They appear to be mostly covering separate topics, but they don’t reference each other and there is no indication of what the intended function is of each. I’m not sure they are entirely consistent with each other, either.
Comment from @petermattson:
Intent is:
Training rules is specifically about what to implement for MLPerf training.
Submissions rules are generic to all MLPerf benchmarks for things like code availability and review process.
Cross references in intro with description at a minimum would be good and easy
Currently there are two separate rules docs for MLPerf : https://github.com/mlperf/training_policies/blob/master/training_rules.adoc https://github.com/mlperf/policies/blob/master/submission_rules.adoc
They appear to be mostly covering separate topics, but they don’t reference each other and there is no indication of what the intended function is of each. I’m not sure they are entirely consistent with each other, either.
Comment from @petermattson: Intent is: Training rules is specifically about what to implement for MLPerf training. Submissions rules are generic to all MLPerf benchmarks for things like code availability and review process.
Cross references in intro with description at a minimum would be good and easy