is the GA implemented?

Hello,

I am trying to reproduce the results shown in this [paper](https://idea.edu.cn/person/guojian/assets/papers/FinRl-Podracer.pdf)


I started by looking at the _demo.py_ file, which leads to the _train_and_evaluate_mg_ function in _run_ga.py_

But it seems this code is only going to perform one "pod" training on one "generation". Indeed, I Don't see any code related to a Genetic Algorithm mentioned in the paper.

It would be great if we could have some hints about how to run the GA strategy…

Thanks