# Billion-level sentence pair training

A sentence embedding model based on mpnet-base, trained on 1 billion sentence pairs through self-supervised contrastive learning, capable of generating high-quality semantic vector representations of sentences