LostTech.TensorFlow : API Documentation

Type tf.graph_util

Namespace tensorflow

Methods

Properties

Public static methods

object convert_variables_to_constants(BaseSession sess, Nullable<int> input_graph_def, IEnumerable<object> output_node_names, string variable_names_whitelist, IEnumerable<string> variable_names_blacklist)

Replaces all the variables in a graph with constants of the same values. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.convert_variables_to_constants`

If you have a trained graph containing Variable ops, it can be convenient to convert them all to Const ops holding the same values. This makes it possible to describe the network fully with a single GraphDef file, and allows the removal of a lot of ops related to loading and saving the variables.
Parameters
BaseSession sess
Active TensorFlow session containing the variables.
Nullable<int> input_graph_def
GraphDef object holding the network.
IEnumerable<object> output_node_names
List of name strings for the result nodes of the graph.
string variable_names_whitelist
The set of variable names to convert (by default, all variables are converted).
IEnumerable<string> variable_names_blacklist
The set of variable names to omit converting to constants.
Returns
object
GraphDef containing a simplified version of the original.

object convert_variables_to_constants(BaseSession sess, Nullable<int> input_graph_def, IEnumerable<object> output_node_names, string variable_names_whitelist, string variable_names_blacklist)

Replaces all the variables in a graph with constants of the same values. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.convert_variables_to_constants`

If you have a trained graph containing Variable ops, it can be convenient to convert them all to Const ops holding the same values. This makes it possible to describe the network fully with a single GraphDef file, and allows the removal of a lot of ops related to loading and saving the variables.
Parameters
BaseSession sess
Active TensorFlow session containing the variables.
Nullable<int> input_graph_def
GraphDef object holding the network.
IEnumerable<object> output_node_names
List of name strings for the result nodes of the graph.
string variable_names_whitelist
The set of variable names to convert (by default, all variables are converted).
string variable_names_blacklist
The set of variable names to omit converting to constants.
Returns
object
GraphDef containing a simplified version of the original.

object convert_variables_to_constants(BaseSession sess, Nullable<int> input_graph_def, IEnumerable<object> output_node_names, IEnumerable<string> variable_names_whitelist, IEnumerable<string> variable_names_blacklist)

Replaces all the variables in a graph with constants of the same values. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.convert_variables_to_constants`

If you have a trained graph containing Variable ops, it can be convenient to convert them all to Const ops holding the same values. This makes it possible to describe the network fully with a single GraphDef file, and allows the removal of a lot of ops related to loading and saving the variables.
Parameters
BaseSession sess
Active TensorFlow session containing the variables.
Nullable<int> input_graph_def
GraphDef object holding the network.
IEnumerable<object> output_node_names
List of name strings for the result nodes of the graph.
IEnumerable<string> variable_names_whitelist
The set of variable names to convert (by default, all variables are converted).
IEnumerable<string> variable_names_blacklist
The set of variable names to omit converting to constants.
Returns
object
GraphDef containing a simplified version of the original.

object convert_variables_to_constants(BaseSession sess, Nullable<int> input_graph_def, IEnumerable<object> output_node_names, IEnumerable<string> variable_names_whitelist, string variable_names_blacklist)

Replaces all the variables in a graph with constants of the same values. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.convert_variables_to_constants`

If you have a trained graph containing Variable ops, it can be convenient to convert them all to Const ops holding the same values. This makes it possible to describe the network fully with a single GraphDef file, and allows the removal of a lot of ops related to loading and saving the variables.
Parameters
BaseSession sess
Active TensorFlow session containing the variables.
Nullable<int> input_graph_def
GraphDef object holding the network.
IEnumerable<object> output_node_names
List of name strings for the result nodes of the graph.
IEnumerable<string> variable_names_whitelist
The set of variable names to convert (by default, all variables are converted).
string variable_names_blacklist
The set of variable names to omit converting to constants.
Returns
object
GraphDef containing a simplified version of the original.

object convert_variables_to_constants_dyn(object sess, object input_graph_def, object output_node_names, object variable_names_whitelist, object variable_names_blacklist)

Replaces all the variables in a graph with constants of the same values. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.convert_variables_to_constants`

If you have a trained graph containing Variable ops, it can be convenient to convert them all to Const ops holding the same values. This makes it possible to describe the network fully with a single GraphDef file, and allows the removal of a lot of ops related to loading and saving the variables.
Parameters
object sess
Active TensorFlow session containing the variables.
object input_graph_def
GraphDef object holding the network.
object output_node_names
List of name strings for the result nodes of the graph.
object variable_names_whitelist
The set of variable names to convert (by default, all variables are converted).
object variable_names_blacklist
The set of variable names to omit converting to constants.
Returns
object
GraphDef containing a simplified version of the original.

object extract_sub_graph(Nullable<int> graph_def, IEnumerable<string> dest_nodes)

Extract the subgraph that can reach any of the nodes in 'dest_nodes'. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.extract_sub_graph`
Parameters
Nullable<int> graph_def
A graph_pb2.GraphDef proto.
IEnumerable<string> dest_nodes
A list of strings specifying the destination node names.
Returns
object
The GraphDef of the sub-graph.

object extract_sub_graph(Nullable<int> graph_def, object dest_nodes)

Extract the subgraph that can reach any of the nodes in 'dest_nodes'. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.extract_sub_graph`
Parameters
Nullable<int> graph_def
A graph_pb2.GraphDef proto.
object dest_nodes
A list of strings specifying the destination node names.
Returns
object
The GraphDef of the sub-graph.

object extract_sub_graph(Nullable<int> graph_def, string dest_nodes)

Extract the subgraph that can reach any of the nodes in 'dest_nodes'. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.extract_sub_graph`
Parameters
Nullable<int> graph_def
A graph_pb2.GraphDef proto.
string dest_nodes
A list of strings specifying the destination node names.
Returns
object
The GraphDef of the sub-graph.

object extract_sub_graph_dyn(object graph_def, object dest_nodes)

Extract the subgraph that can reach any of the nodes in 'dest_nodes'. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.extract_sub_graph`
Parameters
object graph_def
A graph_pb2.GraphDef proto.
object dest_nodes
A list of strings specifying the destination node names.
Returns
object
The GraphDef of the sub-graph.

bool must_run_on_cpu(Operation node, bool pin_variables_on_cpu)

Returns True if the given node_def must run on CPU, otherwise False. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.must_run_on_cpu`
Parameters
Operation node
The node to be assigned to a device. Could be either an ops.Operation or NodeDef.
bool pin_variables_on_cpu
If True, this function will return False if node_def represents a variable-related op.
Returns
bool
True if the given node must run on CPU, otherwise False.

bool must_run_on_cpu(object node, bool pin_variables_on_cpu)

Returns True if the given node_def must run on CPU, otherwise False. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.must_run_on_cpu`
Parameters
object node
The node to be assigned to a device. Could be either an ops.Operation or NodeDef.
bool pin_variables_on_cpu
If True, this function will return False if node_def represents a variable-related op.
Returns
bool
True if the given node must run on CPU, otherwise False.

object must_run_on_cpu_dyn(object node, ImplicitContainer<T> pin_variables_on_cpu)

Returns True if the given node_def must run on CPU, otherwise False. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.must_run_on_cpu`
Parameters
object node
The node to be assigned to a device. Could be either an ops.Operation or NodeDef.
ImplicitContainer<T> pin_variables_on_cpu
If True, this function will return False if node_def represents a variable-related op.
Returns
object
True if the given node must run on CPU, otherwise False.

object remove_training_nodes(object input_graph, IEnumerable<string> protected_nodes)

Prunes out nodes that aren't needed for inference. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.remove_training_nodes`

There are nodes like Identity and CheckNumerics that are only useful during training, and can be removed in graphs that will be used for nothing but inference. Here we identify and remove them, returning an equivalent graph. To be specific, CheckNumerics nodes are always removed, and Identity nodes that aren't involved in control edges are spliced out so that their input and outputs are directly connected.
Parameters
object input_graph
Model to analyze and prune.
IEnumerable<string> protected_nodes
An optional list of names of nodes to be kept unconditionally. This is for example useful to preserve Identity output nodes.
Returns
object
A list of nodes with the unnecessary ones removed.

object remove_training_nodes_dyn(object input_graph, object protected_nodes)

Prunes out nodes that aren't needed for inference. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.remove_training_nodes`

There are nodes like Identity and CheckNumerics that are only useful during training, and can be removed in graphs that will be used for nothing but inference. Here we identify and remove them, returning an equivalent graph. To be specific, CheckNumerics nodes are always removed, and Identity nodes that aren't involved in control edges are spliced out so that their input and outputs are directly connected.
Parameters
object input_graph
Model to analyze and prune.
object protected_nodes
An optional list of names of nodes to be kept unconditionally. This is for example useful to preserve Identity output nodes.
Returns
object
A list of nodes with the unnecessary ones removed.

string tensor_shape_from_node_def_name(object graph, object input_name)

Convenience function to get a shape from a NodeDef's input string. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.tensor_shape_from_node_def_name`

object tensor_shape_from_node_def_name_dyn(object graph, object input_name)

Convenience function to get a shape from a NodeDef's input string. (deprecated)

Warning: THIS FUNCTION IS DEPRECATED. It will be removed in a future version. Instructions for updating: Use `tf.compat.v1.graph_util.tensor_shape_from_node_def_name`

Public properties

PythonFunctionContainer convert_variables_to_constants_fn get;

PythonFunctionContainer extract_sub_graph_fn get;

PythonFunctionContainer must_run_on_cpu_fn get;

PythonFunctionContainer remove_training_nodes_fn get;

PythonFunctionContainer tensor_shape_from_node_def_name_fn get;