Jina Embeddings V2 Base De

J

Jina Embeddings V2 Base De

Developed by jinaai

Jina Embeddings V2 Base German is a transformer-based sentence embedding model focused on feature extraction and sentence similarity calculation for German and English texts.

Supports Multiple LanguagesOpen Source License:Apache-2.0 #German-English bilingual sentence embeddings #High-precision text similarity calculation #MTEB benchmark optimization

Downloads 124.30k

Release Time : 1/12/2024

Model Overview

This model is primarily used to generate high-quality sentence embeddings, supporting German and English texts, and is suitable for tasks such as sentence similarity calculation, text classification, and information retrieval.

Model Features

Multilingual support

Supports embedding generation for German and English texts, suitable for cross-language application scenarios.

High-performance feature extraction

Performs excellently in the MTEB benchmark, especially in classification and retrieval tasks for German and English.

Flexible applicability

Suitable for various natural language processing tasks, including text classification, clustering, and information retrieval.

Model Capabilities

Sentence embedding generation

Text classification

Sentence similarity calculation

Information retrieval

Text clustering

Use Cases

E-commerce

Product review classification

Classify German and English product reviews to identify positive and negative feedback.

Achieved 77.52% accuracy in the MTEB AmazonPolarityClassification test.

Information retrieval

Cross-language document retrieval

Retrieve relevant content from German and English documents.

Achieved 98.98% accuracy in the MTEB BUCC (de-en) test.

🚀 jina-embeddings-v2-base-de

This model jina-embeddings-v2-base-de is designed for various natural language processing tasks, including text classification, retrieval, clustering, and more. It supports both German and English languages and has been evaluated on multiple datasets with detailed performance metrics provided.

📚 Documentation

Tags

sentence-transformers
feature-extraction
sentence-similarity
mteb
transformers
transformers.js

Supported Languages

de
en

Inference

Inference is disabled for this model.

License

This model is licensed under the apache-2.0 license.

Model Index

The model named jina-embeddings-v2-base-de has the following evaluation results on different tasks and datasets:

Classification Tasks

Dataset	Language	Accuracy	AP	F1
MTEB AmazonCounterfactualClassification	en	73.76119402985076	35.99577188521176	67.50397431543269
MTEB AmazonCounterfactualClassification	de	68.9186295503212	79.73307115840507	66.66245744831339
MTEB AmazonPolarityClassification	-	77.52215	71.85051037177416	77.4171096157774
MTEB AmazonReviewsClassification	en	38.498	-	38.058193386555956
MTEB AmazonReviewsClassification	de	37.717999999999996	-	37.22674371574757
MTEB Banking77Classification	-	83.93506493506493	-	83.91014949949302

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	25.319999999999997	40.351	41.435	41.443000000000005	25.746999999999996	40.515	41.606	41.614000000000004	25.319999999999997	49.332	53.909	54.089	25.319999999999997	7.831	0.9820000000000001	0.1	25.319999999999997	78.307	98.222	99.57300000000001
MTEB CQADupstackAndroidRetrieval	30.830999999999996	41.355	42.791000000000004	42.918	38.484	47.593	48.388	48.439	38.484	47.27	52.568000000000005	54.729000000000006	38.484	8.927	1.425	0.19	30.830999999999996	57.87799999999999	80.124	94.208
MTEB CQADupstackEnglishRetrieval	25.782	34.492	35.521	35.638	32.357	39.965	40.644000000000005	40.695	32.357	39.644	43.851	46.211999999999996	32.357	7.344	1.201	0.168	25.782	49.132999999999996	67.24	83.045
MTEB CQADupstackGamingRetrieval	35.778999999999996	47.038000000000004	48.064	48.128	41.254000000000005	50.556999999999995	51.296	51.331	41.254000000000005	52.454	56.776	58.181000000000004	41.254000000000005	8.464	1.157	0.133	35.778999999999996	64.85300000000001	83.98400000000001	94.18299999999999
MTEB CQADupstackGisRetrieval	21.719	29.326999999999998	30.314000000000004	30.397000000000002	23.503	31.225	32.096000000000004	32.159	23.503	33.842	39.038000000000004	41.214	23.503	5.266	-	-	21.719	-	-	-

Clustering Tasks

Dataset	V - Measure
MTEB ArxivClusteringP2P	41.43100588255654
MTEB ArxivClusteringS2S	32.08988904593667
MTEB BiorxivClusteringP2P	34.970675877585144
MTEB BiorxivClusteringS2S	28.779230269190954
MTEB BlurbsClusteringP2P	35.490175601567216
MTEB BlurbsClusteringS2S	16.16638280560168

Reranking Task

Dataset	MAP	MRR
MTEB AskUbuntuDupQuestions	60.55514765595906	73.51393835465858

STS Task

Dataset	Cosine Similarity Pearson	Cosine Similarity Spearman	Euclidean Pearson	Euclidean Spearman	Manhattan Pearson	Manhattan Spearman
MTEB BIOSSES	79.6723823121172	76.90596922214986	77.87910737957918	76.66319260598262	77.37039493457965	76.09872191280964

Bitext Mining Task

Dataset	Accuracy	F1	Precision	Recall
MTEB BUCC (de - en)	98.97703549060543	98.86569241475296	98.81002087682673	98.97703549060543

Featured Recommended AI Models

AIbase

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

© 2025AIbase