Benchmark inputs¶
prepared-runtime.json is the fixed synthetic workload and acceptance thresholds used by
benchmark_prepared_runtime.py, benchmark_prepared_checkpoint.py and
benchmark_prepared_real_window.py. It is benchmark input, not a performance result.
Keep a profile change explicit when comparing results. Preserve the source/environment, trial counts and limitations with each measurement. See the tool index and verification.
Maintained source: tools/fixtures/README.md