Paper in ICLR 2020 on “DD-PPO: Learning Near-Perfect PointGoal Navigators from 2.5 Billion Frames”
Abstract
We present Decentralized Distributed Proximal Policy Optimization (DD-PPO), a method for distributed reinforcement learning in resource-intensive simulated environments. DD-PPO is distributed (uses multiple machines), decentralized (lacks a centralized server), and synchronous (no computation is ever ‘stale’), making it conceptually simple and easy to implement. In our experiments on training virtual robots to navigate in Habitat-Sim, DD-PPO exhibits near-linear scaling – achieving a speedup of 107x on 128 GPUs over a serial implementation. We leverage this scaling to train an agent for 2.5 Billion steps of experience (the equivalent of 80 years of human experience) – over 6 months of GPU-time training in under 3 days of wall-clock time with 64 GPUs.
This massive-scale training not only sets the state of art on Habitat Autonomous Navigation Challenge 2019, but essentially ‘solves’ the task – near-perfect autonomous navigation in an unseen environment without access to a map, directly from an RGB-D camera and a GPS+Compass sensor. Fortuitously, error vs computation exhibits a power-law-like distribution; thus, 90% of peak performance is obtained relatively early (at 100 million steps) and relatively cheaply (under 1 day with 8 GPUs). Finally, we show that the scene understanding and navigation policies learned can be transferred to other navigation tasks – the analog of ‘ImageNet pre-training + task-specific fine-tuning’ for embodied AI. Our model outperforms ImageNet pre-trained CNNs on these transfer tasks and can serve as a universal resource (all models + code will be publicly available).
Citation
- E. Wijmans, A. Kadian, A. Morcos, S. Lee, I. Essa, D. Parikh, M. Savva, and D. Batra (2020), “Decentralized Distributed PPO: Solving PointGoal Navigation,” in Proceedings of International Conference on Learning Representations (ICLR), 2020. [PDF] [arXiv] [BIBTEX]
@InProceedings{ 2020-Wijmans-DDSPN, archiveprefix = {arXiv}, arxiv = {https://arxiv.org/abs/1911.00357}, author = {Erik Wijmans and Abhishek Kadian and Ari Morcos and Stefan Lee and Irfan Essa and Devi Parikh and Manolis Savva and Dhruv Batra}, booktitle = {{Proceedings of International Conference on Learning Representations (ICLR)}}, eprint = {1911.00357}, month = {April}, pdf = {https://arxiv.org/pdf/1911.00357}, primaryclass = {cs.CV}, title = {Decentralized Distributed PPO: Solving PointGoal Navigation}, year = {2020} }