toddlerbot.locomotion package¶
Submodules¶
toddlerbot.locomotion.cartwheel_env module¶
toddlerbot.locomotion.crawl_env module¶
toddlerbot.locomotion.mjx_config module¶
toddlerbot.locomotion.mjx_env module¶
toddlerbot.locomotion.on_policy_runner module¶
toddlerbot.locomotion.ppo_config module¶
toddlerbot.locomotion.rsl_rl_wrapper module¶
toddlerbot.locomotion.train_mjx module¶
toddlerbot.locomotion.walk_env module¶
Module contents¶
Locomotion and movement control for ToddlerBot.
This package contains implementations for various locomotion strategies and movement patterns for the ToddlerBot humanoid robot, including:
Walking gaits using reinforcement learning (PPO) and classical control
Cartwheel and acrobatic movement patterns
Crawling and low-profile locomotion
Balance and stability control
Environment interaction and terrain adaptation
The locomotion modules support both MuJoCo-based training environments and real robot deployment, with configurable parameters for different robot configurations and movement requirements.
Key components include environment definitions, training configurations, reward functions, and policy interfaces for various locomotion behaviors.