ztlearn.dl.layers package

Submodules

ztlearn.dl.layers.base module

class ztlearn.dl.layers.base.Layer(layer_name='zeta_squential')[source]

Bases: abc.ABC

input_shape
layer_name
layer_parameters
output_shape
pass_backward()[source]
pass_forward()[source]

ztlearn.dl.layers.convolutional module

class ztlearn.dl.layers.convolutional.Conv(filters=32, kernel_size=(3, 3), activation=None, input_shape=(1, 8, 8), strides=(1, 1), padding='valid')[source]

Bases: ztlearn.dl.layers.base.Layer

layer_activation
layer_parameters
output_shape
prep_layer()[source]
trainable
weight_initializer
weight_optimizer
class ztlearn.dl.layers.convolutional.Conv2D(filters=32, kernel_size=(3, 3), activation=None, input_shape=(1, 8, 8), strides=(1, 1), padding='valid')[source]

Bases: ztlearn.dl.layers.convolutional.Conv

pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
class ztlearn.dl.layers.convolutional.ConvLoop2D(filters=32, kernel_size=(3, 3), activation=None, input_shape=(1, 8, 8), strides=(1, 1), padding='valid')[source]

Bases: ztlearn.dl.layers.convolutional.Conv

pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
class ztlearn.dl.layers.convolutional.ConvToeplitzMat(filters=32, kernel_size=(3, 3), activation=None, input_shape=(1, 8, 8), strides=(1, 1), padding='valid')[source]

Bases: ztlearn.dl.layers.convolutional.Conv

pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]

ztlearn.dl.layers.core module

class ztlearn.dl.layers.core.Activation(function_name, input_shape=None, **kwargs)[source]

Bases: ztlearn.dl.layers.base.Layer

layer_name
output_shape
pass_backward(grad)[source]
pass_forward(input_signal, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable
class ztlearn.dl.layers.core.Dense(units, activation=None, input_shape=None)[source]

Bases: ztlearn.dl.layers.base.Layer

layer_activation
layer_parameters
output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True)[source]
prep_layer()[source]
trainable
weight_initializer
weight_optimizer
class ztlearn.dl.layers.core.Dropout(drop=0.5)[source]

Bases: ztlearn.dl.layers.base.Layer

output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable
class ztlearn.dl.layers.core.Flatten(input_shape=None)[source]

Bases: ztlearn.dl.layers.base.Layer

output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable
class ztlearn.dl.layers.core.Reshape(target_shape, input_shape=None)[source]

Bases: ztlearn.dl.layers.base.Layer

output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable
class ztlearn.dl.layers.core.UpSampling2D(size=(2, 2), input_shape=None)[source]

Bases: ztlearn.dl.layers.base.Layer

output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable

ztlearn.dl.layers.embedding module

class ztlearn.dl.layers.embedding.Embedding(input_dim, output_dim, activation='relu', input_shape=(1, 10))[source]

Bases: ztlearn.dl.layers.base.Layer

layer_activation
output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable
weight_initializer
weight_optimizer

ztlearn.dl.layers.normalization module

class ztlearn.dl.layers.normalization.BatchNormalization(eps=0.01, momentum=0.99)[source]

Bases: ztlearn.dl.layers.base.Layer

layer_parameters
output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable
weight_optimizer
class ztlearn.dl.layers.normalization.LayerNormalization1D(eps=1e-05)[source]

Bases: ztlearn.dl.layers.base.Layer

output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable
weight_optimizer

ztlearn.dl.layers.pooling module

class ztlearn.dl.layers.pooling.AveragePool2D(pool_size=(2, 2), strides=(1, 1), padding='valid')[source]

Bases: ztlearn.dl.layers.pooling.Pool

pool_backward(d_input_col, grad_col, pool_cache=None)[source]
pool_forward(input_col)[source]
class ztlearn.dl.layers.pooling.MaxPooling2D(pool_size=(2, 2), strides=(1, 1), padding='valid')[source]

Bases: ztlearn.dl.layers.pooling.Pool

pool_backward(d_input_col, grad_col, pool_cache)[source]
pool_forward(input_col)[source]
class ztlearn.dl.layers.pooling.Pool(pool_size=(2, 2), strides=(1, 1), padding='valid')[source]

Bases: ztlearn.dl.layers.base.Layer

output_shape
pass_backward(grad)[source]
pass_forward(inputs, train_mode=True, **kwargs)[source]
prep_layer()[source]
trainable

Module contents