Massively Parallel Methods for Deep Reinforcement Learning
This work addresses the computational bottleneck in deep reinforcement learning for researchers and practitioners, enabling faster and more efficient training.
The authors tackled the challenge of scaling deep reinforcement learning by developing a massively distributed architecture, which surpassed non-distributed DQN performance in 41 out of 49 Atari games and reduced training time by an order of magnitude.
We present the first massively distributed architecture for deep reinforcement learning. This architecture uses four main components: parallel actors that generate new behaviour; parallel learners that are trained from stored experience; a distributed neural network to represent the value function or behaviour policy; and a distributed store of experience. We used our architecture to implement the Deep Q-Network algorithm (DQN). Our distributed algorithm was applied to 49 games from Atari 2600 games from the Arcade Learning Environment, using identical hyperparameters. Our performance surpassed non-distributed DQN in 41 of the 49 games and also reduced the wall-time required to achieve these results by an order of magnitude on most games.