the issue is with the skill itself or the input data.
  7. Studio Activity Configuration:

    • Ensure that the ML Skill activity in UiPath Studio is configured correctly. Check the input and output parameters, and make sure the activity is placed in the correct sequence in your workflow.
  8. Version Compatibility:

    • Verify that the versions of the UiPath Document Understanding package, the AI Center, and the ML model are compatible with each other. Incompatibility between versions can lead to errors.
  9. UiPath Support:

    • If you have followed all the steps above and the issue persists, consider reaching out to UiPath Support for further assistance. They can provide guidance and help troubleshoot more complex issues.

By carefully reviewing and verifying these aspects, you should be able to diagnose and resolve the “InternalServerError, Prediction failed” error and the “Invalid document” error when using the ML Skill in UiPath Document Understanding.