Modly extension for Hunyuan3D 2 Mini, Tencent's lightweight image-to-3D pipeline.
- installs an isolated Python environment for the extension
- loads the Hunyuan3D 2 Mini shape pipeline inside Modly
- optionally uses the texture pipeline when the required native extensions are available
At install time, setup.py creates a virtual environment and selects the PyTorch stack from the platform information that Modly passes in:
gpu_smcuda_version- operating system / CPU architecture
Linux ARM64 support is handled explicitly in setup.py.
The installer now:
- uses ARM64-specific PyTorch/TorchVision wheel selection
- switches to the CUDA 12.8 path when Modly reports
cuda_version >= 128 - installs pinned direct wheel URLs with published
sha256hashes - keeps the matching PyTorch index available for transitive dependencies such as
triton - increases pip timeout/retry settings for large wheel downloads
- uses
rembg+onnxruntimeon ARM64 instead ofrembg[gpu]
This path was validated successfully on Linux ARM64 in Modly for both:
- extension installation
- mesh generation
The Windows path remains unchanged.
The ARM64-specific logic is gated to:
platform.system() == "Linux"platform.machine() in {"aarch64", "arm64"}
So Windows still uses the original virtualenv executable layout (Scripts/python.exe, Scripts/pip.exe) and the existing non-ARM64 dependency resolution path.
- If installation fails after changing installer logic, run Repair in Modly so the extension venv is recreated.
- If texture generation fails, verify the optional native texture-generation extensions are available for your environment.
- Model weights:
tencent/Hunyuan3D-2mini - Project source:
Tencent/Hunyuan3D-2