Sentence Transformers (SBERT)
State-of-the-art Python framework for sentence, text, and image embeddings using siamese BERT networks, providing access to 15,000+ pre-trained models for semantic search, similarity comparison, and clustering.
About this tool
Overview
SBERT is the go-to Python module for accessing, using, and training state-of-the-art embedding and reranker models. Sentence-BERT (SBERT) is a modification of the pretrained BERT network that uses siamese and triplet network structures to derive semantically meaningful sentence embeddings.
Architecture
SBERT implements an additional pooling layer on top of BERT, enabling the model to produce fixed-size sentence embeddings that can be compared using cosine similarity without requiring inference computation for every sentence-pair comparison.
Key Features
- 15,000+ Pre-trained Models: Available on 🤗 Hugging Face
- Semantic Embeddings: Generate dense vector representations for sentences
- Fast Comparison: Cosine similarity for sentence-pair comparison
- Multiple Tasks: Semantic search, similarity, clustering, reranking
- Easy to Use: Simple Python API
Performance
SBERT outperformed the previous state-of-the-art (SOTA) models for all common semantic textual similarity (STS) tasks, achieving significantly better performance while being much faster than computing BERT representations for every pair.
Applications
- Semantic search
- Semantic textual similarity
- Paraphrase mining
- Text clustering
- Document classification
- Question answering
Model Categories
- Text Embedding Models: For semantic search and similarity
- Cross-Encoders: For reranking and classification
- Multilingual Models: Support for 100+ languages
- Domain-Specific Models: Fine-tuned for specific domains
Popular Models
- all-MiniLM-L6-v2: Fast and efficient
- all-mpnet-base-v2: Best quality
- paraphrase-multilingual: Multilingual support
- MS MARCO models: Trained for search
Installation
pip install sentence-transformers
Basic Usage
from sentence_transformers import SentenceTransformer
model = SentenceTransformer('all-MiniLM-L6-v2')
embeddings = model.encode(['First sentence', 'Second sentence'])
Integration
- Hugging Face
- LangChain
- LlamaIndex
- Haystack
- Vector databases
Research Foundation
Paper: "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks" (2019) ArXiv: https://arxiv.org/abs/1908.10084
Resources
- GitHub: https://github.com/huggingface/sentence-transformers
- Documentation: https://sbert.net/
- PyPI: https://pypi.org/project/sentence-transformers/
Pricing
Free and open-source under Apache 2.0 license.
Loading more......
Information
Categories
Tags
Similar Products
6 result(s)