Doc format fixes.

PiperOrigin-RevId: 340494524
Change-Id: I7662470f5d63b860896b2845cd48a4abc00d5348
This commit is contained in:
Mark Daoust 2020-11-03 11:59:49 -08:00 committed by TensorFlower Gardener
parent d7a91161c0
commit 3113c71abd
2 changed files with 40 additions and 36 deletions

View File

@ -70,17 +70,17 @@ class DeviceSpecV2(object):
With eager execution disabled (by default in TensorFlow 1.x and by calling
disable_eager_execution() in TensorFlow 2.x), the following syntax
can be used:
```python
tf.compat.v1.disable_eager_execution()
# Same as previous
device_spec = DeviceSpec(job="ps", device_type="GPU", device_index=0)
# No need of .to_string() method.
with tf.device(device_spec):
my_var = tf.Variable(..., name="my_variable")
squared_var = tf.square(my_var)
```
```
If a `DeviceSpec` is partially specified, it will be merged with other
`DeviceSpec`s according to the scope in which it is defined. `DeviceSpec`
@ -160,39 +160,43 @@ class DeviceSpecV2(object):
def parse_from_string(self, spec):
"""Parse a `DeviceSpec` name into its components.
2.x behavior change:
In TensorFlow 1.x, this function mutates its own state and returns itself.
In 2.x, DeviceSpecs are immutable, and this function will return a
DeviceSpec which contains the spec.
**2.x behavior change**:
Recommended:
```
# my_spec and my_updated_spec are unrelated.
my_spec = tf.DeviceSpec.from_string("/CPU:0")
my_updated_spec = tf.DeviceSpec.from_string("/GPU:0")
with tf.device(my_updated_spec):
...
```
In TensorFlow 1.x, this function mutates its own state and returns itself.
In 2.x, DeviceSpecs are immutable, and this function will return a
DeviceSpec which contains the spec.
Will work in 1.x and 2.x (though deprecated in 2.x):
```
my_spec = tf.DeviceSpec.from_string("/CPU:0")
my_updated_spec = my_spec.parse_from_string("/GPU:0")
with tf.device(my_updated_spec):
...
```
* Recommended:
Will NOT work in 2.x:
```
my_spec = tf.DeviceSpec.from_string("/CPU:0")
my_spec.parse_from_string("/GPU:0") # <== Will not update my_spec
with tf.device(my_spec):
...
```
```
# my_spec and my_updated_spec are unrelated.
my_spec = tf.DeviceSpec.from_string("/CPU:0")
my_updated_spec = tf.DeviceSpec.from_string("/GPU:0")
with tf.device(my_updated_spec):
...
```
In general, `DeviceSpec.from_string` should completely replace
`DeviceSpec.parse_from_string`, and `DeviceSpec.replace` should
completely replace setting attributes directly.
* Will work in 1.x and 2.x (though deprecated in 2.x):
```
my_spec = tf.DeviceSpec.from_string("/CPU:0")
my_updated_spec = my_spec.parse_from_string("/GPU:0")
with tf.device(my_updated_spec):
...
```
* Will NOT work in 2.x:
```
my_spec = tf.DeviceSpec.from_string("/CPU:0")
my_spec.parse_from_string("/GPU:0") # <== Will not update my_spec
with tf.device(my_spec):
...
```
In general, `DeviceSpec.from_string` should completely replace
`DeviceSpec.parse_from_string`, and `DeviceSpec.replace` should
completely replace setting attributes directly.
Args:
spec: an optional string of the form

View File

@ -4003,7 +4003,7 @@ def batch_to_space_v2(input, block_shape, crops, name=None): # pylint: disable=
Examples:
(1) For the following input of shape `[4, 1, 1, 1]`,
1. For the following input of shape `[4, 1, 1, 1]`,
`block_shape = [2, 2]`, and `crops = [[0, 0], [0, 0]]`:
```python
@ -4020,7 +4020,7 @@ def batch_to_space_v2(input, block_shape, crops, name=None): # pylint: disable=
[[3], [4]]]]
```
(2) For the following input of shape `[4, 1, 1, 3]`,
2. For the following input of shape `[4, 1, 1, 3]`,
`block_shape = [2, 2]`, and `crops = [[0, 0], [0, 0]]`:
```python
@ -4037,7 +4037,7 @@ def batch_to_space_v2(input, block_shape, crops, name=None): # pylint: disable=
[[7, 8, 9], [10, 11, 12]]]]
```
(3) For the following
3. For the following
input of shape `[4, 2, 2, 1]`,
`block_shape = [2, 2]`, and `crops = [[0, 0], [0, 0]]`:
@ -4057,7 +4057,7 @@ def batch_to_space_v2(input, block_shape, crops, name=None): # pylint: disable=
[[13], [14], [15], [16]]]
```
(4) For the following input of shape
4. For the following input of shape
`[8, 1, 3, 1]`,
`block_shape = [2, 2]`, and `crops = [[0, 0], [2, 0]]`: