tf.keras.initializers.Orthogonal
Stay organized with collections
Save and categorize content based on your preferences.
Initializer that generates an orthogonal matrix.
Inherits From: Initializer
tf.keras.initializers.Orthogonal(
gain=1.0, seed=None
)
If the shape of the tensor to initialize is two-dimensional, it is initialized
with an orthogonal matrix obtained from the QR decomposition of a matrix of
random numbers drawn from a normal distribution.
If the matrix has fewer rows than columns then the output will have orthogonal
rows. Otherwise, the output will have orthogonal columns.
If the shape of the tensor to initialize is more than two-dimensional,
a matrix of shape (shape[0] * ... * shape[n - 2], shape[n - 1])
is initialized, where n is the length of the shape vector.
The matrix is subsequently reshaped to give a tensor of the desired shape.
Args |
|---|
gain
|
multiplicative factor to apply to the orthogonal matrix
|
seed
|
A Python integer. Used to create random seeds. See
tf.compat.v1.set_random_seed
for behavior.
|
References:
Saxe et al., 2014
(pdf)
Methods
from_config
View source
@classmethod
from_config(
config
)
Instantiates an initializer from a configuration dictionary.
Example:
initializer = RandomUniform(-1, 1)
config = initializer.get_config()
initializer = RandomUniform.from_config(config)
| Args |
|---|
config
|
A Python dictionary.
It will typically be the output of get_config.
|
| Returns |
|---|
|
An Initializer instance.
|
get_config
View source
get_config()
Returns the configuration of the initializer as a JSON-serializable dict.
| Returns |
|---|
|
A JSON-serializable Python dict.
|
__call__
View source
__call__(
shape, dtype=tf.dtypes.float32
)
Returns a tensor object initialized as specified by the initializer.
| Args |
|---|
shape
|
Shape of the tensor.
|
dtype
|
Optional dtype of the tensor. Only floating point types are
supported.
|
| Raises |
|---|
ValueError
|
If the dtype is not floating point or the input shape is not
valid.
|
Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.
Last updated 2020-10-01 UTC.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2020-10-01 UTC."],[],[],null,["# tf.keras.initializers.Orthogonal\n\n\u003cbr /\u003e\n\n|------------------------------------------------------------------------------------------------------------------------------|\n| [View source on GitHub](https://github.com/tensorflow/tensorflow/blob/v2.0.0/tensorflow/python/ops/init_ops_v2.py#L448-L515) |\n\nInitializer that generates an orthogonal matrix.\n\nInherits From: [`Initializer`](../../../tf/keras/initializers/Initializer)\n\n#### View aliases\n\n\n**Main aliases**\n\n[`tf.initializers.Orthogonal`](/api_docs/python/tf/keras/initializers/Orthogonal), [`tf.initializers.orthogonal`](/api_docs/python/tf/keras/initializers/Orthogonal), [`tf.keras.initializers.orthogonal`](/api_docs/python/tf/keras/initializers/Orthogonal)\n\n\u003cbr /\u003e\n\n tf.keras.initializers.Orthogonal(\n gain=1.0, seed=None\n )\n\nIf the shape of the tensor to initialize is two-dimensional, it is initialized\nwith an orthogonal matrix obtained from the QR decomposition of a matrix of\nrandom numbers drawn from a normal distribution.\nIf the matrix has fewer rows than columns then the output will have orthogonal\nrows. Otherwise, the output will have orthogonal columns.\n\nIf the shape of the tensor to initialize is more than two-dimensional,\na matrix of shape `(shape[0] * ... * shape[n - 2], shape[n - 1])`\nis initialized, where `n` is the length of the shape vector.\nThe matrix is subsequently reshaped to give a tensor of the desired shape.\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Args ---- ||\n|--------|------------------------------------------------------------------------------------------------------------------------------------------|\n| `gain` | multiplicative factor to apply to the orthogonal matrix |\n| `seed` | A Python integer. Used to create random seeds. See [`tf.compat.v1.set_random_seed`](../../../tf/compat/v1/set_random_seed) for behavior. |\n\n\u003cbr /\u003e\n\n#### References:\n\n[Saxe et al., 2014](https://openreview.net/forum?id=_wzZwKpTDF_9C)\n([pdf](https://arxiv.org/pdf/1312.6120.pdf))\n\nMethods\n-------\n\n### `from_config`\n\n[View source](https://github.com/tensorflow/tensorflow/blob/v2.0.0/tensorflow/python/ops/init_ops_v2.py#L69-L89) \n\n @classmethod\n from_config(\n config\n )\n\nInstantiates an initializer from a configuration dictionary.\n\n#### Example:\n\n initializer = RandomUniform(-1, 1)\n config = initializer.get_config()\n initializer = RandomUniform.from_config(config)\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Args ||\n|----------|-----------------------------------------------------------------------|\n| `config` | A Python dictionary. It will typically be the output of `get_config`. |\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Returns ||\n|---|---|\n| An Initializer instance. ||\n\n\u003cbr /\u003e\n\n### `get_config`\n\n[View source](https://github.com/tensorflow/tensorflow/blob/v2.0.0/tensorflow/python/ops/init_ops_v2.py#L514-L515) \n\n get_config()\n\nReturns the configuration of the initializer as a JSON-serializable dict.\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Returns ||\n|---|---|\n| A JSON-serializable Python dict. ||\n\n\u003cbr /\u003e\n\n### `__call__`\n\n[View source](https://github.com/tensorflow/tensorflow/blob/v2.0.0/tensorflow/python/ops/init_ops_v2.py#L478-L512) \n\n __call__(\n shape, dtype=tf.dtypes.float32\n )\n\nReturns a tensor object initialized as specified by the initializer.\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Args ||\n|---------|------------------------------------------------------------------------|\n| `shape` | Shape of the tensor. |\n| `dtype` | Optional dtype of the tensor. Only floating point types are supported. |\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n\u003cbr /\u003e\n\n| Raises ||\n|--------------|---------------------------------------------------------------------|\n| `ValueError` | If the dtype is not floating point or the input shape is not valid. |\n\n\u003cbr /\u003e"]]
View source on GitHub