ruby/tensorflow_privacy
1
0
Fork 0
forked from 626_privacy/tensorflow_privacy
Code Pull requests Activity

Change dp_optimizer_vectorized_test to use more examples to test accuracy of standard deviation.

Browse source 
PiperOrigin-RevId: 267468611
This commit is contained in:
Steve Chien 2019-09-05 15:26:08 -07:00 committed by A. Unique TensorFlower
parent 0776ed124b
commit 543b624169
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
privacy/optimizers/dp_optimizer_vectorized_test.py
View file

@ -109,7 +109,7 @@ class DPOptimizerTest(tf.test.TestCase, parameterized.TestCase):
gradient_op = opt.compute_gradients(self._loss(data0, var0), [var0]) gradient_op = opt.compute_gradients(self._loss(data0, var0), [var0])
grads = [] grads = []
for _ in range(1000): for _ in range(5000):
grads_and_vars = sess.run(gradient_op) grads_and_vars = sess.run(gradient_op)
grads.append(grads_and_vars[0][0]) grads.append(grads_and_vars[0][0])