BayesSimIG: Scalable Parameter Inference for Adaptive Domain Randomization with IsaacGym
This work addresses the challenge of efficient domain randomization for robotics tasks, though it is incremental as it builds on existing methods by providing a scalable implementation.
The paper tackles the problem of scalable parameter inference for domain randomization in reinforcement learning by introducing BayesSimIG, a library that integrates BayesSim with NVIDIA IsaacGym, enabling large-scale GPU-accelerated inference and simulation with support for over 10K parallel environments and more than 100 parameters.
BayesSim is a statistical technique for domain randomization in reinforcement learning based on likelihood-free inference of simulation parameters. This paper outlines BayesSimIG: a library that provides an implementation of BayesSim integrated with the recently released NVIDIA IsaacGym. This combination allows large-scale parameter inference with end-to-end GPU acceleration. Both inference and simulation get GPU speedup, with support for running more than 10K parallel simulation environments for complex robotics tasks that can have more than 100 simulation parameters to estimate. BayesSimIG provides an integration with TensorBoard to easily visualize slices of high-dimensional posteriors. The library is built in a modular way to support research experiments with novel ways to collect and process the trajectories from the parallel IsaacGym environments.