...
Make sure that all files in the same directory:
- models.py
- sg_ad.ipynb
- util.py
Make sure you have the following libraries installed:
numpy
torch
torchvision
scipy
tqdm
matplotlib
pandas
start executing the sg_ad.ipynb
notebook. Before running the detection mechanism, ensure you upload your collected gradients to the following path from the cell under the header 1) Loading FSHA and honest gradients
:
drive_path = '/content/drive/MyDrive/grads'
Continue with executing the notebook from the cell under the header 3) Converting Tensor gradients to NumPy Array
.
Adjust the following parameters in the notebook to fit your needs from 4) Selecting desired reduced data rate
and continue executing all cells until the end of the notebook.
- data_rate_honest_gradients
- window_size_list