tf.keras.utils - LostTech.TensorFlow Documentation

void convert_all_kernels_in_model(object model)

Converts all convolution kernels in a model from Theano to TensorFlow.

Also works from TensorFlow to Theano.

Parameters

object model: target model for the conversion.

object convert_all_kernels_in_model_dyn(object model)

Converts all convolution kernels in a model from Theano to TensorFlow.

Also works from TensorFlow to Theano.

Parameters

object model: target model for the conversion.

CustomObjectScope custom_object_scope(Object[] args)

Provides a scope that changes to `_GLOBAL_CUSTOM_OBJECTS` cannot escape.

Convenience wrapper for `CustomObjectScope`. Code within a `with` statement will be able to access custom objects by name. Changes to global custom objects persist within the enclosing `with` statement. At end of the `with` statement, global custom objects are reverted to state at beginning of the `with` statement.

Example:

Consider a custom object `MyObject`

Parameters

Object[] args: Variable length list of dictionaries of name, class pairs to add to custom objects.

Returns

CustomObjectScope: Object of type `CustomObjectScope`.

Show Example

with custom_object_scope({'MyObject':MyObject}):
                layer = Dense(..., kernel_regularizer='MyObject')
                # save, load, etc. will recognize custom object by name

object custom_object_scope_dyn(Object[] args)

Provides a scope that changes to `_GLOBAL_CUSTOM_OBJECTS` cannot escape.

Convenience wrapper for `CustomObjectScope`. Code within a `with` statement will be able to access custom objects by name. Changes to global custom objects persist within the enclosing `with` statement. At end of the `with` statement, global custom objects are reverted to state at beginning of the `with` statement.

Example:

Consider a custom object `MyObject`

Parameters

Object[] args: Variable length list of dictionaries of name, class pairs to add to custom objects.

Returns

object: Object of type `CustomObjectScope`.

Show Example

with custom_object_scope({'MyObject':MyObject}):
                layer = Dense(..., kernel_regularizer='MyObject')
                # save, load, etc. will recognize custom object by name

object deserialize_keras_object(PythonClassContainer identifier, IDictionary<string, object> module_objects, IDictionary<string, object> custom_objects, string printable_module_name)

object deserialize_keras_object(object identifier, IDictionary<string, object> module_objects, IDictionary<string, object> custom_objects, string printable_module_name)

object deserialize_keras_object_dyn(object identifier, object module_objects, object custom_objects, ImplicitContainer<T> printable_module_name)

IDictionary<object, object> get_custom_objects()

Retrieves a live reference to the global dictionary of custom objects.

Updating and clearing custom objects using `custom_object_scope` is preferred, but `get_custom_objects` can be used to directly access `_GLOBAL_CUSTOM_OBJECTS`.

Example:

Returns

IDictionary<object, object>: Global dictionary of names to classes (`_GLOBAL_CUSTOM_OBJECTS`).

Show Example

get_custom_objects().clear()
            get_custom_objects()['MyObject'] = MyObject

object get_custom_objects_dyn()

Retrieves a live reference to the global dictionary of custom objects.

Updating and clearing custom objects using `custom_object_scope` is preferred, but `get_custom_objects` can be used to directly access `_GLOBAL_CUSTOM_OBJECTS`.

Example:

Returns

object: Global dictionary of names to classes (`_GLOBAL_CUSTOM_OBJECTS`).

Show Example

get_custom_objects().clear()
            get_custom_objects()['MyObject'] = MyObject

object get_file(string fname, string origin, bool untar, string md5_hash, string file_hash, string cache_subdir, string hash_algorithm, bool extract, string archive_format, object cache_dir)

Downloads a file from a URL if it not already in the cache.

By default the file at the url `origin` is downloaded to the cache_dir `~/.keras`, placed in the cache_subdir `datasets`, and given the filename `fname`. The final location of a file `example.txt` would therefore be `~/.keras/datasets/example.txt`.

Files in tar, tar.gz, tar.bz, and zip formats can also be extracted. Passing a hash will verify the file after download. The command line programs `shasum` and `sha256sum` can compute the hash.

Parameters

string fname: Name of the file. If an absolute path `/path/to/file.txt` is specified the file will be saved at that location.
string origin: Original URL of the file.
bool untar: Deprecated in favor of 'extract'. boolean, whether the file should be decompressed
string md5_hash: Deprecated in favor of 'file_hash'. md5 hash of the file for verification
string file_hash: The expected hash string of the file after download. The sha256 and md5 hash algorithms are both supported.
string cache_subdir: Subdirectory under the Keras cache dir where the file is saved. If an absolute path `/path/to/folder` is specified the file will be saved at that location.
string hash_algorithm: Select the hash algorithm to verify the file. options are 'md5', 'sha256', and 'auto'. The default 'auto' detects the hash algorithm in use.
bool extract: True tries extracting the file as an Archive, like tar or zip.
string archive_format: Archive format to try for extracting the file. Options are 'auto', 'tar', 'zip', and None. 'tar' includes tar, tar.gz, and tar.bz files. The default 'auto' is ['tar', 'zip']. None or an empty list will return no matches found.
object cache_dir: Location to store cached files, when None it defaults to the [Keras Directory](/faq/#where-is-the-keras-configuration-filed-stored).

Returns

object: Path to the downloaded file

object get_file_dyn(object fname, object origin, ImplicitContainer<T> untar, object md5_hash, object file_hash, ImplicitContainer<T> cache_subdir, ImplicitContainer<T> hash_algorithm, ImplicitContainer<T> extract, ImplicitContainer<T> archive_format, object cache_dir)

Downloads a file from a URL if it not already in the cache.

By default the file at the url `origin` is downloaded to the cache_dir `~/.keras`, placed in the cache_subdir `datasets`, and given the filename `fname`. The final location of a file `example.txt` would therefore be `~/.keras/datasets/example.txt`.

Files in tar, tar.gz, tar.bz, and zip formats can also be extracted. Passing a hash will verify the file after download. The command line programs `shasum` and `sha256sum` can compute the hash.

Parameters

object fname: Name of the file. If an absolute path `/path/to/file.txt` is specified the file will be saved at that location.
object origin: Original URL of the file.
ImplicitContainer<T> untar: Deprecated in favor of 'extract'. boolean, whether the file should be decompressed
object md5_hash: Deprecated in favor of 'file_hash'. md5 hash of the file for verification
object file_hash: The expected hash string of the file after download. The sha256 and md5 hash algorithms are both supported.
ImplicitContainer<T> cache_subdir: Subdirectory under the Keras cache dir where the file is saved. If an absolute path `/path/to/folder` is specified the file will be saved at that location.
ImplicitContainer<T> hash_algorithm: Select the hash algorithm to verify the file. options are 'md5', 'sha256', and 'auto'. The default 'auto' detects the hash algorithm in use.
ImplicitContainer<T> extract: True tries extracting the file as an Archive, like tar or zip.
ImplicitContainer<T> archive_format: Archive format to try for extracting the file. Options are 'auto', 'tar', 'zip', and None. 'tar' includes tar, tar.gz, and tar.bz files. The default 'auto' is ['tar', 'zip']. None or an empty list will return no matches found.
object cache_dir: Location to store cached files, when None it defaults to the [Keras Directory](/faq/#where-is-the-keras-configuration-filed-stored).

Returns

object: Path to the downloaded file

object get_source_inputs(IGraphNodeBase tensor, object layer, object node_index)

Returns the list of input tensors necessary to compute `tensor`.

Output will always be a list of tensors (potentially with 1 element).

Parameters

IGraphNodeBase tensor: The tensor to start from.
object layer: Origin layer of the tensor. Will be determined via tensor._keras_history if not provided.
object node_index: Origin node index of the tensor.

Returns

object: List of input tensors.

object get_source_inputs(IEnumerable<IGraphNodeBase> tensor, object layer, object node_index)

Returns the list of input tensors necessary to compute `tensor`.

Output will always be a list of tensors (potentially with 1 element).

Parameters

IEnumerable<IGraphNodeBase> tensor: The tensor to start from.
object layer: Origin layer of the tensor. Will be determined via tensor._keras_history if not provided.
object node_index: Origin node index of the tensor.

Returns

object: List of input tensors.

object get_source_inputs_dyn(object tensor, object layer, object node_index)

Returns the list of input tensors necessary to compute `tensor`.

Output will always be a list of tensors (potentially with 1 element).

Parameters

object tensor: The tensor to start from.
object layer: Origin layer of the tensor. Will be determined via tensor._keras_history if not provided.
object node_index: Origin node index of the tensor.

Returns

object: List of input tensors.

object model_to_dot(Layer model, bool show_shapes, bool show_layer_names, string rankdir, bool expand_nested, int dpi, bool subgraph)

Convert a Keras model to dot format.

Parameters

Layer model: A Keras model instance.
bool show_shapes: whether to display shape information.
bool show_layer_names: whether to display layer names.
string rankdir: `rankdir` argument passed to PyDot, a string specifying the format of the plot: 'TB' creates a vertical plot; 'LR' creates a horizontal plot.
bool expand_nested: whether to expand nested models into clusters.
int dpi: Dots per inch.
bool subgraph: whether to return a `pydot.Cluster` instance.

Returns

object: A `pydot.Dot` instance representing the Keras model or a `pydot.Cluster` instance representing nested model if `subgraph=True`.

object model_to_dot(Network model, bool show_shapes, bool show_layer_names, string rankdir, bool expand_nested, int dpi, bool subgraph)

Convert a Keras model to dot format.

Parameters

Network model: A Keras model instance.
bool show_shapes: whether to display shape information.
bool show_layer_names: whether to display layer names.
string rankdir: `rankdir` argument passed to PyDot, a string specifying the format of the plot: 'TB' creates a vertical plot; 'LR' creates a horizontal plot.
bool expand_nested: whether to expand nested models into clusters.
int dpi: Dots per inch.
bool subgraph: whether to return a `pydot.Cluster` instance.

Returns

object: A `pydot.Dot` instance representing the Keras model or a `pydot.Cluster` instance representing nested model if `subgraph=True`.

object model_to_dot_dyn(object model, ImplicitContainer<T> show_shapes, ImplicitContainer<T> show_layer_names, ImplicitContainer<T> rankdir, ImplicitContainer<T> expand_nested, ImplicitContainer<T> dpi, ImplicitContainer<T> subgraph)

Convert a Keras model to dot format.

Parameters

object model: A Keras model instance.
ImplicitContainer<T> show_shapes: whether to display shape information.
ImplicitContainer<T> show_layer_names: whether to display layer names.
ImplicitContainer<T> rankdir: `rankdir` argument passed to PyDot, a string specifying the format of the plot: 'TB' creates a vertical plot; 'LR' creates a horizontal plot.
ImplicitContainer<T> expand_nested: whether to expand nested models into clusters.
ImplicitContainer<T> dpi: Dots per inch.
ImplicitContainer<T> subgraph: whether to return a `pydot.Cluster` instance.

Returns

object: A `pydot.Dot` instance representing the Keras model or a `pydot.Cluster` instance representing nested model if `subgraph=True`.

Model multi_gpu_model(Model model, IEnumerable<int> gpus, bool cpu_merge, bool cpu_relocation)

Replicates a model on different GPUs.

Specifically, this function implements single-machine multi-GPU data parallelism. It works in the following way:

- Divide the model's input(s) into multiple sub-batches. - Apply a model copy on each sub-batch. Every model copy is executed on a dedicated GPU. - Concatenate the results (on CPU) into one big batch.

E.g. if your `batch_size` is 64 and you use `gpus=2`, then we will divide the input into 2 sub-batches of 32 samples, process each sub-batch on one GPU, then return the full batch of 64 processed samples.

This induces quasi-linear speedup on up to 8 GPUs.

This function is only available with the TensorFlow backend for the time being.

Parameters

Model model: A Keras model instance. To avoid OOM errors, this model could have been built on CPU, for instance (see usage example below).
IEnumerable<int> gpus: Integer >= 2, number of on GPUs on which to create model replicas.
bool cpu_merge: A boolean value to identify whether to force merging model weights under the scope of the CPU or not.
bool cpu_relocation: A boolean value to identify whether to create the model's weights under the scope of the CPU. If the model is not defined under any preceding device scope, you can still rescue it by activating this option.

Returns

Model

A Keras `Model` instance which can be used just like the initial `model` argument, but which distributes its workload on multiple GPUs.

Example 1: Training models with weights merge on CPU

```python import tensorflow as tf from keras.applications import Xception from keras.utils import multi_gpu_model import numpy as np

num_samples = 1000 height = 224 width = 224 num_classes = 1000

# Instantiate the base model (or "template" model). # We recommend doing this with under a CPU device scope, # so that the model's weights are hosted on CPU memory. # Otherwise they may end up hosted on a GPU, which would # complicate weight sharing. with tf.device('/cpu:0'): model = Xception(weights=None, input_shape=(height, width, 3), classes=num_classes)

# Replicates the model on 8 GPUs. # This assumes that your machine has 8 available GPUs. parallel_model = multi_gpu_model(model, gpus=8) parallel_model.compile(loss='categorical_crossentropy', optimizer='rmsprop')

# Generate dummy data. x = np.random.random((num_samples, height, width, 3)) y = np.random.random((num_samples, num_classes))

# This `fit` call will be distributed on 8 GPUs. # Since the batch size is 256, each GPU will process 32 samples. parallel_model.fit(x, y, epochs=20, batch_size=256)

# Save model via the template model (which shares the same weights): model.save('my_model.h5') ```

Example 2: Training models with weights merge on CPU using cpu_relocation

```python .. # Not needed to change the device scope for model definition: model = Xception(weights=None,..)

try: model = multi_gpu_model(model, cpu_relocation=True) print("Training using multiple GPUs..") except: print("Training using single GPU or CPU..")

model.compile(..) .. ```

Example 3: Training models with weights merge on GPU (recommended for NV-link)

```python .. # Not needed to change the device scope for model definition: model = Xception(weights=None,..)

try: model = multi_gpu_model(model, cpu_merge=False) print("Training using multiple GPUs..") except: print("Training using single GPU or CPU..") model.compile(..) .. ```

Model multi_gpu_model(Model model, ValueTuple<IEnumerable<object>, object> gpus, bool cpu_merge, bool cpu_relocation)

Replicates a model on different GPUs.

Specifically, this function implements single-machine multi-GPU data parallelism. It works in the following way:

- Divide the model's input(s) into multiple sub-batches. - Apply a model copy on each sub-batch. Every model copy is executed on a dedicated GPU. - Concatenate the results (on CPU) into one big batch.

E.g. if your `batch_size` is 64 and you use `gpus=2`, then we will divide the input into 2 sub-batches of 32 samples, process each sub-batch on one GPU, then return the full batch of 64 processed samples.

This induces quasi-linear speedup on up to 8 GPUs.

This function is only available with the TensorFlow backend for the time being.

Parameters

Model model: A Keras model instance. To avoid OOM errors, this model could have been built on CPU, for instance (see usage example below).
ValueTuple<IEnumerable<object>, object> gpus: Integer >= 2, number of on GPUs on which to create model replicas.
bool cpu_merge: A boolean value to identify whether to force merging model weights under the scope of the CPU or not.
bool cpu_relocation: A boolean value to identify whether to create the model's weights under the scope of the CPU. If the model is not defined under any preceding device scope, you can still rescue it by activating this option.