all-MiniLM-L6-v2 is a compact Sentence-Embedding model from the sentence-transformers project, based on Microsoft's MiniLM backbone. It maps sentences and short paragraphs into a 384-dimensional dense vector space and is suited for semantic search, clustering, and similarity comparisons. The model was Fine-Tuned with a contrastive learning objective on over 1 billion sentence pairs and is Open Source under the Apache-2.0 license on Hugging Face. Due to its small size (~22.7 million parameters), it serves as the standard baseline for fast, resource-efficient embedding applications.

Zdroje (3)