Installation¶
From PyPI (Recommended)¶
Supported platforms:
- Linux (x86_64)
- macOS (arm64 - Apple Silicon)
- Windows (x86_64)
Requirements¶
- Python 3.10 or higher
- PyTorch 2.0 or higher
Verify Installation¶
import torch_webgpu
import torch
# Create a tensor on WebGPU
x = torch.tensor([1.0, 2.0, 3.0], device="webgpu")
print(x)
# tensor([1., 2., 3.], device='webgpu')
From Source¶
For development or if you need the latest changes:
# Clone the repository
git clone https://github.com/jmaczan/torch-webgpu.git
cd torch-webgpu
# Build Dawn (WebGPU runtime)
./scripts/build-dawn.sh
# Install in development mode
pip install -e .
See Building from Source for detailed instructions.
Troubleshooting¶
Import Error¶
If you get an import error, make sure PyTorch is installed:
GPU Not Found¶
torch-webgpu uses Dawn (Google's WebGPU implementation) which supports:
- Vulkan (Linux, Windows)
- Metal (macOS)
- D3D12 (Windows)
Make sure you have appropriate GPU drivers installed.