Update README.md

README.md

@@ -16,7 +16,7 @@ datasets:
 
 ***Note**: The same model trained with negatives yields better performance. [Find it here](https://huggingface.co/jamescalam/mpnet-snli-negatives).*
 
-This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search. It has been fine-tuned using the **S**tanford **N**atural **L**anguage **I**nference (SNLI) dataset and returns MRR@10 and MAP scores of ~0.95 on the SNLI test set.
+This is a [sentence-transformers](https://www.SBERT.net) model: It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for tasks like clustering or semantic search. It has been fine-tuned using the **S**tanford **N**atural **L**anguage **I**nference (SNLI) dataset and returns MRR@10 and MAP scores of ~0.92 on the SNLI test set.
 
 Find more info from [James Briggs on YouTube](https://youtube.com/c/jamesbriggs) or in the [**free** NLP for Semantic Search ebook](https://pinecone.io/learn/nlp).