Commit graph

21 commits

Author SHA1 Message Date
Steve Chien
f37c9d1ea1 Use mean loss within each microbatch.
PiperOrigin-RevId: 233832864
2019-02-13 14:42:12 -08:00
Steve Chien
d75f1b80ba Fix copybara to limit some transformations to the beginning of the line.
PiperOrigin-RevId: 233151293
2019-02-08 16:55:57 -08:00
A. Unique TensorFlower
4d0ab48c35 Add privacy ledger.
The privacy ledger keeps a record of all sampling and query events for analysis post hoc by the privacy accountant.

PiperOrigin-RevId: 233094012
2019-02-08 11:21:43 -08:00
A. Unique TensorFlower
36d9959c19 internal change
PiperOrigin-RevId: 233093203
2019-02-08 11:17:09 -08:00
A. Unique TensorFlower
098c5220b5 Remove test broken by upstream tf changes.
tf.nest.map_structure_up_to has changed so that map_structure_up_to(x, func, x, y) no longer raises an error when y is longer than x, for example x=[1,2], y=[1,2,3]. This broke one of our tests for nested query. Remove the test until (if and when) the old, more reasonable, behavior is restored.

PiperOrigin-RevId: 232057385
2019-02-01 16:22:05 -08:00
Galen Andrew
c8cb3c6b70 General cleanup.
1. Rename PrivateQuery to DPQuery.
2. Move construction of DPQuery to outside of optimizer.
3. Remove PrivateAverageQuery and PrivateSumQuery, and rename DPQuery's 'get_query_result' method to 'get_noised_result'. Rename private_queries.py to dp_query.py.
4. Remove thrice-replicated run_query function from the test classes and replace with a single function in new test_utils.py.
5. Add functions gaussian_sum_query_from_noise_multplier and gaussian_average_query_from_noise_multplier.

PiperOrigin-RevId: 230595991
2019-01-23 14:41:44 -08:00
Steve Chien
7e2d796bde Minor fixes for Python 2/3 compatibility.
PiperOrigin-RevId: 230022543
2019-01-18 17:33:22 -08:00
Peter Kairouz
0b56f7c016 Add optional argument for weighted sum and weighted average queries.
PiperOrigin-RevId: 230021515
2019-01-18 17:25:58 -08:00
Alex Pine
6c5c39c4f2 Created the optional unroll_microbatches parameter for the DpOptimizerClass as a workaround for b/122613513.
PiperOrigin-RevId: 229955297
2019-01-18 11:51:44 -08:00
Peter Kairouz
5ee12803f3 Create NoPrivacySumQuery and NoPrivacyAverageQuery.
PiperOrigin-RevId: 229273971
2019-01-14 17:39:36 -08:00
Nicolas Papernot
398d1d052f Closes #4
PiperOrigin-RevId: 229212766
2019-01-14 10:55:15 -08:00
Steve Chien
c30f6d776e Add test to ensure DP optimizers work with tf.estimator Estimators.
PiperOrigin-RevId: 228920704
2019-01-11 15:50:56 -08:00
schien
251d6298c6 Fix Python 3 compatibility issues
PiperOrigin-RevId: 228232503
2019-01-07 14:09:31 -08:00
Steve Chien
1689cf3a77
Revert "Fix Python 3 compatibility issues" 2019-01-07 13:44:35 -08:00
cclauss
d72378f913 Fix Python 3 compatibility issues 2018-12-24 03:55:12 +01:00
A. Unique TensorFlower
b4188446e0 Project import generated by Copybara.
PiperOrigin-RevId: 226345615
2018-12-20 09:17:59 -08:00
Steve Chien
1595ed3cd1 Project import generated by Copybara.
PiperOrigin-RevId: 226056146
2018-12-18 15:44:04 -08:00
A. Unique TensorFlower
ceee90b1ac Add GaussianSumQuery and express GaussianAverageQuery in terms of it.
Also:
1. Add unit tests for both types of query.
2. Add function "get_query_result" to PrivateQuery. (The utility of having this function is made clear in the test class, where the function _run_query operates on either GaussianSum- or GaussianAverageQueries.)
PiperOrigin-RevId: 225609398
2018-12-18 15:41:38 -08:00
Steve Chien
0af76c7b3d Update to allow bazel on tensorflow_privacy to work out of the box.
PiperOrigin-RevId: 225605386
2018-12-18 15:41:26 -08:00
Steve Chien
b8418b0523 PiperOrigin-RevId: 224078477 2018-12-05 14:03:22 -08:00
Steve Chien
afb8189dba PiperOrigin-RevId: 224061027 2018-12-04 17:01:39 -08:00