Prevent error in github action if no benchmarks are run.

[tompollard]

We currently have a github action that runs benchmarks when a format is added or edited.

The action errors out when base formats are edited because the "Run benchmark tests" step is skipped but the "Concatenate results" step is still looking for benchmark_results_*.txt file/s to report. See for example: https://github.com/chorus-ai/chorus_waveform/actions/runs/11615143978/job/32344989477

This pull request should fix the problem, by wrapping the cat benchmark_results_*.txt step in an if clause. If the file doesn't exist, we now report "No benchmarks were run.", instead of raising an error.

[briangow]

Thanks @tompollard , this looks good me!