zuoxingdong / lagom

lagom: A PyTorch infrastructure for rapid prototyping of reinforcement learning algorithms.
MIT License
373 stars 30 forks source link

Migrate obs/reward normalization from env.wrappers into Agent itself #205

Open zuoxingdong opened 4 years ago

zuoxingdong commented 4 years ago