Add a --verbose mode to eval

[PaliC]

Currently we need to do some analysis on a test by test case like atol, rtol, timing etc. To make this easier as this won't be the first time we need to do this + it's likely useful for developers we should create a verbose mode which reports per test the following averages, standard deviations, mins, and maxes. (can add other things if you think its useful)

1. absolute difference from ground truth
2. relative difference from ground truth
3. Speedup relative to eager
4. Wall clock time
5. Errors per test

[PaliC]

Closed with #92