cross-encoder-ettin-17m-infoNCE

This model is a cross-encoder based on jhu-clsp/ettin-encoder-17m. It was trained on Ms-Marco using loss infoNCE 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 infoNCE

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-17m-infoNCE")
model = AutoModelForSequenceClassification.from_pretrained("xpmir/cross-encoder-ettin-17m-infoNCE")

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	33.68	39.60
trec2019	86.89	63.96
trec2020	93.83	66.26
fever	68.14	69.24
arguana	15.19	22.81
climate_fever	23.51	17.06
dbpedia	63.08	35.27
fiqa	36.53	28.82
hotpotqa	79.28	60.76
nfcorpus	48.40	28.25
nq	41.62	46.09
quora	78.60	79.68
scidocs	23.12	12.80
scifact	63.43	65.78
touche	70.24	35.01
trec_covid	89.45	65.81
robust04	58.72	34.81
lotte_writing	59.45	49.31
lotte_recreation	51.77	46.72
lotte_science	41.26	34.41
lotte_technology	43.04	35.38
lotte_lifestyle	65.88	56.38
Mean In Domain	71.47	56.61
BEIR 13	53.89	43.64
LoTTE (OOD)	53.35	42.83