search

PathmindAI / nativerl

Train reinforcement learning agents using AnyLogic or Python-based simulations
Apache License 2.0
19 stars 4 forks source link

pathmind simulation reward terms #450

Closed thetwotravelers closed 2 years ago

slinlee commented 2 years ago

@brettskymind @maxpumperla is this still relevant after our discussion?

maxpumperla commented 2 years ago

@brettskymind @slinlee I've purged all uses of the Python custom reward function idea, as per our discussion, and fixed all return types (plus some minor things I spotted). Good to go from my end.