Semantic Textual Similarity On Sts15
Métriques
Spearman Correlation
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | Spearman Correlation |
|---|---|
| diffcse-difference-based-contrastive-learning | 0.8390 |
| scaling-sentence-embeddings-with-large | 0.8951 |
| on-the-sentence-embeddings-from-pre-trained | 0.7492 |
| trans-encoder-unsupervised-sentence-pair | 0.8577 |
| deep-continuous-prompt-for-contrastive-1 | 0.8808 |
| trans-encoder-unsupervised-sentence-pair | 0.8508 |
| simcse-simple-contrastive-learning-of | 0.8666 |
| trans-encoder-unsupervised-sentence-pair | 0.8816 |
| angle-optimized-text-embeddings | 0.8956 |
| diffcse-difference-based-contrastive-learning | 0.8281 |
| trans-encoder-unsupervised-sentence-pair | 0.8444 |
| sentence-bert-sentence-embeddings-using | 0.8185 |
| an-unsupervised-sentence-embedding-method | 0.7523 |
| angle-optimized-text-embeddings | 0.8943 |
| generating-datasets-with-pretrained-language | 0.8049 |
| fast-effective-and-self-supervised | 0.798 |
| trans-encoder-unsupervised-sentence-pair | 0.8863 |
| fast-effective-and-self-supervised | 0.814 |
| scaling-sentence-embeddings-with-large | 0.8952 |
| scaling-sentence-embeddings-with-large | 0.9004 |