How rhoban/frasa works
FRASA is a reinforcement-learning environment and training setup for teaching a Sigmaban humanoid to recover and stand from prone or supine falls. The main path registers the Gymnasium environment, runs the stand-up task on MuJoCo, and lets RL training learn a recovery policy; a separate runner loads trained agents for demonstration. Initial-state generation and direct environment testing support this workflow. The source excerpts establish the environment-to-simulator connection; training, evaluation, and configuration links are based on README descriptions where runner internals were not sampled.
Main entry points: __init__.py Gym registration · train_sbx.py Training runner · standup_env.py Stand-up task · enjoy_sbx.py Agent runner
Components 13
RL workflow
- Training runner
train_sbx.py - Algorithm configs
hyperparams - RL algorithms
- Agent runner
enjoy_sbx.py
Stand-up task
- Gym registration
frasa_env/__init__.py - FRASA defaults
frasa_env/env/standup.py - Stand-up task
frasa_env/env/standup_env.py
Simulation model
- MuJoCo simulator
frasa_env/mujoco_simulator/simulator.py - Sigmaban model
frasa_env/mujoco_simulator/model
Task support
- Initial-state generator
standup_generate_initial.py - Environment tester
test_env.py - Fall-state cache
frasa_env/env/standup_env.py
Other components
- Researcher
Connections 15
- Researcher → Training runner: starts training
README.md - Training runner → Algorithm configs: loads parameters
README.md - Training runner → RL algorithms: trains policy
README.md - Training runner → Gym registration: trains environment
README.md - Researcher → Agent runner: runs agent
README.md - Agent runner → Gym registration: loads environment
README.md - Gym registration → FRASA defaults: registers entry point
frasa_env/__init__.py - FRASA defaults → Stand-up task: inherits task
frasa_env/env/standup.py - Stand-up task → MuJoCo simulator: steps simulation
frasa_env/env/standup_env.py - MuJoCo simulator → Sigmaban model: loads scene
frasa_env/mujoco_simulator/simulator.py - Researcher → Initial-state generator: generates fall states
README.md - Initial-state generator → Gym registration: creates environment
standup_generate_initial.py - Stand-up task → Fall-state cache: loads cached states
frasa_env/env/standup_env.py - Researcher → Environment tester: tests environment
README.md - Environment tester → Gym registration: tests environment