RLgraph: Modular Computation Graphs for Deep Reinforcement Learning. Reinforcement learning (RL) tasks are challenging to implement, execute and test due to algorithmic instability, hyper-parameter sensitivity, and heterogeneous distributed communication patterns. We argue for the separation of logical component composition, backend graph definition, and distributed execution. To this end, we introduce RLgraph, a library for designing and executing reinforcement learning tasks in both static graph and define-by-run paradigms. The resulting implementations are robust, incrementally testable, and yield high performance across different deep learning frameworks and distributed backends.
Keywords for this software
References in zbMATH (referenced in 2 articles , 1 standard article )
Showing results 1 to 2 of 2.
- Sergey Kolesnikov, Oleksii Hrinchuk: Catalyst.RL: A Distributed Framework for Reproducible RL Research (2019) arXiv
- Michael Schaarschmidt, Sven Mika, Kai Fricke, Eiko Yoneki: RLgraph: Modular Computation Graphs for Deep Reinforcement Learning (2018) arXiv