For datasets that are configured to upload output files to a location in cloud storage, the path now uses the dataset identifier instead of the job identifier.
Users can now download the output files from JSON output datasets.