Closed VpkPrasanna closed 3 years ago
@VpkPrasanna If you created this issue for the purpose of creating webapp then I will not put this as a priority at the moment. Streamlit app will not need an api if we are invoking the code through a package something like:
from cassava.pretrained import get_model
model = get_model('resnet50') # We can directly use this model to infer in the app
So in my opinion first we will have to package this and then later as an added feature we can have an api if anyone wants to consume through their webapp/interface, makes sense?
@VpkPrasanna If you created this issue for the purpose of creating webapp then I will not put this as a priority at the moment. Streamlit app will not need an api if we are invoking the code through a package something like:
from cassava.pretrained import get_model model = get_model('resnet50') # We can directly use this model to infer in the app
So in my opinion first we will have to package this and then later as an added feature we can have an api if anyone wants to consume through their webapp/interface, makes sense?
Yeah i totally agree with you . I saw some small demo to create a streamlit app so i can work on that .
This issue can also be closed as it is completed @p-s-vishnu
I will implement the inference with fast API as a backend