Keras distribute strategy: Simple fix the styling in docstring.
PiperOrigin-RevId: 336537385
Change-Id: Iac943f023805a0b535afc48b832eb18cf56b9add
diff --git a/tensorflow/python/distribute/distribute_lib.py b/tensorflow/python/distribute/distribute_lib.py
index 11c50fb..c1cdeed 100644
--- a/tensorflow/python/distribute/distribute_lib.py
+++ b/tensorflow/python/distribute/distribute_lib.py
@@ -1087,7 +1087,7 @@
This method can be used for several purposes. First, it allows you to
specify your own batching and sharding logic. (In contrast,
`tf.distribute.experimental_distribute_dataset` does batching and sharding
- for you.)For example, where
+ for you.) For example, where
`experimental_distribute_dataset` is unable to shard the input files, this
method might be used to manually shard the dataset (avoiding the slow
fallback behavior in `experimental_distribute_dataset`). In cases where the
