Sample Efficient Ensemble Learning with Catalyst.RL
This addresses the challenge of computationally expensive and high-dimensional RL environments for researchers, though it is incremental as it builds on existing methods.
The authors tackled the problem of sample-efficient reinforcement learning by developing Catalyst.RL, a framework that achieved second place in a NeurIPS 2019 locomotion challenge, training agents in only a few hours.
We present Catalyst.RL, an open-source PyTorch framework for reproducible and sample efficient reinforcement learning (RL) research. Main features of Catalyst.RL include large-scale asynchronous distributed training, efficient implementations of various RL algorithms and auxiliary tricks, such as n-step returns, value distributions, hyperbolic reinforcement learning, etc. To demonstrate the effectiveness of Catalyst.RL, we applied it to a physics-based reinforcement learning challenge "NeurIPS 2019: Learn to Move -- Walk Around" with the objective to build a locomotion controller for a human musculoskeletal model. The environment is computationally expensive, has a high-dimensional continuous action space and is stochastic. Our team took the 2nd place, capitalizing on the ability of Catalyst.RL to train high-quality and sample-efficient RL agents in only a few hours of training time. The implementation along with experiments is open-sourced so results can be reproduced and novel ideas tried out.