Run benchmarking with the supported models on BEIR MSMARCO

[HAKSOAT]

We need to run benchmarking on the BEIR MSMARCO dataset, to have a better understanding of how the models are performing for retrieval tasks.

We can use the test split available on Hugging Face hub:

QRels
Corpus

Proposed metrics:

• NDCG@10
• Precision@10
• Recall@100

Considering non-judged documents as non-relevant.

[HAKSOAT]

The InformationRetrievalEvaluator from SentenceTransformers can be helpful for this.