This PR adds a number of benchmarks and also changes the benchmarks to be runnable through a single program that outputs the results in JSON format for current-bench.
I feel that there is still work to do. In particular I'd like to have the benchmarks work adaptively such that the benchmarks automatically determine the number of iterations per round and number of rounds to perform based on desired time to finish the benchmarks and perhaps other criteria. However, I also want to merge this soon so that we can see a sequence of results and be able to see that the graphs are readable enough.
This PR adds a number of benchmarks and also changes the benchmarks to be runnable through a single program that outputs the results in JSON format for current-bench.
I feel that there is still work to do. In particular I'd like to have the benchmarks work adaptively such that the benchmarks automatically determine the number of iterations per round and number of rounds to perform based on desired time to finish the benchmarks and perhaps other criteria. However, I also want to merge this soon so that we can see a sequence of results and be able to see that the graphs are readable enough.