search

jsphon / reinforcement_learning

Python Package For Reinforcement Learning
0 stars 0 forks source link

Refactor the State Machine States #36

Closed jsphon closed 7 years ago

jsphon commented 7 years ago

The current implementation of having IntExtStates, which hold internal and external states, is not working out.

When doing vectorized learning, it seems a bit contrived that we must create a new IntExtState per iteration.

However, if we remove the IntExtState, what will we pass to the target generators?

jsphon commented 7 years ago

Keeping INtExtStates for the time being, though we should think of a better name.