Decomposing a Recurrent Neural Network into Modules for Enabling Reusability and Replacement

Open in new window