v0.1.28

Full Changelog: v0.1.27...v0.1.28

v0.1.27

Full Changelog: v0.1.26...v0.1.27

v0.1.26

Full Changelog: v0.1.25...v0.1.26

v0.1.25

Highlights:

• Fix multi-GPU sensitivity analysis.

Full Changelog: v0.1.24...v0.1.25

v0.1.24

Highlights:

• Tighter bounds on DP sensitivity in tf_shell_ml.
• Better scaling factor tracking.

Errata:

• Convolutional models sometimes fail during setup.

Full Changelog: v0.1.23...v0.1.24

v0.1.23

Full Changelog: v0.1.22...v0.1.23

v0.1.22

Highlights:

• Randomized rounding
• tf_shell_ml layer fixes for multi-GPU training.

Full Changelog: v0.1.21...v0.1.22

v0.1.21 Release

Highlights:

• Workaround tensorflow to fix convolution models.
• Fix: DG scale.

Full Changelog: v0.1.19...v0.1.21

v0.1.19

Full Changelog: v0.1.18...v0.1.19

v0.1.18

Highlights:

• Support sending large (>4GB) tensors between parties.

Full Changelog: v0.1.17...v0.1.18j