Note that these aren't necessarily representative of the languages used, but rather what I could write in them without spending too long thinking about it.
To run all defined benchmarks, just run python run.py
, which will dump the plot into plots/
, and the results into data.dat
and memories
.
To skip some specific benchmarks, define the env vars BENCH_OVERRIDE=SKIP
and BENCH_<language>=TRUE
(or BENCH_<mode>=TRUE
to skip a specific mode)
And to run only some specific benchmarks, set BENCH_OVERRIDE=RUN
instead of SKIP
.
To run N benchmarks in parallel, set BENCH_JOB=N
, or to anything that's not a number to make it use all CPUs (as defined by os.cpu_count()
).
Fix it yourself (and throw me a PR)