Runtime Performance Benchmarks for Dynamical Systems

Supported experiments

On Ubuntu, the following commands will build code and save result data to a user supplied directory, under relatively controlled conditions:

$ bazel run //systems/benchmarking:framework_experiment -- --output_dir=trial1

$ bazel run //systems/benchmarking:multilayer_perceptron_experiment -- --output_dir=trial2

Additional information

Documentation for command line arguments is here: https://github.com/google/benchmark#command-line