e5-small Open-source Sentence Transformer Model - Efficiently Handle Sentence Similarity and Text Embedding Tasks

E5 Small

Developed by intfloat

E5-small is a compact sentence transformer model focused on sentence similarity and text embedding tasks, performing well on multiple classification and retrieval tasks.

Text Embedding

Safetensors

EnglishOpen Source License:MIT #Sentence similarity calculation #Multi-task text embedding #English text processing

Downloads 16.96k

Release Time : 12/7/2022

Model Overview

This model is primarily used to generate sentence-level embedding representations, suitable for text similarity calculation, information retrieval, and classification tasks.

Model Features

Multi-task performance

Balanced performance across various NLP tasks including classification, clustering, and retrieval

Efficient embedding

Generates high-quality sentence-level embedding representations

Broad applicability

Supports multiple text processing tasks including similarity calculation and information retrieval

Model Capabilities

Sentence embedding generation

Text similarity calculation

Text classification

Text clustering

Information retrieval

Reranking tasks

Use Cases

E-commerce

Product review sentiment analysis

Analyzing sentiment polarity of Amazon product reviews

Achieved 87.53% accuracy on AmazonPolarity dataset

Counterfactual review detection

Identifying counterfactual reviews on Amazon platform

Achieved 76.22% accuracy on AmazonCounterfactual dataset

Customer service

Bank inquiry classification

Automatic classification of bank customer inquiries

Achieved 81.87% accuracy on Banking77 dataset

Academic research

Academic paper clustering

Topic clustering for arXiv and biorxiv papers

Achieved 44.14 V-measure on arXiv clustering task

🚀 e5-small

This document presents the performance results of the e5-small model on various natural language processing tasks, including classification, retrieval, clustering, reranking, and semantic text similarity (STS).

📚 Documentation

Model Information

Property	Details
Model Type	e5-small
Tags	mteb, Sentence Transformers, sentence-similarity, sentence-transformers

Results on Different Tasks

Classification Tasks

Task	Dataset	Accuracy	AP	F1
Classification	MTEB AmazonCounterfactualClassification (en)	76.22388059701493	40.27466219523129	70.60533006025108
Classification	MTEB AmazonPolarityClassification	87.525775	83.51063993897611	87.49342736805572
Classification	MTEB AmazonReviewsClassification (en)	42.611999999999995	-	42.05088045932892
Classification	MTEB Banking77Classification	81.87337662337661	-	81.76647866926402

Retrieval Tasks

Task	Dataset	MAP@1	MAP@10	MAP@100	MAP@1000	MRR@1	MRR@10	MRR@100	MRR@1000	NDCG@1	NDCG@10	NDCG@100	NDCG@1000	Precision@1	Precision@10	Precision@100	Precision@1000	Recall@1	Recall@10	Recall@100	Recall@1000
Retrieval	MTEB ArguAna	23.826	38.269	39.322	39.344	24.253	38.425	39.478	39.5	23.826	46.693	51.469	52.002	23.826	7.383000000000001	0.9530000000000001	0.099	23.826	73.82600000000001	95.306	99.431
Retrieval	MTEB CQADupstackAndroidRetrieval	32.054	40.699999999999996	41.818	41.959999999999994	38.769999999999996	46.150000000000006	46.865	46.925	38.769999999999996	45.778	50.38	52.922999999999995	38.769999999999996	8.269	1.278	0.178	32.054	54.947	74.79599999999999	91.40899999999999
Retrieval	MTEB CQADupstackEnglishRetrieval	29.035	38.007000000000005	39.125	39.251999999999995	36.497	44.077	44.743	44.79	36.497	42.986000000000004	47.323	49.624	36.497	7.8340000000000005	1.269	0.178	29.035	51.06	69.64099999999999	84.49
Retrieval	MTEB CQADupstackGamingRetrieval	37.239	47.873	48.842999999999996	48.913000000000004	42.508	51.44	52.087	52.129999999999995	42.508	53.31399999999999	57.245000000000005	58.794000000000004	42.508	8.458	1.133	0.132	37.239	65.99000000000001	82.99499999999999	94.128
Retrieval	MTEB CQADupstackGisRetrieval	23.039	29.694	30.587999999999997	30.692999999999998	24.633	31.478	32.299	32.381	24.633	33.697	38.080000000000005	40.812	24.633	5.0729999999999995	0.753	0.10300000000000001	23.039	44.275999999999996	64.4	85.135
Retrieval	MTEB CQADupstackMathematicaRetrieval	13.594999999999999	19.933999999999997	20.966	21.087	17.662	24.407	25.385	25.465	17.662	24.391	29.681	32.923	17.662	4.44	0.8200000000000001	0.125	13.594999999999999	29.933999999999997	40.966	41.087

Clustering Tasks

Task	Dataset	V-Measure
Clustering	MTEB ArxivClusteringP2P	44.13995374767436
Clustering	MTEB ArxivClusteringS2S	37.13950072624313
Clustering	MTEB BiorxivClusteringP2P	35.80600542614507
Clustering	MTEB BiorxivClusteringS2S	31.86321613256603

Reranking Task

Task	Dataset	MAP	MRR
Reranking	MTEB AskUbuntuDupQuestions	59.35843292105327	73.72312359846987

STS Task

Task	Dataset	Cosine Similarity Pearson	Cosine Similarity Spearman	Euclidean Pearson	Euclidean Spearman	Manhattan Pearson	Manhattan Spearman
STS	MTEB BIOSSES	84.55140418324174	84.21637675860022	81.26069614610006	83.25069210421785	80.17441422581014	81.87596198487877

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご