pytorch-dp: Train PyTorch models with Differential Privacy

pytorch-dp is a library that enables training PyTorch models with differential privacy. It supports training with minimal code changes required on the client, has little impact on training performance and allows the client to online track the privacy budget expended at any given moment.

PyTorch Privacy is currently a preview beta and under active development!

Target audience

This code release is aimed at two target audiences:

1. ML practicioners will find this code a gentle introduction to training a model with differential privacy as it requires minimal code changes.
2. Differential Privacy scientists will find this code easy to experiment and tinker with, allowing them to focus on what matters.

Installation

pip:

pip install pytorch-dp

From source:

git clone https://github.com/facebookresearch/pytorch-dp.git
cd pytorch-dp
pip install -e .

Getting started

To train your model with differential privacy, all you need to do is to declare a PrivacyEngine and attach it to your optimizer before running, eg:

model = Net()
optimizer = SGD(model.parameters(), lr=0.05)
privacy_engine = PrivacyEngine(
    model,
    batch_size,
    sample_size,
    alphas=[1, 10, 100],
    noise_multiplier=1.3,
    max_grad_norm=1.0,
)
privacy_engine.attach(optimizer)
# Now it's business as usual

The MNIST example contains an end to end run.

Environment

1. Create environment

conda env create --prefix="./.venv" --file=environment.yml

1. Activate environment

conda activate "./.venv"

1. Deactivate environment

conda deactivate

Contributing

See the CONTRIBUTING file for how to help out.

References

• Mironov, Ilya. "Rényi differential privacy." 2017 IEEE 30th Computer Security Foundations Symposium (CSF). IEEE, 2017.
• Abadi, Martin, et al. "Deep learning with differential privacy." Proceedings of the 2016 ACM SIGSAC Conference on Computer and Communications Security. ACM, 2016.
• Mironov, Ilya, Kunal Talwar, and Li Zhang. "R'enyi Differential Privacy of the Sampled Gaussian Mechanism." arXiv preprint arXiv:1908.10530 (2019).
• Goodfellow, Ian. "Efficient per-example gradient computations." arXiv preprint arXiv:1510.01799 (2015).
• McMahan, H. Brendan, and Galen Andrew. "A general approach to adding differential privacy to iterative training procedures." arXiv preprint arXiv:1812.06210 (2018).

License

This code is released under Apache 2.0, as found in the LICENSE file.