Then we create an MXNet context for CPU-based computation. Since I am doing it by CPU, I instantiated for the CPU. Feel free to use the GPU if you have already configured it by providing the device ID:

// Retrieves the name of this Context object 
val ctx = Context.cpu()