Sentence-transformers

why Streamlit could not load sentence bert model?
model = SentenceTransformer(‘paraphrase-multilingual-mpnet-base-v2’)

Probably too large?

Thanks for the help, will the huggingface API works?

Another question is the compatibility with CV2, it always show error when it is not open-CV- headless, but many app like paddleocr is not using headless