Venusaur Open - Source Sentence Embedding Model - Freely Assist with Sentence Similarity and Feature Extraction Tasks

Venusaur

Developed by Mihaiii

Venusaur is a sentence embedding model developed based on the Mihaiii/Bulbasaur foundation model, focusing on sentence similarity and feature extraction tasks.

Text Embedding

Safetensors

Open Source License:MIT #Text Similarity Calculation #Multi-task Evaluation #High-precision Classification

Downloads 290

Release Time : 4/29/2024

Model Overview

This model is a sentence-transformer primarily used for calculating similarity between sentences, suitable for scenarios like text matching and information retrieval.

Model Features

Efficient Sentence Embedding

Capable of converting sentences into high-dimensional vector representations for calculating similarity between sentences.

Multi-task Evaluation

Comprehensively evaluated on multiple benchmarks including MTEB.

Lightweight Model

Compared to large language models, this model is more lightweight and suitable for resource-constrained environments.

Model Capabilities

Sentence similarity calculation

Text feature extraction

Text classification

Information retrieval

Text clustering

Use Cases

E-commerce

Product Review Classification

Sentiment classification for Amazon product reviews

Achieved 79.99% accuracy in the MTEB AmazonPolarityClassification test.

Product Similarity Matching

Calculating similarity between different product descriptions

Information Retrieval

Q&A System

Used for answer retrieval in Q&A systems

Achieved NDCG@10 of 34.8 in the MTEB ArguAna test.

🚀 Venusaur

Venusaur is a model based on sentence-transformers for sentence similarity tasks. It uses the base model Mihaiii/Bulbasaur and is trained on the dataset Mihaiii/qa-assistant.

✨ Features

Sentence Similarity: Capable of calculating the similarity between sentences, useful for tasks such as text matching and retrieval.
Feature Extraction: Can extract features from text, facilitating further analysis and processing.
Multiple Task Support: Demonstrates performance in various tasks including classification, retrieval, and clustering.

📦 Model Information

Property	Details
Base Model	Mihaiii/Bulbasaur
Library Name	sentence-transformers
Pipeline Tag	sentence-similarity
Tags	sentence-transformers, feature-extraction, sentence-similarity, gte, mteb
Datasets	Mihaiii/qa-assistant
License	mit

📚 Model Performance

Model Name: Venusaur

1. Classification Tasks

MTEB AmazonCounterfactualClassification (en):
- Accuracy: 73.17910447761194
- AP: 35.29994612283548
- F1: 66.87845205993153
MTEB AmazonPolarityClassification:
- Accuracy: 79.993525
- AP: 74.7042261687233
- F1: 79.9004149386498
MTEB AmazonReviewsClassification (en):
- Accuracy: 39.656000000000006
- F1: 39.287139345446256

2. Retrieval Task (MTEB ArguAna)

MAP@1: 16.643
MAP@10: 28.276
MAP@100: 29.543999999999997
MAP@1000: 29.595
MAP@20: 29.043000000000003
MAP@3: 24.739
MAP@5: 26.592
MRR@1: 17.639
MRR@10: 28.631
MRR@100: 29.891000000000002
MRR@1000: 29.942999999999998
MRR@20: 29.391000000000002
MRR@3: 25.107000000000003
MRR@5: 26.942
NDCG@1: 16.643
NDCG@10: 34.8
NDCG@100: 41.179
NDCG@1000: 42.564
NDCG@20: 37.601
NDCG@3: 27.356
NDCG@5: 30.725
Precision@1: 16.643
Precision@10: 5.576
Precision@100: 0.861
Precision@1000: 0.097
Precision@20: 3.343
Precision@3: 11.641
Precision@5: 8.634
Recall@1: 16.643
Recall@10: 55.761
Recall@100: 86.06
Recall@1000: 97.013
Recall@20: 66.85600000000001
Recall@3: 34.922
Recall@5: 43.172

3. Clustering Task (MTEB ArxivClusteringP2P)

V-Measure: 31.76467048453136
V-Measures: [0.2646936786804572, 0.27790871012280266, 0.29027802989910717, 0.27400555976615254, 0.2823478131745678, 0.25739544436992295, 0.3014171939280134, 0.2862214695233955, 0.2856734533249879, 0.2870107976688266, 0.3709000837926645, 0.3702167780750079, 0.36556393540769305, 0.37650336515785243, 0.3699811227722488, 0.36806220730606526, 0.3696328229784335, 0.3852970338255622, 0.37157613433218695, 0.368267862192135, 0.3715516752706066, 0.26093751350716654, 0.24003989063421033, 0.31112640151573373, 0.2509815194812587, 0.19256512170374224, 0.2638556294764011, 0.08503820346290819, 0.1374194639615466, 1.0, 0.21057893489306592, 0.2646936786804572, 0.27790871012280266, 0.29027802989910717, 0.27400555976615254, 0.2823478131745678, 0.25739544436992295, 0.3014171939280134, 0.2862214695233955, 0.2856734533249879, 0.2870107976688266, 0.3709000837926645, 0.3702167780750079, 0.36556393540769305, 0.37650336515785243, 0.3699811227722488, 0.36806220730606526, 0.3696328229784335, 0.3852970338255622, 0.37157613433218695, 0.368267862192135, 0.3715516752706066, 0.26093751350716654, 0.24003989063421033, 0.31112640151573373, 0.2509815194812587, 0.19256512170374224, 0.2638556294764011, 0.08503820346290819, 0.1374194639615466, 1.0, 0.21057893489306592, 0.2646936786804572, 0.27790871012280266, 0.29027802989910717, 0.27400555976615254, 0.2823478131745678, 0.25739544436992295, 0.3014171939280134, 0.2862214695233955, 0.2856734533249879, 0.2870107976688266, 0.3709000837926645, 0.3702167780750079, 0.36556393540769305, 0.37650336515785243, 0.3699811227722488, 0.36806220730606526, 0.3696328229784335, 0.3852970338255622, 0.37157613433218695, 0.368267862192135, 0.3715516752706066, 0.26093751350716654, 0.24003989063421033, 0.31112640151573373, 0.2509815194812587, 0.19256512170374224, 0.2638556294764011, 0.08503820346290819, 0.1374194639615466, 1.0, 0.21057893489306592, 0.2646936786804572, 0.27790871012280266, 0.29027802989910717, 0.27400555976615254, 0.2823478131745678, 0.25739544436992295, 0.3014171939280134, 0.2862214695233955, 0.2856734533249879, 0.2870107976688266, 0.3709000837926645, 0.3702167780750079, 0.36556393540769305, 0.37650336515785243, 0.3699811227722488, 0.36806220730606526, 0.3696328229784335, 0.3852970338255622, 0.37157613433218695, 0.368267862192135, 0.3715516752706066, 0.26093751350716654, 0.24003989063421033, 0.31112640151573373, 0.2509815194812587, 0.19256512170374224, 0.2638556294764011, 0.08503820346290819, 0.1374194639615466, 1.0, 0.21057893489306592]

📄 License

This project is licensed under the MIT 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