Rank-DistiLLM: Closing the Effectiveness Gap Between Cross-Encoders and LLMs for Passage Re-Ranking
Paper • 2405.07920 • Published • 3
How to use webis/monoelectra-large with Lightning IR:
#install from https://github.com/webis-de/lightning-ir
from lightning_ir import CrossEncoderModule
model = CrossEncoderModule("webis/monoelectra-large")
model.score("query", ["doc1", "doc2", "doc3"])This repository contains the model described in the paper Rank-DistiLLM: Closing the Effectiveness Gap Between Cross-Encoders and LLMs for Passage Re-Ranking.
The code for training and evaluation can be found at https://github.com/webis-de/msmarco-llm-distillation.
Base model
google/electra-large-discriminator