# NLI Fine-tuning

The EUBERT MLM model is fine-tuned as a cross-encoder using sentence transformer training sets, supporting 24 EU languages

An improved model based on BioBERT, fine-tuned with SNLI and MultiNLI datasets for generating universal sentence embeddings

A model based on SciBERT, fine-tuned with SNLI and MultiNLI datasets, for generating universal sentence embeddings

SGPT is a sentence embedding model based on the GPT architecture, specifically designed for semantic search tasks, generating sentence representations through weighted mean pooling.