Add Embeddings base class, SentenceTransformerEmbeddings class, EmbeddingGeneration and FaissNearestNeighbour steps

[gabrielmbmb]

Description

This PR adds a new Embeddings base class which purpose is to define an interface for classes that generates sentence embeddings. In addition, it adds the first implementation of this class, the SentenceTransformerEmbeddings class which uses sentence-transformers package.

In addition, it adds EmbeddingGeneration step which uses an Embeddings model to generate sentence embeddings, and FaissNearestNeighbour step which uses faiss to create an index used to search the top k nearest neighbour and the scores for each input row.

[github-actions]

Documentation for this PR has been built. You can view it at: https://distilabel.argilla.io/pr-830/

[codspeed-hq]

CodSpeed Performance Report

Merging #830 will not alter performance

_{Comparing embeddings (8527538) with develop (6ac0267)}

Summary

✅ 1 untouched benchmarks