Skip to content

Injecting First-stage Retriever score into the input of cross-encoder re-rankers in a knowledge distillation training setup

Notifications You must be signed in to change notification settings

arian-askari/RelevanceCAT

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

66 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RelevanceCAT: Injecting Relevance score of BM25 or Dense Retriever Improves Cross-encoder re-rankers

Quick run notebook

To quickly train a cross-encoder_BM25CAT or cross-encoder_DPRCAT re-ranker in a knowledge distillation setup, you could use the implementation below:

train_cross-encoder_kd_BM25CAT

train_cross-encoder_kd_DPRCAT

About

Injecting First-stage Retriever score into the input of cross-encoder re-rankers in a knowledge distillation training setup

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages