All the judge outputs (including rubric) and model outputs are in the samples reports.

Sorry you don't like the displayed metrics. I find them very useful / revealing of the things I'm trying to measure with this benchmark.