I\'m working with a task of deploy Elastic Inference on AWS server.
Our model is Tacotron model with encoder, decoder and postnet as it\'s children.
When I cal