626_privacy/tensorflow_privacy
2
0
Fork 1
Code Issues Pull requests 1 Projects Releases Packages Wiki Activity Actions

[tfds] Fix tfds.as_numpy which now returns a reusable iterable.

Browse source 
`next(ds)` -> `next(iter(ds))`

Context: https://github.com/tensorflow/datasets/issues/2270
PiperOrigin-RevId: 324951455
This commit is contained in:
A. Unique TensorFlower 2020-08-04 21:24:33 -07:00 committed by Steve Chien
parent 29651216cd
commit 191f2461c5
1 changed files with 2 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
tutorials/lm_dpsgd_tutorial.py
View file

@ -139,8 +139,8 @@ def load_data():
test_dataset = tfds.load(name='lm1b/subwords8k', test_dataset = tfds.load(name='lm1b/subwords8k',
split=tfds.Split.TEST, split=tfds.Split.TEST,
batch_size=10000) batch_size=10000)
train_data = next(tfds.as_numpy(train_dataset)) train_data = next(iter(tfds.as_numpy(train_dataset)))
test_data = next(tfds.as_numpy(test_dataset)) test_data = next(iter(tfds.as_numpy(test_dataset)))
train_data = train_data['text'].flatten() train_data = train_data['text'].flatten()
test_data = test_data['text'].flatten() test_data = test_data['text'].flatten()
else: else: