| Property | Value |
|---|---|
| Parameter Count | 135M |
| Output Dimensions | 768 |
| License | Apache 2.0 |
| Framework Support | PyTorch, TensorFlow, ONNX |
| Paper | Sentence-BERT Paper |
What is distilbert-multilingual-nli-stsb-quora-ranking?
This is a sophisticated sentence embedding model based on DistilBERT architecture, designed to convert sentences and paragraphs into fixed-length vector representations. It's specifically optimized for multilingual applications and trained on a combination of Natural Language Inference (NLI), Semantic Textual Similarity Benchmark (STSB), and Quora question pair datasets.
Implementation Details
The model implements a two-step architecture combining a DistilBERT transformer with a pooling layer. It processes text sequences up to 128 tokens and outputs 768-dimensional embeddings. The implementation supports both sentence-transformers and HuggingFace Transformers frameworks, with mean pooling as the default aggregation strategy.
- Utilizes DistilBERT's efficient architecture for reduced computational requirements
- Implements mean pooling over token embeddings
- Supports multiple deep learning frameworks including PyTorch and TensorFlow
- Downloaded over 270,000 times, indicating strong community adoption
Core Capabilities
- Multilingual sentence embedding generation
- Semantic similarity computation
- Text clustering and classification
- Cross-lingual information retrieval
- Question-answer matching
Frequently Asked Questions
Q: What makes this model unique?
This model stands out for its multilingual capabilities while maintaining a relatively compact size (135M parameters). It's specifically optimized for semantic similarity tasks and can be used across multiple languages without requiring separate models.
Q: What are the recommended use cases?
The model excels in semantic search applications, document clustering, similarity matching, and multilingual text comparison. It's particularly useful for applications requiring cross-lingual semantic understanding or large-scale text similarity computations.