LucasAlegre / morl-baselines

Multi-Objective Reinforcement Learning algorithms implementations.
https://lucasalegre.github.io/morl-baselines
MIT License
294 stars 47 forks source link

Debug eupg #91

Closed omidsbhn closed 8 months ago

omidsbhn commented 8 months ago

The EUPG algorithm has been updated to incorporate the discount factor (gamma) in the calculation of accrued and future rewards. In the current implementation, gamma is not utilized. Furthermore, the scalarization function in eupg_fishwood.py has been revised to handle scalarization for both episodic rewards and the combined sum of accrued and future rewards

ffelten commented 8 months ago

Fixes #88