gte-tiny Open-source Text Embedding Model - Free Deployment Suitable for Various Natural Language Processing Tasks

Home

Gte Tiny

Developed by TaylorAI

GTE Tiny is a small general-purpose text embedding model suitable for various natural language processing tasks.

Text Embedding

Transformers

#Text Classification #Semantic Similarity Calculation #Information Retrieval

Downloads 74.46k

Release Time : 10/5/2023

Model Overview

GTE Tiny is an efficient text embedding model that supports multiple text-related tasks, including classification, clustering, retrieval, and semantic similarity calculation.

Model Features

Multi-task Support

Supports various text-related tasks, including classification, clustering, retrieval, and semantic similarity calculation.

Efficient Performance

Performs well on multiple benchmarks, especially in classification and retrieval tasks.

Lightweight

The model is small in size, making it suitable for resource-constrained environments.

Model Capabilities

Text classification

Text clustering

Text retrieval

Semantic similarity calculation

Text embedding generation

Use Cases

E-commerce

Amazon Review Classification

Sentiment classification of Amazon product reviews.

Accuracy 86.61%, F1 score 86.59%

Amazon Counterfactual Classification

Identifying counterfactual statements in Amazon product reviews.

Accuracy 71.76%, F1 score 65.89%

Finance

Bank Customer Service Classification

Classifying bank customer service requests.

Accuracy 81.73%, F1 score 81.66%

Academic Research

Paper Clustering

Topic clustering of arXiv and biorxiv papers.

v_measure score 36.01-46.64

Q&A Systems

Technical Q&A Retrieval

Retrieving related questions in technical Q&A communities.

MAP@10 score 36.39-40.47

🚀 gte_tiny

This document presents the performance metrics of the gte_tiny model across various natural language processing tasks and datasets.

📚 Documentation

Model Information

Property	Details
Model Name	gte_tiny

Task and Dataset Metrics

Classification Tasks

Dataset	Accuracy	AP	F1
MTEB AmazonCounterfactualClassification (en)	71.76119402985076	34.63659287952359	65.88939512571113
MTEB AmazonPolarityClassification	86.61324999999998	81.7476302802319	86.5863470912001
MTEB AmazonReviewsClassification (en)	42.61000000000001	-	42.2217180000715
MTEB Banking77Classification	81.73051948051948	-	81.66368364988331

Retrieval Tasks

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
MTEB ArguAna	28.377999999999997	44.565	45.48	45.487	29.445	44.956	45.877	45.884	28.377999999999997	53.638	57.354000000000006	57.513000000000005	28.377999999999997	8.272	0.984	0.1	28.377999999999997	82.717	98.43499999999999	99.644
MTEB CQADupstackAndroidRetrieval	29.160000000000004	40.474	41.905	42.041000000000004	36.91	46.495999999999995	47.288000000000004	47.339999999999996	36.91	46.722	51.969	54.232	36.91	9.013	1.455	0.193	29.160000000000004	58.521	80.323	95.13000000000001
MTEB CQADupstackEnglishRetrieval	27.750000000000004	36.39	37.5	37.625	34.14	41.841	42.469	42.521	34.14	41.409	45.668	47.916	34.14	7.739	1.2630000000000001	0.173	27.750000000000004	49.882	68.556	83.186
MTEB CQADupstackGamingRetrieval	36.879	48.878	49.918	49.978	42.696	52.342	53.044000000000004	53.077	42.696	54.469	58.664	59.951	42.696	8.734	1.1769999999999998	0.133	36.879	67.669	85.822	95.092
MTEB CQADupstackGisRetrieval	22.942	31.741999999999997	32.721000000000004	32.809	24.746000000000002	33.517	34.451	34.522000000000006	24.746000000000002	36.553000000000004	41.53	43.811	24.746000000000002	5.684	0.859	0.109	22.942	49.58	72.614	89.89200000000001
MTEB CQADupstackMathematicaRetrieval	15.345	22.428	23.756	23.872	19.279	27.1	28.211000000000002	28.279	19.279	27.36	33.499	36.452	19.279	5.149	0.938	0.133	15.345	37.974999999999994	64.472	-

Clustering Tasks

Dataset	V-Measure
MTEB ArxivClusteringP2P	46.637318326729876
MTEB ArxivClusteringS2S	36.01134479855804
MTEB BiorxivClusteringP2P	39.18623707448217
MTEB BiorxivClusteringS2S	32.12697757150375

Reranking Task

Dataset	MAP	MRR
MTEB AskUbuntuDupQuestions	59.82917555338909	74.7888361254012

STS Task

Dataset	Cosine Similarity Pearson	Cosine Similarity Spearman	Euclidean Pearson	Euclidean Spearman	Manhattan Pearson	Manhattan Spearman
MTEB BIOSSES	87.1657730995964	86.62787748941281	85.48127914481798	86.48148861167424	85.07496934780823	86.39473964708843

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