Datasets:

facebook

/

wiki_dpr

like 33

Follow

AI at Meta 3,429

Hi @BUGLI ,

You have the description in the docstring of the loading script: https://huggingface.co/datasets/wiki_dpr/blob/main/wiki_dpr.py#L53

            with_embeddings (`bool`, defaults to `True`): Load the 768-dimensional embeddings from DPR.
            with_index (`bool`, defaults to `True`): Load the faiss index trained on the embeddings.
            wiki_split (`str`, defaults to `psgs_w100`): name of the splitting method of wiki articles.
            embeddings_name (`str`, defaults to `nq`): "nq" or "multiset", depending on which dataset DPR was trained on.
            index_name (`str`, defaults to `compressed`): "compressed" or "exact", the configuration of the faiss index to use.

Maybe @lhoestq can give more details.

I agree we should better document this in the dataset card.

There are two types of DPR embeddings, based on two different models. The first model is trained on Natural Questions (nq) and the other one on multiple datasets (multiset).

Then there are two FAISS indexes that come with this dataset. The first one is an exact index but it requires lot of RAM). The other one is a compressed (but approximate) index that is much more RAM efficient.

There are two types of DPR embeddings, based on two different models. The first model is trained on Natural Questions (nq) and the other one on multiple datasets (multiset).

Then there are two FAISS indexes that come with this dataset. The first one is an exact index but it requires lot of RAM). The other one is a compressed (but approximate) index that is much more RAM efficient.

Thanks, it's clear!