-
Notifications
You must be signed in to change notification settings - Fork 0
Installation
Gaurav14cs17 edited this page Jun 21, 2026
·
1 revision
- Python 3.8+
- PyTorch 2.0+
- TorchVision 0.15+
git clone https://github.com/FlashVision/FlashFusion.git
cd FlashFusion
pip install -e .# With ONNX export support
pip install -e ".[export]"
# With analytics (matplotlib, pandas)
pip install -e ".[analytics]"
# With FlashVision model support
pip install -e ".[flash]"
# Everything
pip install -e ".[all]"
# Development tools
pip install -e ".[dev]"cd docker
docker compose up -d
docker exec -it flashfusion bashflashfusion checkThis will verify all required and optional dependencies are installed correctly.
For a quick environment setup with conda:
bash setup_env.shFlashFusion automatically detects CUDA availability. To verify:
flashfusion versionThis shows your PyTorch version and CUDA status.
| Issue | Solution |
|---|---|
ImportError: torch |
Install PyTorch: pip install torch torchvision
|
CUDA not available |
Install CUDA-enabled PyTorch from pytorch.org |
onnx export fails |
Install export deps: pip install -e ".[export]"
|
matplotlib not found |
Install analytics deps: pip install -e ".[analytics]"
|
FlashFusion — Multi-model vision fusion | PyPI | MIT License