utils.py

import os

def calculate_similarity():
    image_features = embeddings[ModalityType.VISION]
    text_features = embeddings[ModalityType.TEXT]

    image_features /= image_features.norm(dim=-1, keepdim=True)
    text_features /= text_features.norm(dim=-1, keepdim=True)
    similarity = text_features.cpu().numpy() @ image_features.cpu().numpy().T