dm_control: Software and Tasks for Continuous Control

Abstract

The dm_control software package is a collection of Python libraries and task suites for reinforcement learning agents in an articulated-body simulation. A MuJoCo wrapper provides convenient bindings to functions and data structures. The PyMJCF and Composer libraries enable procedural model manipulation and task authoring. The Control Suite is a fixed set of tasks with standardised structure, intended to serve as performance benchmarks. The Locomotion framework provides high-level abstractions and examples of locomotion tasks. A set of configurable manipulation tasks with a robot arm and snap-together bricks is also included. dm_control is publicly available at https://www.github.com/deepmind/dm_control

arXiv admin note: text overlap with arXiv:1801.00690

References in corpus (17)

• Continuous control with deep reinforcement learning
• Trust Region Policy Optimization
• Asynchronous Methods for Deep Reinforcement Learning
• Benchmarking Deep Reinforcement Learning for Continuous Control
• Emergence of Locomotion Behaviours in Rich Environments
• DeepMind Control Suite
• Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning
• Maximum a Posteriori Policy Optimisation
• Learning human behaviors from motion capture by adversarial imitation
• Reinforcement and Imitation Learning for Diverse Visuomotor Skills
• Neural probabilistic motor primitives for humanoid control
• Emergent Coordination Through Competition
• Hierarchical visuomotor control of humanoids
• DeepMind Lab
• Learning Awareness Models
• IKEA Furniture Assembly Environment for Long-Horizon Complex Manipulation Tasks
• Simultaneously Learning Vision and Feature-based Control Policies for Real-world Ball-in-a-Cup

Cited by in corpus (40)

• Embodied Intelligence via Learning and Evolution
• A Survey on Offline Reinforcement Learning: Taxonomy, Review, and Open Problems
• Learning Agile Soccer Skills for a Bipedal Robot with Deep Reinforcement Learning
• ManiSkill: Generalizable Manipulation Skill Benchmark with Large-Scale Demonstrations
• Contrastive Behavioral Similarity Embeddings for Generalization in Reinforcement Learning
• Knowledge Transfer for Cross-Domain Reinforcement Learning: A Systematic Review
• Sample-efficient Reinforcement Learning Representation Learning with Curiosity Contrastive Forward Dynamics Model
• State Entropy Maximization with Random Encoders for Efficient Exploration
• StARformer: Transformer with State-Action-Reward Representations for Visual Reinforcement Learning
• MBRL-Lib: A Modular Library for Model-based Reinforcement Learning
• PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-training
• CCLF: A Contrastive-Curiosity-Driven Learning Framework for Sample-Efficient Reinforcement Learning
• Behavior From the Void: Unsupervised Active Pre-Training
• MADE: Exploration via Maximizing Deviation from Explored Regions
• Learning and Planning in Complex Action Spaces
• SINDy-RL: Interpretable and Efficient Model-Based Reinforcement Learning
• Deep Reinforcement Learning for Concentric Tube Robot Path Following
• rl_reach: Reproducible Reinforcement Learning Experiments for Robotic Reaching Tasks
• Learning Markov State Abstractions for Deep Reinforcement Learning
• Generalized Policy Improvement Algorithms with Theoretically Supported Sample Reuse
• Physically Embedded Planning Problems: New Challenges for Reinforcement Learning
• Reward Bonuses with Gain Scheduling Inspired by Iterative Deepening Search
• Scenic4RL: Programmatic Modeling and Generation of Reinforcement Learning Environments
• A Comparative Tutorial of Bayesian Sequential Design and Reinforcement Learning
• Is Bang-Bang Control All You Need? Solving Continuous Control with Bernoulli Policies
• SILG: The Multi-environment Symbolic Interactive Language Grounding Benchmark
• Building Minimal and Reusable Causal State Abstractions for Reinforcement Learning
• $\mathtt{MORALS}$: Analysis of High-Dimensional Robot Controllers via Topological Tools in a Latent Space
• Eigensubspace of Temporal-Difference Dynamics and How It Improves Value Approximation in Reinforcement Learning
• Learning Temporally-Consistent Representations for Data-Efficient Reinforcement Learning
• Multi-Step First: A Lightweight Deep Reinforcement Learning Strategy for Robust Continuous Control with Partial Observability
• Revisiting Experience Replayable Conditions
• Intentionally-underestimated Value Function at Terminal State for Temporal-difference Learning with Mis-designed Reward
• Hierarchical Skills for Efficient Exploration
• Evaluating model-based planning and planner amortization for continuous control
• Phase-Amplitude Reduction-Based Imitation Learning
• Active Offline Policy Selection
• KeyCLD: Learning Constrained Lagrangian Dynamics in Keypoint Coordinates from Images
• Dynamic Horizon Value Estimation for Model-based Reinforcement Learning
• Multitask Adaptation by Retrospective Exploration with Learned World Models