A few small things that would be great to improve in benchmarking CLI:
- We should use
structop instead of clap as thats what we use for all other CLIs (e.g., here).
- We should keep CLI parameters as consistent as possible across different CLIs (e.g., for how we specify security, number of outputs etc.). See example here.
- Let's change the executable file name to something like
miden-bench (miden-benchmarking-cli is way too long).
miden::prove() also includes execution time, I believe. So, to make things more clear, we should probably label it as "Execution + proving time" or something like that.
- Would also be great to print out proof size.
A few small things that would be great to improve in benchmarking CLI:
structopinstead ofclapas thats what we use for all other CLIs (e.g., here).miden-bench(miden-benchmarking-cliis way too long).miden::prove()also includes execution time, I believe. So, to make things more clear, we should probably label it as "Execution + proving time" or something like that.