PyTorchFI is a runtime fault injector tool for PyTorch to simulate bit flips within the neural network. Check us out on PyPI here.
pip install pytorchfi Then in your project source files:
Download this repository into your project folder. Then in your project source files:
from src import PyTorchFI_Core
The documentation can be found at https://pytorchfi.github.io/docs/.
The main source code of PyTorchFI is held in
src, which carries both
All python code is formatted with black.
- Sarita V. Adve (UIUC)
- Neeraj Aggarwal (UIUC)
- Christopher W. Fletcher (UIUC)
- Siva Kumar Sastry Hari (NVIDIA)
- Abdulrahman Mahmoud (UIUC)
- Alex Nobbe (UIUC)