search

beir-cellar / beir

A Heterogeneous Benchmark for Information Retrieval. Easy to use, evaluate your models across 15+ diverse IR datasets.
http://beir.ai
Apache License 2.0
1.61k stars 192 forks source link

chg: title optional #25

Closed svakulenk0 closed 3 years ago

svakulenk0 commented 3 years ago

Many passage retrieval datasets dont have titles only text

thakur-nandan commented 3 years ago

Hi @svakulenk0,

This is handled in the BEIR data loader, where I always output the title key in the corpus dictionary with an empty string for passage retrieval datasets without titles. But, I can see the use of the code when someone provides a custom corpus dictionary and forget to mention the titles as keys.

I will merge this soon. Thanks for your help!

Kind Regards, Nandan

svakulenk0 commented 3 years ago

right :) I was just compressing the size of my corpus file