Snowflake Arctic Embed M v2.0 Open-Source Multilingual Sentence Embedding Model - Free Deployment, Supports Similarity Comparison for Over 50 Languages

Snowflake Arctic Embed M V2.0 Cpu

Developed by cnmoro

Snowflake Arctic Embed M v2.0 is a multilingual sentence embedding model focused on sentence similarity tasks, supporting over 50 languages.

Text Embedding

Transformers

Supports Multiple LanguagesOpen Source License:Apache-2.0 #Multilingual sentence embedding #High-dimensional semantic matching #Cross-lingual retrieval

Downloads 502

Release Time : 1/22/2025

Model Overview

This model is a sentence transformer for feature extraction and sentence similarity calculation, performing well on the MTEB benchmark.

Model Features

Multilingual support

Supports sentence embedding calculation for over 50 languages

High performance

Excellent performance on the MTEB benchmark

Sentence similarity

Specifically optimized for calculating similarity between sentences

Model Capabilities

Sentence embedding

Feature extraction

Sentence similarity calculation

Multilingual processing

Use Cases

Text retrieval

Semantic search

Document retrieval based on semantics rather than keywords

Classification tasks

Sentiment analysis

Sentiment classification based on sentence embeddings

Achieved 33.766% accuracy on Amazon review classification task

Question answering systems

Question-answer matching

Calculating similarity between questions and candidate answers

🚀 snowflake-arctic-embed-m-v2.0

This is a model for sentence similarity tasks, supporting a wide range of languages. It has achieved certain performance results on multiple datasets of MTEB.

📚 Documentation

Model Information

Property	Details
Pipeline Tag	sentence-similarity
Tags	sentence-transformers, feature-extraction, sentence-similarity, mteb, arctic, snowflake-arctic-embed, transformers.js
License	apache-2.0
Supported Languages	af, ar, az, be, bg, bn, ca, ceb, cs, cy, da, de, el, en, es, et, eu, fa, fi, fr, gl, gu, he, hi, hr, ht, hu, hy, id, is, it, ja, jv, ka, kk, km, kn, ko, ky, lo, lt, lv, mk, ml, mn, mr, ms, my, ne, nl, pa, pl, pt, qu, ro, ru, si, sk, sl, so, sq, sr, sv, sw, ta, te, th, tl, tr, uk, ur, vi, yo, zh

Model Results

The model snowflake-arctic-embed-m-v2.0 has the following performance results on multiple MTEB datasets:

1. MTEB AmazonCounterfactualClassification (en-ext)

Task Type: Classification
Metrics:
- Accuracy: 66.6867
- F1: 55.0373
- F1 Weighted: 73.07430000000001
- AP: 18.077399999999997
- AP Weighted: 18.077399999999997
- Main Score: 66.6867

2. MTEB AmazonCounterfactualClassification (en)

Task Type: Classification
Metrics:
- Accuracy: 66.194
- F1: 60.854299999999995
- F1 Weighted: 69.57339999999999
- AP: 30.279099999999996
- AP Weighted: 30.279099999999996
- Main Score: 66.194

3. MTEB AmazonPolarityClassification (default)

Task Type: Classification
Metrics:
- Accuracy: 70.3589
- F1: 70.0409
- F1 Weighted: 70.0409
- AP: 64.81949999999999
- AP Weighted: 64.81949999999999
- Main Score: 70.3589

4. MTEB AmazonReviewsClassification (en)

Task Type: Classification
Metrics:
- Accuracy: 33.766
- F1: 33.3656
- F1 Weighted: 33.3656
- Main Score: 33.766

5. MTEB ArguAna (default)

Task Type: Retrieval
Metrics:
- NDCG@1: 33.144
- NDCG@3: 47.909
- NDCG@5: 52.932
- NDCG@10: 58.011
- NDCG@20: 60.168
- NDCG@100: 60.928000000000004
- NDCG@1000: 61.046
- MAP@1: 33.144
- MAP@3: 44.156
- MAP@5: 46.951
- MAP@10: 49.071999999999996
- MAP@20: 49.692
- MAP@100: 49.809
- MAP@1000: 49.815
- Recall@1: 33.144
- Recall@3: 58.819
- Recall@5: 70.982
- Recall@10: 86.558
- Recall@20: 94.879
- Recall@100: 98.791
- Recall@1000: 99.644
- Precision@1: 33.144
- Precision@3: 19.606
- Precision@5: 14.196
- Precision@10: 8.656
- Precision@20: 4.744000000000001
- Precision@100: 0.988
- Precision@1000: 0.1
- MRR@1: 33.4993
- MRR@3: 44.393100000000004
- MRR@5: 47.131299999999996
- MRR@10: 49.264599999999994
- MRR@20: 49.8707
- MRR@100: 49.987700000000004
- MRR@1000: 49.993700000000004
- NAUC NDCG@1 Max: -10.8287
- NAUC NDCG@1 Std: -17.1177
- NAUC NDCG@1 Diff1: 14.4508
- **... (Other metrics are omitted for brevity) **
- Main Score: 58.011

6. MTEB ArxivClusteringP2P (default)

Task Type: Clustering
Metrics:
- V-Measure: 44.684400000000004
- V-Measure Std: 13.5064
- Main Score: 44.684400000000004

7. MTEB ArxivClusteringS2S (default)

Task Type: Clustering
Metrics:
- V-Measure: 35.0503
- V-Measure Std: 13.9543
- Main Score: 35.0503

8. MTEB AskUbuntuDupQuestions (default)

Task Type: Reranking
Metrics:
- MAP: 60.648500000000006
- MRR: 74.528
- nAUC MAP Max: 19.4239
- nAUC MAP Std: 20.0729
- nAUC MAP Diff1: 10.0382
- nAUC MRR Max: 30.693199999999997
- nAUC MRR Std: 27.1279
- nAUC MRR Diff1: 23.0291
- Main Score: 60.648500000000006

9. MTEB BIOSSES (default)

Task Type: STS
Metrics:
- Pearson: 89.5081
- Spearman: 87.0568
- Cosine Pearson: 89.5081
- Cosine Spearman: 87.0568
- Manhattan Pearson: 88.1247
- Manhattan Spearman: 87.2556
- Euclidean Pearson: 88.3266
- Euclidean Spearman: 87.0568
- Main Score: 87.0568

10. MTEB Banking77Classification (default)

Task Type: Classification
Metrics:
- Accuracy: 80.18180000000001
- F1: 79.5538
- F1 Weighted: 79.5538
- Main Score: 80.18180000000001

11. MTEB BiorxivClusteringP2P (default)

Task Type: Clustering
Metrics:
- V-Measure: 36.0126
- V-Measure Std: 0.47019999999999995
- Main Score: 36.0126

12. MTEB BiorxivClusteringS2S (default)

Task Type: Clustering
Metrics:
- V-Measure: 28.6331
- V-Measure Std: 0.8607999999999999
- Main Score: 28.6331

13. MTEB CQADupstackAndroidRetrieval (default)

Task Type: Retrieval
Metrics:
- NDCG@1: 45.207
- NDCG@3: 51.31400000000001
- NDCG@5: 54.093999999999994
- NDCG@10: 56.31
- NDCG@20: 58.378
- NDCG@100: 61.307
- NDCG@1000: 62.724999999999994
- MAP@1: 37.732
- MAP@3: 46.263
- MAP@5: 48.553000000000004
- MAP@10: 49.984
- MAP@20: 50.888999999999996
- MAP@100: 51.568999999999996
- MAP@1000: 51.666999999999994
- Recall@1: 37.732
- Recall@3: 53.736
- Recall@5: 60.95399999999999
- Recall@10: 68.062
- Recall@20: 75.149
- Recall@100: 88.075
- Recall@1000: 96.878
- Precision@1: 45.207
- Precision@3: 24.368000000000002
- Precision@5: 17.854
- Precision@10: 10.558
- Precision@20: 6.23
- Precision@100: 1.614
- Precision@1000: 0.202
- MRR@1: 45.2074
- MRR@3: 52.9804
- MRR@5: 54.718599999999995
- MRR@10: 55.5713
- MRR@20: 55.94
- MRR@100: 56.21699999999999
- MRR@1000: 56.2504
- NAUC NDCG@1 Max: 43.7697
- NAUC NDCG@1 Std: -3.9530000000000003
- NAUC NDCG@1 Diff1: 57.75320000000001
- **... (Other metrics are omitted for brevity) **
- Main Score: (Not shown in the original, can be further calculated according to requirements)

📄 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