Snowflake Arctic M Long Open-Source Sentence Embedding Model - Free for Sentence Similarity and Feature Extraction

Snowflake Arctic Embed M Long

Developed by Snowflake

Snowflake Arctic M Long is a sentence-transformers-based sentence embedding model focused on sentence similarity and feature extraction tasks.

Text Embedding

Transformers

Open Source License:Apache-2.0 #Sentence similarity calculation #Multi-task text embedding #High-precision semantic matching

Downloads 23.79k

Release Time : 4/12/2024

Model Overview

This model is primarily used for sentence similarity calculation and feature extraction, supporting various natural language processing tasks such as classification, clustering, retrieval, and reranking.

Model Features

High-performance sentence embedding

Excels in various tasks (e.g., classification, retrieval, clustering) by providing high-quality sentence embedding representations.

Multi-task support

Supports multiple natural language processing tasks including sentence similarity, classification, clustering, and retrieval.

Comprehensive dataset evaluation

Thoroughly evaluated on multiple public datasets (e.g., MTEB) with outstanding performance.

Model Capabilities

Sentence similarity calculation

Feature extraction

Text classification

Text clustering

Information retrieval

Reranking

Use Cases

E-commerce

Product review classification

Performs sentiment classification on Amazon product reviews to distinguish between positive and negative feedback.

Achieves 78.74% accuracy on the MTEB AmazonPolarityClassification dataset.

Q&A systems

Question-answer retrieval

Retrieves similar questions in Q&A communities like AskUbuntu.

Achieves 61.36% MAP on the MTEB AskUbuntuDupQuestions dataset.

Academic research

Paper clustering

Performs topic clustering on academic papers from arXiv and bioRxiv.

Achieves 45.56% V-measure on the MTEB ArxivClusteringP2P dataset.

🚀 snowflake-arctic-m-long

This model is designed for sentence - similarity tasks. It has been tested on multiple datasets from MTEB, showing its performance in various natural language processing tasks such as classification, retrieval, clustering, reranking, and STS.

📚 Documentation

Model Information

Property	Details
Model Type	snowflake - arctic - m - long
Pipeline Tag	sentence - similarity
Tags	sentence - transformers, feature - extraction, sentence - similarity, mteb, arctic, snowflake - arctic - embed, transformers.js

Results

The model has been evaluated on multiple tasks and datasets. Here are some of the key results:

Classification Tasks

Dataset	Accuracy	AP	F1
MTEB AmazonCounterfactualClassification (en)	78.4776119402985	42.34374238166049	72.51164234732224
MTEB AmazonPolarityClassification	78.7416	73.12074819362377	78.64057339708795
MTEB AmazonReviewsClassification (en)	39.926	-	39.35531993117573
MTEB Banking77Classification	80.26298701298703	-	79.68356764080303

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	34.851	51.473	52.103	52.105000000000004	35.491	51.73799999999999	52.37500000000001	52.378	34.851	60.364	62.888999999999996	62.946000000000005	34.851	8.855	0.992	0.1	34.851	88.549	99.21799999999999	99.644
MTEB CQADupstackAndroidRetrieval	35.612	47.117	48.711	48.826	42.918	52.806	53.564	53.596999999999994	42.918	53.291999999999994	58.711999999999996	60.317	42.918	9.927999999999999	1.592	0.201	35.612	64.671	86.97	96.99600000000001
MTEB CQADupstackEnglishRetrieval	33.742	44.49	45.781	45.902	42.357	50.463	51.17	51.205999999999996	42.357	50.175000000000004	54.491	56.282	42.357	9.382	1.473	0.191	33.742	59.760999999999996	77.89500000000001	89.005
MTEB CQADupstackGamingRetrieval	43.883	56.464999999999996	57.394	57.443999999999996	50.470000000000006	59.997	60.586	60.61	50.470000000000006	62.134	65.69500000000001	66.674	50.470000000000006	9.812	1.25	0.13699999999999998	43.883	75.417	90.545	97.44500000000001
MTEB CQADupstackGisRetrieval	29.813000000000002	38.627	39.735	39.806000000000004	32.316	40.752	41.699000000000005	41.749	32.316	43.524	48.648	50.405	32.316	6.451999999999999	0.9490000000000001	0.11299999999999999	29.813000000000002	56.562999999999995	79.452	92.715
MTEB CQADupstackMathematicaRetrieval	19.961000000000002	28.026	29.212	29.332	24.627	33.045	33.944	34.013	24.627	-	-	-	24.627	-	-	-	19.961000000000002	-	-	-

Clustering Tasks

Dataset	V - Measure
MTEB ArxivClusteringP2P	45.5554998405317
MTEB ArxivClusteringS2S	35.614248811397005
MTEB BigPatentClustering	20.923883720813706
MTEB BiorxivClusteringP2P	36.16058801465044
MTEB BiorxivClusteringS2S	30.1402356118627

Reranking Task

Dataset	MAP	MRR
MTEB AskUbuntuDupQuestions	61.355489424753884	75.49443784900849

STS Task

Dataset	Cosine Similarity Pearson	Cosine Similarity Spearman	Euclidean Pearson	Euclidean Spearman	Manhattan Pearson	Manhattan Spearman
MTEB BIOSSES	89.17311056578292	88.24237210809322	87.3188065853646	88.24237210809322	86.89499710049658	87.85441146091777

📄 License

This project is licensed under the Apache - 2.0 license.

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