It's a suite that has benchmarks in it.
| Benchmark | Description |
|---|---|
| aurpkg | |
| bio | Bioinformatics. |
| covid-mts | COVID-19 multivariate time series. |
| file-enc | File encoding. |
| log-analysis | Log analysis. |
| max-temp | Maximum temperature. |
| media-conv | Media conversion. |
| nlp | Natural language processing. |
| oneliners | One-liners. |
| riker | |
| sklearn | Machine learning. |
| teraseq | |
| tuft-weather | Tuft weather. |
| uniq-ips | Unique IPs. |
| unix50 | Unix 50. |
| web-index | Web index. |
First, set the shell runtime to benchmark with the $BENCHMARK_SHELL shell variable.
By default, this is set to bash.
You can also pass in flags/options to use with the shell.
For example, to benchmark PaSh with --width 4, run export BENCHMARK_SHELL="$PASH_TOP/pa.sh --width 4".
# Build the container
$ docker build -t bensh .
# Run the container
$ docker run --cap-add NET_ADMIN --cap-add NET_RAW -it bensh
# For development, mount the benchmarks directory
docker run --cap-add NET_ADMIN --cap-add NET_RAW -it -v "$(pwd):/benchmarks" bensh