ctf4science.benchmark_module.main

ctf4science.benchmark_module.main()

Run the ModelBenchmarker class from the command line.

This function parses command-line arguments for benchmarking a CTF model using optimal hyperparameters. It instantiates the ModelBenchmarker class, performs the specified number of independent evaluation runs, and prints a summary of the benchmark results.

Parameters:

None

Returns:

None

Raises:

SystemExit

If the benchmark fails due to an exception.

Notes

Command-line Arguments:

–configstr

Path to the model configuration file (required).

–num-evalsint, optional

Number of independent evaluation runs to perform (default is 5).

Side Effects:

Prints benchmark summary to stdout. On failure, prints an error message and exits with code 1.

Examples

Run from command line:

$ python benchmark_module.py –config config.yaml –num-evals 10