Installation
1. Environment Setup
Install as a Package
You can install LeIsaac as a dependency. The script below provisions IsaacLab, IsaacSim, and all required components.
conda create -n leisaac python=3.11
conda activate leisaac
# Install cuda-toolkit
conda install -c "nvidia/label/cuda-12.8.1" cuda-toolkit
# Install PyTorch (CUDA 12.8 wheels)
pip install -U torch==2.7.0 torchvision==0.22.0 --index-url https://download.pytorch.org/whl/cu128
# Install LeIsaac and IsaacLab/IsaacSim extras
pip install 'leisaac[isaaclab] @ git+https://github.com/LightwheelAI/leisaac.git#subdirectory=source/leisaac' --extra-index-url https://pypi.nvidia.com
Install as a Package may expose edge cases. If you encounter issues, please open an issue on GitHub and consider switching to the “install from source” workflow described below.
Install from Source
You can also install directly from the source for local development. First, clone our repository and related submodules.
git clone https://github.com/LightwheelAI/leisaac.git --recursive
Then follow the IsaacLab official installation guide to install IsaacLab. We recommend using Conda for easier environment management. In summary, you only need to run the following command.
# Create and activate environment
conda create -n leisaac python=3.11
conda activate leisaac
# Install cuda-toolkit
conda install -c "nvidia/label/cuda-12.8.1" cuda-toolkit
# Install PyTorch
pip install -U torch==2.7.0 torchvision==0.22.0 --index-url https://download.pytorch.org/whl/cu128
# Install IsaacSim
pip install --upgrade pip
pip install "isaacsim[all,extscache]==5.1.0" --extra-index-url https://pypi.nvidia.com
# Install IsaacLab
sudo apt install cmake build-essential
cd leisaac/dependencies/IsaacLab
./isaaclab.sh --install
Finally, install leisaac as dependency.
cd ../..
pip install -e source/leisaac
The steps above are essentially the same as the official IsaacLab documentation; please adjust according to the versions you use. Below is the compatibility between LeIsaac and IsaacLab and the related version dependencies.
If you are using a 50-series GPU, we recommend using IsaacSim 5.0+ and IsaacLab v2.2.1+. We have tested on IsaacSim 5.0 and it works properly.
| Dependency | IsaacSim4.5 | IsaaSim5.0 | IsaacSim5.1 |
|---|---|---|---|
| Python | 3.10 | 3.11 | 3.11 |
| IsaacLab | v2.1.1 | v2.2.1 | v2.3.0 |
| CUDA | 11.8 | 12.8 | 12.8 |
| PyTorch | 2.5.1 | 2.7.0 | 2.7.0 |
2. Asset Preparation
We provide an example USD asset—a kitchen scene. Please download related scene here and extract it into the assets directory. The directory structure should look like this:
<assets>
├── robots/
│ └── so101_follower.usd
└── scenes/
└── kitchen_with_orange/
├── scene.usd
├── assets
└── objects/
├── Orange001
├── Orange002
├── Orange003
└── Plate
Below are the download links for the scenes we provide. For more high-quality scene assets, please visit our official website or the Releases page.
| Scene Name | Description | Download Link |
|---|---|---|
| Kitchen with Orange | Example kitchen scene with oranges | Download |
| Lightwheel Toyroom | Modern room with many toys | Download |
| Table with Cube | Simple table with one cube | Download |
| Lightwheel Bedroom | Realistic bedroom scene with cloth | Download |
You can also download scenes from huggingface, which be stored in the assets directory.
3. Device Setup
We use the SO101Leader as the teleoperation device. Please follow the official documentation for connection and configuration.
Note that you do not need to use the LeRobot repository for calibration; our codebase provides guided steps for the calibration process.