The first step is to save model. After finishing the neural network training, save Your model as file with .h5 extension.
Next, insert Your Deep Neural Network model into MLAPI.io and write simple config for Your model.
Then, write simple Class for Your Deep Neural Network model. For example, prediction of cats.
The access management in our API is powered by JWT. Every request to the service is signed by a unique token which you get every time you login.
Every user request is stored in a sqlite Data Base. The entries contain information on request type, data, headers and models response.
Just copy a folder containing model, config and Python file containing a class that overrides the loading and processing methods from our base class .