How to use vsearch/vdr-cross-modal with Transformers:
# Load model directly from transformers import Retriever model = Retriever.from_pretrained("vsearch/vdr-cross-modal", dtype="auto")
