cross-encoder-ELECTRA-DistillRankNET

This model is a cross-encoder based on google/electra-base-discriminator. It was trained on Ms-Marco using loss distillRankNET as part of a reproducibility paper for training cross encoders: "Reproducing and Comparing Distillation Techniques for Cross-Encoders", see the paper for more details.

Contents

• Model Description
• Usage
• Evals

Model Description

This model is intended for re-ranking the top results returned by a retrieval system (like BM25, Bi-Encoders or SPLADE).

• Training Data: MS MARCO Passage
• Language: English
• Loss distillRankNET

Training can be easily reproduced using the assiciated repository. The exact training configuration used for this model is also detailed in config.yaml.

Usage

Quick Start:

from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch

tokenizer = AutoTokenizer.from_pretrained("google/electra-base-discriminator")
model = AutoModelForSequenceClassification.from_pretrained("xpmir/cross-encoder-ELECTRA-DistillRankNET")

features = tokenizer("What is experimaestro ?", "Experimaestro is a powerful framework for ML experiments management...", padding=True, truncation=True, return_tensors="pt")

model.eval()
with torch.no_grad():
    scores = model(**features).logits
    print(scores)

Evaluations

We provide evaluations of this cross-encoder re-ranking the top 1000 documents retrieved by naver/splade-v3-distilbert.

dataset	RR@10	nDCG@10
msmarco_dev	37.50	44.08
trec2019	100.00	77.88
trec2020	95.00	74.82
fever	79.89	80.03
arguana	15.87	24.53
climate_fever	22.70	17.38
dbpedia	77.35	47.24
fiqa	46.89	38.68
hotpotqa	86.53	67.52
nfcorpus	55.78	34.33
nq	55.00	60.02
quora	77.07	79.32
scidocs	27.87	15.98
scifact	62.64	65.76
touche	68.69	35.77
trec_covid	87.97	70.20
robust04	70.36	49.20
lotte_writing	70.07	61.35
lotte_recreation	62.44	56.76
lotte_science	47.24	40.02
lotte_technology	55.93	47.04
lotte_lifestyle	74.60	64.90
Mean In Domain	77.50	65.59
BEIR 13	58.79	48.98
LoTTE (OOD)	63.44	53.21