Steps

• Init repository
• Scaffold plugin
• Depend on socket.io client plugin
• Define C++ tensorflow base component
• sub-class tensorflow-remote component
• refactor into machine learning generic api
• build simple python server with socket.io listener, calls TFPluginAPI scripts in same format as current tensorflow-ue4 component does. Selects script remotely from client (based on name).
• add a non-component c++ interface used as one parent for component base
• finish Initial build ml-remote-server#1
• initial working build test
• documentation