5 ms·
Jupyter Notebook Tutorial – Privacy-Preserving Deep Learning in PyTorch
- williamtrask 8y agoBased on PySyft: https://github.com/OpenMined/PySyft https://github.com/OpenMined/PySyft Paper: https://arxiv.org/abs/1811.04017.pdf https://arxiv.org/abs/1811.04017.pdf
- brandoncarl 8y agoIt's fantastic to see great research combined with great tutorials – bravo. You all have bitten off an incredibly hard problem with pragmatism and tenacity. Keep up the great work.
- williamtrask 8y agoThank you!
- themiratrix 8y ago
- Sol- 8y agoMy first thought was that this complements differentially private systems nicely, but seems like it was your intention to use it as a building block for things like that anyway. Very cool work! I think it's really important to have such secure primitives that are easy to build upon, maybe secure multiparty computation and such will finally become more widespread.
- williamtrask 8y agoThank you so much!