[Feature] Add separated `limit` flag

[DzmitryPihulski]

--limit : Accepts an integer, or a float between 0.0 and 1.0 . If passed, will limit the number of samples to evaluate to the first X samples (if an integer) or first X%. Useful for debugging.

[DzmitryPihulski]

As in lm-evaluation-harness: https://github.com/EleutherAI/lm-evaluation-harness/blob/main/docs/interface.md#:~:text=%2D%2Dlimit,For%20testing%20only.