Streamlit importing HDF5 FILES

Question

I am developing a car damage assessment project using streamlit. I have trained three densenet models and also yolov5 model for bounding boxes. Now i have connected with frontend using streamlit application and want to deloy that application in their cloud platform streamlit cloud.

For this application to be deployed my project needs to be on github. But i cannot upload my hdf5 files into github as they are very large. I uploaded into drive and tried like this:

@st.cache_resource
def load_models():
    stage1 = 'https://drive.google.com/file/d/1jwz47Yu7uUmPPtmCrzdF3q10lHtkuVyy/view?usp=drive_link'
    stage2 = 'https://drive.google.com/file/d/12wpjWBxuhxQLE-L5GcmIIiI7O_TaciIK/view?usp=drive_link'
    stage3 = 'https://drive.google.com/file/d/1_4drud9QqAjzZzZ4g1Q8bFG_FUixw-A7/view?usp=drive_link'
    model1 = load_model(stage1)
    model2 = load_model(stage2)
    model3 = load_model(stage3) 
    
    return model1,model2,model3

Any help on how can i load my models with the links so that i can upload on github.