Low-Precision Reinforcement Learning Running Soft Actor-Critic in Half Precision

Low-PrecisionReinforcementLearning:RunningSoftActor-CriticinHalfPrecisionJohanBjorck1XiangyuChen1ChristopherDeSa1CarlaP.Gomes1KilianQ.Weinberger1Abstractlearning,anemergingtrendforacceleratingdeepl...