python - How to deploy and serve prediction using TensorFlow from API? -
from google tutorial know how train model in tensorflow. best way save trained model, serve prediction using basic minimal python api in production server.
my question tensorflow best practices save model , serve prediction on live server without compromising speed , memory issue. since api server running on background forever.
a small snippet of python code appreciated.
tensorflow serving high performance, open source serving system machine learning models, designed production environments , optimized tensorflow. initial release contains c++ server , python client examples based on grpc. basic architecture shown in diagram below.
to started quickly, check out tutorial.
Comments
Post a Comment