We could provide more comprehensive results, e.g.:
- Performance profile plots
- Per-workload runtimes
- Per-workload performance plots (metric vs. runtime)
- ...
Ideally, this would automatically be created once new logs are available.
Related to # and also the main benchmark codebase, since the scoring/plotting code is part of this codebase.
We could provide more comprehensive results, e.g.:
Ideally, this would automatically be created once new logs are available.
Related to # and also the main benchmark codebase, since the scoring/plotting code is part of this codebase.