README.md

metadata

pipeline_tag: zero-shot-image-classification

This repository contains the models of the paper Preserving Multi-Modal Capabilities of Pre-trained VLMs for Improving Vision-Linguistic Compositionality.