Add a prompt engineering evaluation example notebook

[gilinachum]

Add an example notebook to show how you can compare different prompts using the same model.
In details:
As a user, I want to try and improve a given prompt by measuring different prompts, utilizing different prompting strategies like: zero-shot, few-shot, chain-of-thought, and possibly others. While using the same given model.