Sentence Transformers v3.0
Major update to the Sentence Transformers library introducing a new SentenceTransformerTrainer for easier fine-tuning, multi-GPU support, improved loss logging, and access to 15,000+ pre-trained models on HuggingFace.
About this tool
Overview
Sentence Transformers v3.0 represents the largest update since the project's inception, with a completely redesigned training approach and extensive new features.
Major Updates in v3.0
New Training System:
- SentenceTransformerTrainer for unified training
- Multi-GPU training support
- Enhanced loss logging
- Backward compatible with old training code
Model Access:
- 15,000+ pre-trained models on HuggingFace
- State-of-the-art MTEB leaderboard models
- Easy model loading and inference
Latest Version (2026)
Version 5.3.0 released March 12, 2026
- Python 3.10-3.13 support
- Regular updates and improvements
Core Capabilities
- Sentence and text embeddings
- Semantic similarity computation
- Cross-encoder reranking
- Sparse embeddings
Use Cases
- Fine-tuning embedding models
- Semantic search development
- Similarity computation
- Clustering and classification
Availability
PyPI: sentence-transformers
GitHub: UKPLab/sentence-transformers
Documentation: sbert.net
Surveys
Loading more......
Information
Websitesbert.net
PublishedMar 20, 2026
Categories
Tags
Similar Products
6 result(s)