Do you know of a search dataset with very large document length differences? MSMarco for example is pretty consistent in length.
Do you know of a search dataset with very large document length differences? MSMarco for example is pretty consistent in length.