Dataiku DSS

Overview

This node integrates with the Dataiku DSS API to perform various operations related to machine learning saved models, specifically including retrieving scoring PMML for a particular version of a saved model. It is useful in scenarios where users want to programmatically access or manage saved machine learning models within Dataiku DSS projects, such as fetching scoring artifacts for deployment or evaluation.

For the Machine Learning - Saved Model resource and the Get Version Scoring PMML operation, the node fetches the PMML (Predictive Model Markup Language) scoring file associated with a specific version of a saved model. This can be used to deploy the model in environments that support PMML scoring or to inspect the model's scoring logic.

Practical Examples

• Automatically retrieve the PMML scoring file of a saved model version to integrate it into an external scoring system.
• Use the node in a workflow to validate or audit the scoring logic of different versions of saved models.
• Automate the process of downloading scoring artifacts for continuous integration/deployment pipelines.

Properties

Output

The output contains the binary data of the PMML scoring file for the specified saved model version. This binary data represents the PMML XML content used for scoring by compatible systems.

• The output JSON will include a binary field with a data property holding the PMML file content prepared for downstream use.
• The filename for the binary data is typically "model_scoring.pmml" indicating the PMML format.

Example output structure:

{
  "binary": {
    "data": "<Buffer ...>"  // Binary content of the PMML file
  }
}

Dependencies

• Requires valid credentials for the Dataiku DSS API, including the server URL and an API authentication token.
• The node uses HTTP requests to communicate with the Dataiku DSS REST API.
• No additional external services are required beyond the configured Dataiku DSS instance.

Troubleshooting

• Missing Credentials Error: If the node throws an error about missing API credentials, ensure that the Dataiku DSS API credentials are properly configured in n8n.
• Required Parameters Missing: Errors like "Project Key is required", "Save Model ID is required", or "Version ID is required" indicate that these mandatory input properties were not provided. Verify that all required fields are set.
• HTTP Request Failures: Network issues or incorrect server URLs can cause request failures. Confirm connectivity to the Dataiku DSS server and correctness of the API endpoint.
• Permission Issues: Insufficient permissions on the Dataiku DSS project or saved model may result in authorization errors. Ensure the API key has appropriate access rights.
• Unexpected Response Format: If the response cannot be parsed, check if the API endpoint is correct and the saved model version exists.

Links and References

• Dataiku DSS API Documentation — Official documentation for Dataiku DSS REST API.
• PMML Specification — Details on the Predictive Model Markup Language standard.