TensorFlow Agents: Efficient Batched Reinforcement Learning in TensorFlow
This provides a flexible and efficient starting point for researchers and practitioners working on parallel reinforcement learning algorithms, though it is incremental as it builds on existing TensorFlow infrastructure.
The authors tackled the problem of inefficient reinforcement learning computation by introducing TensorFlow Agents, an infrastructure paradigm that enables batched parallel simulation of multiple environments and neural network computation, resulting in significantly accelerated training without manual synchronization.
We introduce TensorFlow Agents, an efficient infrastructure paradigm for building parallel reinforcement learning algorithms in TensorFlow. We simulate multiple environments in parallel, and group them to perform the neural network computation on a batch rather than individual observations. This allows the TensorFlow execution engine to parallelize computation, without the need for manual synchronization. Environments are stepped in separate Python processes to progress them in parallel without interference of the global interpreter lock. As part of this project, we introduce BatchPPO, an efficient implementation of the proximal policy optimization algorithm. By open sourcing TensorFlow Agents, we hope to provide a flexible starting point for future projects that accelerates future research in the field.