Layer Class Reference

Inheritance diagram for Layer:

Collaboration diagram for Layer:

Public Member Functions
	__init__ (self, name, tf_activation_str, dim_input, dim_output, p_bias, p_w, random_seed=None)
	__call__ (self, x)
	variable_to_summary (self, var, step, writer)
	all_to_summary (self, step, writer)

Public Attributes
	tf_activation = tf_activation_dict[tf_activation_str]
	activation function
list	shape = [dim_input, dim_output]
	layer shape
	w = self._init_weight(self.shape, p_w, random_seed)
	init parameters for uniform distribution
	b = self._init_bias(self.shape[1], p_bias)
	init parameters for bias
	input = None
	input
	output = None
	output

Protected Member Functions
	_init_bias (self, width, init_val, name=None)
	_init_weight (self, shape, stddev, operation_seed, name=None)

Detailed Description

definition of a layer obj

Definition at line 21 of file tensorflow_dnn_model.py.

Constructor & Destructor Documentation

__init__()

__init__	(		self,
			name,
			tf_activation_str,
			dim_input,
			dim_output,
			p_bias,
			p_w,
			random_seed = None )

:param name: name of the layer.
:param tf_activation: string, name of an available tensorflow activations function
:param dim_input: dimension of the input
:param dim_output: dimension of the output
:param p_bias: initial bias
:param p_w: stddev of uniform distribution to initialize
:param random_seed: random seed used in initialising the weights
:return: None

Definition at line 26 of file tensorflow_dnn_model.py.

                 random_seed=None):
        """
        :param name: name of the layer.
        :param tf_activation: string, name of an available tensorflow activations function
        :param dim_input: dimension of the input
        :param dim_output: dimension of the output
        :param p_bias: initial bias
        :param p_w: stddev of uniform distribution to initialize
        :param random_seed: random seed used in initialising the weights
        :return: None
        """
        
        super().__init__(name=name)
 
        tf_activation_dict = {
            'tanh': tf.nn.tanh,
            'sigmoid': tf.nn.sigmoid,
            'relu': tf.nn.relu,
            'leaky_relu': tf.nn.leaky_relu,
        }
 
        if tf_activation_str not in tf_activation_dict:
            raise ValueError
 
        
        self.tf_activation = tf_activation_dict[tf_activation_str]
 
        
        self.shape = [dim_input, dim_output]
 
        
        self.w = self._init_weight(self.shape, p_w, random_seed)
 
        
        self.b = self._init_bias(self.shape[1], p_bias)
 
        
        self.input = None
 
        
        self.output = None
 

Member Function Documentation

__call__()

__call__	(		self,
			x )

evaluate the layer

Definition at line 88 of file tensorflow_dnn_model.py.

    def __call__(self, x):
        """
        evaluate the layer
        """
        return self.tf_activation(tf.matmul(x, self.w) + self.b)
 

_init_bias()

_init_bias ( self, width, init_val, name = None )

protected

define bias variables

Definition at line 69 of file tensorflow_dnn_model.py.

    def _init_bias(self, width, init_val, name=None):
        """
        define bias variables
        """
        if name is None:
            name = self.name + '_b'
        initial = tf.constant(init_val, shape=[width], name=name)
        return tf.Variable(initial, name=name, trainable=True)
 

_init_weight()

_init_weight ( self, shape, stddev, operation_seed, name = None )

protected

define weight variables

Definition at line 78 of file tensorflow_dnn_model.py.

    def _init_weight(self, shape, stddev, operation_seed, name=None):
        """
        define weight variables
        """
        if name is None:
            name = self.name + '_w'
        initial = tf.random.truncated_normal(shape, stddev=stddev, seed=operation_seed, name=name)
        return tf.Variable(initial, name=name, trainable=True)
 

all_to_summary()

all_to_summary	(		self,
			step,
			writer )

Passes all layer variables to the tf.summary writer.

Definition at line 109 of file tensorflow_dnn_model.py.

    def all_to_summary(self, step, writer):
        """
        Passes all layer variables to the tf.summary writer.
        """
        self.variable_to_summary(self.w, step=step, writer=writer)
        self.variable_to_summary(self.b, step=step, writer=writer)
        return
 
 

variable_to_summary()

variable_to_summary	(		self,
			var,
			step,
			writer )

Passes information about each variable to the summary writer.

Definition at line 94 of file tensorflow_dnn_model.py.

    def variable_to_summary(self, var, step, writer):
        """
        Passes information about each variable to the summary writer.
        """
        with writer.as_default():
            mean = tf.reduce_mean(var)
            stddev = tf.sqrt(tf.reduce_mean(tf.square(var - mean)))
            tf.summary.scalar(f'{var.name}_mean', mean, step=step)
            tf.summary.scalar(f'{var.name}_stddev', stddev, step=step)
            tf.summary.scalar(f'{var.name}_max', tf.reduce_max(var), step=step)
            tf.summary.scalar(f'{var.name}_min', tf.reduce_min(var), step=step)
            tf.summary.histogram(f'{var.name}_histogram', var, step=step)
            writer.flush()
        return
 

Member Data Documentation

b

b = self._init_bias(self.shape[1], p_bias)

init parameters for bias

Definition at line 61 of file tensorflow_dnn_model.py.

input

input = None

input

Definition at line 64 of file tensorflow_dnn_model.py.

output

output = None

output

Definition at line 67 of file tensorflow_dnn_model.py.

shape

list shape = [dim_input, dim_output]

layer shape

Definition at line 55 of file tensorflow_dnn_model.py.

tf_activation

tf_activation = tf_activation_dict[tf_activation_str]

activation function

Definition at line 52 of file tensorflow_dnn_model.py.

w

w = self._init_weight(self.shape, p_w, random_seed)

init parameters for uniform distribution

Definition at line 58 of file tensorflow_dnn_model.py.

---

The documentation for this class was generated from the following file:

• analysis/scripts/dft/tensorflow_dnn_model.py