About

General

Explore and validate if all important use cases and scenarios outlined at MLflow Tracking are covered, and work successfully. When not, add corresponding information to the "Caveats" section of the documentation.

Backlog

Those are specific items we would like to explore, or need to be addressed.

• Fix or evaluate impact of failing test cases #11 enumerates a few failing test cases. Their impact needs to be assessed.
• Directly use an SQLAlchemy database URI as MLFLOW_TRACKING_URI, without using a Tracking Server. See Where Runs are recorded.
• Evaluate Automatic Logging on behalf of a few examples.
• Evaluate the databricks and/or databricks-uc (Databricks Unity) backends / URI schemes.
• Identify other subsystems of MLflow, which could be applicable to be supported by CrateDB.