Dataiku DSSInstall

Overview

This node integrates with the Dataiku DSS API, enabling users to perform a wide range of operations on Dataiku DSS resources. Specifically for the Machine Learning - Saved Model resource and the Import MLflow Version From File or Path operation, it allows importing a new MLflow model version into an existing saved model in a project. This is useful when you have an MLflow model stored as a file or in a managed folder path and want to register it as a new version under a saved model in Dataiku DSS.

Common scenarios include:

• Automating the deployment pipeline by programmatically importing MLflow models.
• Managing multiple versions of machine learning models within Dataiku DSS.
• Integrating external MLflow models into Dataiku projects for further use or evaluation.

Example: Importing an MLflow model version from a zipped archive stored in a managed folder into a saved model version in a specific project.

Properties

Output

The node outputs the response from the Dataiku DSS API call:

• For this operation, the output JSON contains details about the imported MLflow model version, such as confirmation of creation, metadata, or error messages.
• If a binary file is downloaded or returned, it is provided in the binary field with the appropriate filename and content type.
• In case of logs or textual responses, these are returned in the json field under relevant keys.

Dependencies

• Requires a configured connection to a Dataiku DSS instance via an API key credential.
• The node uses the Dataiku DSS REST API endpoint, constructed from the server URL and project key.
• The user must have appropriate permissions in the target Dataiku DSS project to import saved model versions.
• The node expects the MLflow model file either uploaded as binary data or accessible via a managed folder path.

Troubleshooting

• Missing Credentials Error: If the API credentials are not set or invalid, the node throws an error indicating missing Dataiku DSS API credentials.
• Required Parameter Missing: The node validates required parameters like Project Key, Save Model ID, and Version ID. Omitting any of these results in an error specifying the missing parameter.
• File Upload Issues: When importing from a file, ensure the file is correctly attached as binary data; otherwise, the import will fail.
• API Errors: Any errors returned by the Dataiku DSS API are wrapped and reported with descriptive messages. Check the API response for details.
• Network/Connection Issues: Ensure the Dataiku DSS server URL is reachable and the API key has sufficient privileges.

Links and References

• Dataiku DSS API Documentation
• MLflow Model Registry Concepts
• Dataiku DSS Saved Models Documentation

This summary focuses on the "Machine Learning - Saved Model" resource and the "Import MLflow Version From File or Path" operation, describing its purpose, inputs, outputs, dependencies, and common troubleshooting points based on static analysis of the node's source code and provided property definitions.