RL#

RL with stable-baselines3#

We provide a basic RL training example.

RL framework: stable-baselines3

RL algorithm: PPO

Simulator: IsaacSim and MuJoCo

python metasim/scripts/RL/train_sb3.py --sim isaaclab --task Stand --num_envs 32 --wandb_entity <your_wandb_entity_name>

python metasim/scripts/RL/train_sb3.py --sim mujoco --task Stand --num_envs 32 --wandb_entity <your_wandb_entity_name>

Tasks:

  • Walk

  • Run

  • Stand