# Sentence Pair Scoring

An automated trainable metric for evaluating semantic preservation between sentences

This is a French sentence similarity calculation model based on CamemBERT, used to predict the semantic similarity score between two sentences.

A cross-encoder based on DistilRoBERTa for natural language inference tasks, capable of determining the relationship between sentence pairs (contradiction, entailment, or neutral).

This model is a cross-encoder trained based on the RoBERTa-large architecture, specifically designed to predict semantic similarity between two sentences, with scores ranging from 0 to 1.