KoSimCSE-bert Open-source Korean Sentence Embedding Model - Precisely Calculate Semantic Similarity of Sentences

Home

Kosimcse Bert

Developed by BM-K

Korean sentence embedding model optimized based on BERT architecture for calculating sentence semantic similarity

Text Embedding

Transformers

Korean#Korean Semantic Similarity #Multi-task Learning #High-precision Embedding

Downloads 444

Release Time : 5/23/2022

Model Overview

This model optimizes sentence representations through contrastive learning, efficiently calculating semantic similarity between Korean sentences, suitable for tasks like information retrieval and text matching

Model Features

High-performance Semantic Matching

Achieves an average score of 83.37 on Korean STS tasks, outperforming similar baseline models

Multi-dimensional Similarity Calculation

Supports various similarity measures such as cosine similarity, Euclidean distance, and Manhattan distance

Ready-to-use Pre-trained Model

Provides an out-of-the-box pre-trained model supporting fast inference

Model Capabilities

Sentence Vector Generation

Semantic Similarity Calculation

Text Matching

Information Retrieval

Use Cases

Text Matching

Q&A Systems

Matching user questions with similar questions in the knowledge base

Improves Q&A accuracy

Document Deduplication

Identifying semantically similar documents

Effectively reduces duplicate content

Information Retrieval

Semantic Search

Search enhancement based on semantics rather than keyword matching

Improves search result relevance

🚀 Korean-Sentence-Embedding

This is a Korean sentence embedding repository. You can download pre-trained models and perform inference immediately. It also provides an environment for individuals to train models.

🚀 Quick Start

💻 Usage Examples

🔍 Basic Usage

import torch
from transformers import AutoModel, AutoTokenizer

def cal_score(a, b):
    if len(a.shape) == 1: a = a.unsqueeze(0)
    if len(b.shape) == 1: b = b.unsqueeze(0)

    a_norm = a / a.norm(dim=1)[:, None]
    b_norm = b / b.norm(dim=1)[:, None]
    return torch.mm(a_norm, b_norm.transpose(0, 1)) * 100

model = AutoModel.from_pretrained('BM-K/KoSimCSE-bert') 
AutoTokenizer.from_pretrained('BM-K/KoSimCSE-bert')

sentences = ['치타가 들판을 가로 질러 먹이를 쫓는다.',
             '치타 한 마리가 먹이 뒤에서 달리고 있다.',
             '원숭이 한 마리가 드럼을 연주한다.']

inputs = tokenizer(sentences, padding=True, truncation=True, return_tensors="pt")
embeddings, _ = model(**inputs, return_dict=False)

score01 = cal_score(embeddings[0][0], embeddings[1][0])
score02 = cal_score(embeddings[0][0], embeddings[2][0])

✨ Performance

Results of the Semantic Textual Similarity test set

Model	AVG	Cosine Pearson	Cosine Spearman	Euclidean Pearson	Euclidean Spearman	Manhattan Pearson	Manhattan Spearman	Dot Pearson	Dot Spearman
KoSBERT^†_SKT	77.40	78.81	78.47	77.68	77.78	77.71	77.83	75.75	75.22
KoSBERT	80.39	82.13	82.25	80.67	80.75	80.69	80.78	77.96	77.90
KoSRoBERTa	81.64	81.20	82.20	81.79	82.34	81.59	82.20	80.62	81.25

KoSentenceBART	77.14	79.71	78.74	78.42	78.02	78.40	78.00	74.24	72.15
KoSentenceT5	77.83	80.87	79.74	80.24	79.36	80.19	79.27	72.81	70.17

KoSimCSE-BERT^†_SKT	81.32	82.12	82.56	81.84	81.63	81.99	81.74	79.55	79.19
KoSimCSE-BERT	83.37	83.22	83.58	83.24	83.60	83.15	83.54	83.13	83.49
KoSimCSE-RoBERTa	83.65	83.60	83.77	83.54	83.76	83.55	83.77	83.55	83.64

KoSimCSE-BERT-multitask	85.71	85.29	86.02	85.63	86.01	85.57	85.97	85.26	85.93
KoSimCSE-RoBERTa-multitask	85.77	85.08	86.12	85.84	86.12	85.83	86.12	85.03	85.99

You can access the repository here.

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