Finnish Sentence-BERT Cased
This Finnish sentence embedding model adapts the cased FinBERT checkpoint for paraphrase similarity. It returns sentence representations rather than generating text.
Model design and training
Training uses the Finnish Paraphrase Corpus and automatically collected candidates, including 500,000 positive and five million negative pairs. Mean pooling combines token representations. The model learns whether two sentences are paraphrases. Its card does not publish an explicit model license; the reviewed repository contains no license file.sbert-cased-finnish-paraphrase: official model card
Practical assessment
Practical guidance. Use the cased release when capitalization may help distinguish names and sentence structure. Build the document index and query vectors with the same encoder and pooling settings. Similarity finds related wording; it does not establish factual agreement or entailment.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | sbert-cased-finnish-paraphrase |
| Context window | 128 tokens Default SentenceTransformer maximum sequence length; longer text is truncated under these packaged settings. sbert-cased-finnish-paraphrase: sentence encoder settings |
| License | Not specified in the reviewed derivative card or repository files.sbert-cased-finnish-paraphrase: official model card |
| Base model | TurkuNLP/bert-base-finnish-cased-v1.sbert-cased-finnish-paraphrase: official model card |
| Pooling | Mean pooling over token embeddings.sbert-cased-finnish-paraphrase: official model card |
| Case handling | do_lower_case is false.sbert-cased-finnish-paraphrase: sentence encoder settings |
Related models
- Finnish Sentence-BERT Uncased
TurkuNLP ยท Finnish Sentence-BERT
This Finnish sentence embedding model adapts the uncased FinBERT checkpoint for paraphrase similarity.
Launch date not confirmedDownloadable