Benchmark hyde using evaluation harness

[mrm1001]

Context

In the past, we have written a successful Hyde blog post, but it did not have any performance metrics. We are aiming to create one blog post by the end of the quarter to showcase the evaluation harness used for advanced techniques. This work will lay the ground-work for that blog post.

Task description

Add another architecture to the evaluation repository, where we show how to improve the performance of a naive RAG pipeline using the evaluation harness and a Hyde architecture.