# JacopoBandoni /BioBertRelationGenesDiseases

The following is a fine-tuning of the BioBert models on the GAD dataset.

The model works by masking the gene string with "@GENE$" and the disease string with "@DISEASE$".

The output is a text classification that can either be:

• "LABEL0" if there is no relation
• "LABEL1" if there is a relation.