gaponet - GapONet sim-to-real humanoid robot control workflows.
Provides workflows for GapONet environment validation, policy training, evaluation, checkpoint export, and lightweight deployment in Isaac Lab.
Tags
Updated: 2026-09-20Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Validate environment readiness
- Run policy training
- Evaluate policy checkpoints
- Export JIT policy checkpoints
- Deploy JIT policy models
Inputs
- JSON configuration files
- NPZ motion data files
- PyTorch model checkpoints
- Robot USD or URDF assets
Outputs
- Run manifest files
- Execution log files
- Training and evaluation metrics JSON
- Exported TorchScript policy files
- Environment check result JSON
Requirements
- Python 3.10
- Isaac Sim
- Isaac Lab
- PyTorch
- CUDA-compatible GPU
- rsl_rl package