Local ML engine integration into an FL system
The successful integration of FL client libraries into a local ML engine is key to conducting FL in distributed environments later on.
The minimal_MLEngine.py file in the examples/minimal directory found in the GitHub repository at https://github.com/tie-set/simple-fl, as shown in Figure 5.2, provides an example of integrating FL client-side libraries into a minimal ML engine package:
Figure 5.2 – The minimal ML engine package
Next, we will explain what libraries need to be imported into the local ML engine in the following section.
Importing libraries for a local ML engine
The following code shows the importing process, where general libraries such as numpy, time, and Dict are imported first. The key part of this process is that Client is imported from the client.py file in the fl_main.agent folder. This way, a developer does not need to know too much about the code inside an FL system and just...
