FastTD3: Simple, Fast, and Capable Reinforcement Learning for Humanoid Control
By: Younggyo Seo , Carmelo Sferrazza , Haoran Geng and more
Potential Business Impact:
Teaches robots to walk much faster.
Reinforcement learning (RL) has driven significant progress in robotics, but its complexity and long training times remain major bottlenecks. In this report, we introduce FastTD3, a simple, fast, and capable RL algorithm that significantly speeds up training for humanoid robots in popular suites such as HumanoidBench, IsaacLab, and MuJoCo Playground. Our recipe is remarkably simple: we train an off-policy TD3 agent with several modifications -- parallel simulation, large-batch updates, a distributional critic, and carefully tuned hyperparameters. FastTD3 solves a range of HumanoidBench tasks in under 3 hours on a single A100 GPU, while remaining stable during training. We also provide a lightweight and easy-to-use implementation of FastTD3 to accelerate RL research in robotics.
Similar Papers
Learning Sim-to-Real Humanoid Locomotion in 15 Minutes
Robotics
Teaches robots to walk in minutes.
Hybrid DQN-TD3 Reinforcement Learning for Autonomous Navigation in Dynamic Environments
Robotics
Helps robots navigate tricky paths safely and smoothly.
TDMPBC: Self-Imitative Reinforcement Learning for Humanoid Robot Control
Robotics
Helps robots learn to do hard tasks faster.