Example of policy inference

[worldveil]

We need an example of policy inference for RLlib in the docs.

Could we include it in the intro example?

https://docs.ray.io/en/latest/rllib.html

I think it's important to show how users can actually use the thing they just trained when they are getting started to envision the end result.

[worldveil]

I'll also be adding a page or section in the docs (this page perhaps: https://docs.ray.io/en/latest/rllib-models.html ?), but once I have the example I can do the writing portion

[sven1977]

Added 3 scripts (normal case, LSTM, attention) to the examples folder and linked to the simple one from the RLlib index docs page. Closing this issue now.