cross-encoder-ettin-68m-ADR-MSE

This model is a cross-encoder based on jhu-clsp/ettin-encoder-68m. It was trained on Ms-Marco using loss ADR 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 ADR

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("xpmir/cross-encoder-ettin-68m-ADR-MSE")
model = AutoModelForSequenceClassification.from_pretrained("xpmir/cross-encoder-ettin-68m-ADR-MSE")

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	34.13	40.57
trec2019	95.54	74.00
trec2020	93.98	68.79
fever	76.77	77.02
arguana	12.56	18.62
climate_fever	16.73	12.38
dbpedia	70.93	41.64
fiqa	42.21	34.62
hotpotqa	82.32	63.86
nfcorpus	53.16	32.11
nq	48.06	53.34
quora	76.63	78.38
scidocs	23.29	12.88
scifact	60.29	61.83
touche	62.63	33.60
trec_covid	89.06	74.14
robust04	66.57	42.06
lotte_writing	69.14	60.07
lotte_recreation	57.79	52.50
lotte_science	45.91	38.86
lotte_technology	52.18	43.12
lotte_lifestyle	68.75	59.76
Mean In Domain	74.55	61.12
BEIR 13	54.97	45.72
LoTTE (OOD)	60.06	49.40