- You could download (using the API) the run_results.json artifact that is generated after each run.
- The discovery API I think can tell you the status of models for the run.
Note: @Jeremy Yeo
originally posted this reply in Slack. It might not have transferred perfectly.