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

Re-assign community