What is the need of separate initialization for img_model and text_model?

#2
by gokulkarthik - opened

The model initialization from SentenceTransformer('clip-ViT-B-32-multilingual-v1') could be used for processing both the text and the images.

Sign up or log in to comment