def __init__(self,
n_components,
model_name='dae',
main_dir='dae/',
models_dir='models/',
data_dir='data/',
summary_dir='logs/',
enc_act_func=tf.nn.tanh,
dec_act_func=None,
loss_func='mean_squared',
num_epochs=10,
batch_size=10,
dataset='mnist',
opt='gradient_descent',
learning_rate=0.01,
momentum=0.5,
corr_type='none',
corr_frac=0.,
verbose=1,
regtype='none',
l2reg=5e-4):
"""Constructor.
:param n_components: number of hidden units
:param enc_act_func: Activation function for the encoder.
[tf.nn.tanh, tf.nn.sigmoid]
:param dec_act_func: Activation function for the decoder.
[tf.nn.tanh, tf.nn.sigmoid, None]
:param corr_type: Type of input corruption.
["none", "masking", "salt_and_pepper"]
:param corr_frac: Fraction of the input to corrupt.
:param verbose: Level of verbosity. 0 - silent, 1 - print accuracy.
:param l2reg: Regularization parameter. If 0, no regularization.
"""
UnsupervisedModel.__init__(self, model_name, main_dir, models_dir,
data_dir, summary_dir)
self._initialize_training_parameters(loss_func=loss_func,
learning_rate=learning_rate,
opt=opt,
num_epochs=num_epochs,
batch_size=batch_size,
dataset=dataset,
momentum=momentum,
regtype=regtype,
l2reg=l2reg)
self.n_components = n_components
self.enc_act_func = enc_act_func
self.dec_act_func = dec_act_func
self.corr_type = corr_type
self.corr_frac = corr_frac
self.verbose = verbose
self.input_data_orig = None
self.input_data = None
self.W_ = None
self.bh_ = None
self.bv_ = None
def __init__(self, n_components, model_name='dae', main_dir='dae/', models_dir='models/', data_dir='data/', summary_dir='logs/',
enc_act_func=tf.nn.tanh, dec_act_func=None, loss_func='mean_squared', num_epochs=10, batch_size=10, dataset='mnist',
opt='gradient_descent', learning_rate=0.01, momentum=0.5, corr_type='none', corr_frac=0., verbose=1, l2reg=5e-4):
"""
:param n_components: number of hidden units
:param enc_act_func: Activation function for the encoder. [tf.nn.tanh, tf.nn.sigmoid]
:param dec_act_func: Activation function for the decoder. [[tf.nn.tanh, tf.nn.sigmoid, None]
:param corr_type: Type of input corruption. ["none", "masking", "salt_and_pepper"]
:param corr_frac: Fraction of the input to corrupt.
:param verbose: Level of verbosity. 0 - silent, 1 - print accuracy.
:param l2reg: Regularization parameter. If 0, no regularization.
"""
UnsupervisedModel.__init__(self, model_name, main_dir, models_dir, data_dir, summary_dir)
self._initialize_training_parameters(loss_func=loss_func, learning_rate=learning_rate, num_epochs=num_epochs,
batch_size=batch_size, dataset=dataset, opt=opt, momentum=momentum,
l2reg=l2reg)
self.n_components = n_components
self.enc_act_func = enc_act_func
self.dec_act_func = dec_act_func
self.corr_type = corr_type
self.corr_frac = corr_frac
self.verbose = verbose
self.input_data_orig = None
self.input_data = None
self.W_ = None
self.bh_ = None
self.bv_ = None
def __init__(self, num_hidden, visible_unit_type='bin', main_dir='rbm', models_dir='models/', data_dir='data/', summary_dir='logs/',
model_name='rbm', dataset='mnist',
gibbs_sampling_steps=1, learning_rate=0.01, batch_size=10, num_epochs=10, stddev=0.1, verbose=0):
"""
:param num_hidden: number of hidden units
:param visible_unit_type: type of the visible units (binary or gaussian)
:param gibbs_sampling_steps: optional, default 1
:param stddev: optional, default 0.1. Ignored if visible_unit_type is not 'gauss'
:param verbose: level of verbosity. optional, default 0
"""
UnsupervisedModel.__init__(self, model_name, main_dir, models_dir, data_dir, summary_dir)
self._initialize_training_parameters(None, learning_rate, num_epochs, batch_size,
dataset, None, None)
self.num_hidden = num_hidden
self.visible_unit_type = visible_unit_type
self.gibbs_sampling_steps = gibbs_sampling_steps
self.stddev = stddev
self.verbose = verbose
self.W = None
self.bh_ = None
self.bv_ = None
self.w_upd8 = None
self.bh_upd8 = None
self.bv_upd8 = None
self.cost = None
self.input_data = None
self.hrand = None
self.vrand = None
def __init__(self, layers, model_name='sdae', main_dir='sdae/', models_dir='models/', data_dir='data/', summary_dir='logs/',
num_epochs=[10], batch_size=[10], dataset='mnist', learning_rate=[0.01], gibbs_k=[1],
momentum=0.5, finetune_dropout=1, verbose=1, finetune_loss_func='cross_entropy', finetune_enc_act_func=[tf.nn.relu],
finetune_dec_act_func=[tf.nn.sigmoid], finetune_opt='gradient_descent', finetune_learning_rate=0.001,
finetune_num_epochs=10, noise=['gauss'], stddev=0.1, finetune_batch_size=20, do_pretrain=True):
"""
:param layers: list containing the hidden units for each layer
:param finetune_loss_func: Loss function for the softmax layer. string, default ['cross_entropy', 'mean_squared']
:param finetune_dropout: dropout parameter
:param finetune_learning_rate: learning rate for the finetuning. float, default 0.001
:param finetune_enc_act_func: activation function for the encoder finetuning phase
:param finetune_dec_act_func: activation function for the decoder finetuning phase
:param finetune_opt: optimizer for the finetuning phase
:param finetune_num_epochs: Number of epochs for the finetuning. int, default 20
:param finetune_batch_size: Size of each mini-batch for the finetuning. int, default 20
:param verbose: Level of verbosity. 0 - silent, 1 - print accuracy. int, default 0
:param do_pretrain: True: uses variables from pretraining, False: initialize new variables.
"""
# WARNING! This must be the first expression in the function or else it will send other variables to expanded_args()
# This function takes all the passed parameters that are lists and expands them to the number of layers, if the number
# of layers is more than the list of the parameter.
expanded_args = utilities.expand_args(layers, **locals())
UnsupervisedModel.__init__(self, model_name, main_dir, models_dir, data_dir, summary_dir)
self._initialize_training_parameters(loss_func=finetune_loss_func, learning_rate=finetune_learning_rate,
num_epochs=finetune_num_epochs, batch_size=finetune_batch_size,
dropout=finetune_dropout, dataset=dataset, opt=finetune_opt, momentum=momentum)
self.do_pretrain = do_pretrain
self.layers = layers
self.verbose = verbose
self.finetune_enc_act_func = expanded_args['finetune_enc_act_func']
self.finetune_dec_act_func = expanded_args['finetune_dec_act_func']
self.input_ref = None
# Model parameters
self.encoding_w_ = [] # list of matrices of encoding weights (one per layer)
self.encoding_b_ = [] # list of arrays of encoding biases (one per layer)
self.decoding_w = [] # list of matrices of decoding weights (one per layer)
self.decoding_b = [] # list of arrays of decoding biases (one per layer)
self.reconstruction = None
self.rbms = []
self.rbm_graphs = []
for l, layer in enumerate(layers):
rbm_str = 'rbm-' + str(l + 1)
new_rbm = rbm.RBM(model_name=self.model_name + '-' + rbm_str,
models_dir=os.path.join(self.models_dir, rbm_str), data_dir=os.path.join(self.data_dir, rbm_str),
summary_dir=os.path.join(self.tf_summary_dir, rbm_str), visible_unit_type=expanded_args['noise'][l],
stddev=stddev, num_hidden=expanded_args['layers'][l], main_dir=self.main_dir, learning_rate=expanded_args['learning_rate'][l],
gibbs_sampling_steps=expanded_args['gibbs_k'][l], verbose=self.verbose, num_epochs=expanded_args['num_epochs'][l],
batch_size=expanded_args['batch_size'][l])
self.rbms.append(new_rbm)
self.rbm_graphs.append(tf.Graph())
def __init__(self,
num_hidden,
visible_unit_type='bin',
main_dir='rbm/',
models_dir='models/',
data_dir='data/',
summary_dir='logs/',
model_name='rbm',
dataset='mnist',
loss_func='mean_squared',
l2reg=5e-4,
regtype='none',
gibbs_sampling_steps=1,
learning_rate=0.01,
batch_size=10,
num_epochs=10,
stddev=0.1,
verbose=0):
"""Constructor.
:param num_hidden: number of hidden units
:param loss_function: type of loss function
:param visible_unit_type: type of the visible units (bin or gauss)
:param gibbs_sampling_steps: optional, default 1
:param stddev: default 0.1. Ignored if visible_unit_type is not 'gauss'
:param verbose: level of verbosity. optional, default 0
"""
UnsupervisedModel.__init__(self, model_name, main_dir, models_dir,
data_dir, summary_dir)
self._initialize_training_parameters(loss_func=loss_func,
learning_rate=learning_rate,
num_epochs=num_epochs,
batch_size=batch_size,
dataset=dataset,
regtype=regtype,
l2reg=l2reg)
self.num_hidden = num_hidden
self.visible_unit_type = visible_unit_type
self.gibbs_sampling_steps = gibbs_sampling_steps
self.stddev = stddev
self.verbose = verbose
self.W = None
self.bh_ = None
self.bv_ = None
self.w_upd8 = None
self.bh_upd8 = None
self.bv_upd8 = None
self.cost = None
self.input_data = None
self.hrand = None
self.vrand = None
def pretrain(self, train_set, validation_set=None):
self.do_pretrain = True
def set_params_func(rbmmachine, rbmgraph):
params = rbmmachine.get_model_parameters(graph=rbmgraph)
self.encoding_w_.append(params['W'])
self.encoding_b_.append(params['bh_'])
return UnsupervisedModel.pretrain_procedure(self, self.rbms, self.rbm_graphs, set_params_func=set_params_func,
train_set=train_set, validation_set=validation_set)
def __init__(
self, num_hidden, visible_unit_type='bin', main_dir='rbm/',
models_dir='models/', data_dir='data/', summary_dir='logs/',
model_name='rbm', dataset='mnist', loss_func='mean_squared',
l2reg=5e-4, regtype='none', gibbs_sampling_steps=1, learning_rate=0.01,
batch_size=10, num_epochs=10, stddev=0.1, verbose=0):
"""Constructor.
:param num_hidden: number of hidden units
:param loss_function: type of loss function
:param visible_unit_type: type of the visible units (bin or gauss)
:param gibbs_sampling_steps: optional, default 1
:param stddev: default 0.1. Ignored if visible_unit_type is not 'gauss'
:param verbose: level of verbosity. optional, default 0
"""
UnsupervisedModel.__init__(self, model_name, main_dir, models_dir,
data_dir, summary_dir)
self._initialize_training_parameters(
loss_func=loss_func, learning_rate=learning_rate,
num_epochs=num_epochs, batch_size=batch_size, dataset=dataset,
regtype=regtype, l2reg=l2reg)
self.num_hidden = num_hidden
self.visible_unit_type = visible_unit_type
self.gibbs_sampling_steps = gibbs_sampling_steps
self.stddev = stddev
self.verbose = verbose
self.W = None
self.bh_ = None
self.bv_ = None
self.w_upd8 = None
self.bh_upd8 = None
self.bv_upd8 = None
self.cost = None
self.input_data = None
self.hrand = None
self.vrand = None
def pretrain(self, train_set, validation_set=None):
self.do_pretrain = True
def set_params_func(autoenc, autoencgraph):
params = autoenc.get_model_parameters(graph=autoencgraph)
self.encoding_w_.append(params['enc_w'])
self.encoding_b_.append(params['enc_b'])
return UnsupervisedModel.pretrain_procedure(self, self.autoencoders, self.autoencoder_graphs,
set_params_func=set_params_func, train_set=train_set,
validation_set=validation_set)
def pretrain(self, train_set, validation_set=None):
self.do_pretrain = True
def set_params_func(rbmmachine, rbmgraph):
params = rbmmachine.get_model_parameters(graph=rbmgraph)
self.encoding_w_.append(params['W'])
self.encoding_b_.append(params['bh_'])
return UnsupervisedModel.pretrain_procedure(
self,
self.rbms,
self.rbm_graphs,
set_params_func=set_params_func,
train_set=train_set,
validation_set=validation_set)
def pretrain(self, train_set, validation_set=None):
"""Perform Unsupervised pretraining of the autoencoder."""
self.do_pretrain = True
def set_params_func(rbmmachine, rbmgraph):
params = rbmmachine.get_model_parameters(graph=rbmgraph)
self.encoding_w_.append(params["W"])
self.encoding_b_.append(params["bh_"])
return UnsupervisedModel.pretrain_procedure(
self,
self.rbms,
self.rbm_graphs,
set_params_func=set_params_func,
train_set=train_set,
validation_set=validation_set,
)
def __init__(self,
layers,
model_name='srbm',
main_dir='srbm/',
models_dir='models/',
data_dir='data/',
summary_dir='logs/',
num_epochs=[10],
batch_size=[10],
dataset='mnist',
learning_rate=[0.01],
gibbs_k=[1],
loss_func=['mean_squared'],
momentum=0.5,
finetune_dropout=1,
verbose=1,
finetune_loss_func='cross_entropy',
finetune_enc_act_func=[tf.nn.relu],
finetune_dec_act_func=[tf.nn.sigmoid],
finetune_opt='gradient_descent',
finetune_learning_rate=0.001,
l2reg=5e-4,
finetune_num_epochs=10,
noise=['gauss'],
stddev=0.1,
finetune_batch_size=20,
do_pretrain=False,
tied_weights=False,
regtype=['none'],
finetune_reg_type='none'):
"""Constructor.
:param layers: list containing the hidden units for each layer
:param finetune_loss_func: Loss function for the softmax layer.
string, default ['cross_entropy', 'mean_squared']
:param finetune_dropout: dropout parameter
:param finetune_learning_rate: learning rate for the finetuning.
float, default 0.001
:param finetune_enc_act_func: activation function for the encoder
finetuning phase
:param finetune_dec_act_func: activation function for the decoder
finetuning phase
:param finetune_opt: optimizer for the finetuning phase
:param finetune_num_epochs: Number of epochs for the finetuning.
int, default 20
:param finetune_batch_size: Size of each mini-batch for the finetuning.
int, default 20
:param verbose: Level of verbosity. 0 - silent, 1 - print accuracy.
int, default 0
:param do_pretrain: True: uses variables from pretraining,
False: initialize new variables.
"""
# WARNING! This must be the first expression in the function or else it
# will send other variables to expanded_args()
# This function takes all the passed parameters that are lists and
# expands them to the number of layers, if the number
# of layers is more than the list of the parameter.
expanded_args = utilities.expand_args(**locals())
UnsupervisedModel.__init__(self, model_name, main_dir, models_dir,
data_dir, summary_dir)
self._initialize_training_parameters(
loss_func=finetune_loss_func,
learning_rate=finetune_learning_rate,
regtype=finetune_reg_type,
num_epochs=finetune_num_epochs,
batch_size=finetune_batch_size,
l2reg=l2reg,
dropout=finetune_dropout,
dataset=dataset,
opt=finetune_opt,
momentum=momentum)
self.do_pretrain = do_pretrain
self.layers = layers
self.tied_weights = tied_weights
self.verbose = verbose
self.finetune_enc_act_func = expanded_args['finetune_enc_act_func']
self.finetune_dec_act_func = expanded_args['finetune_dec_act_func']
self.input_ref = None
# Model parameters
self.encoding_w_ = [] # list of matrices of encoding weights per layer
self.encoding_b_ = [] # list of arrays of encoding biases per layer
self.decoding_w = [] # list of matrices of decoding weights per layer
self.decoding_b = [] # list of arrays of decoding biases per layer
self.reconstruction = None
self.rbms = []
self.rbm_graphs = []
for l, layer in enumerate(layers):
rbm_str = 'rbm-' + str(l + 1)
new_rbm = rbm.RBM(model_name=self.model_name + '-' + rbm_str,
loss_func=expanded_args['loss_func'][l],
models_dir=os.path.join(self.models_dir,
rbm_str),
data_dir=os.path.join(self.data_dir, rbm_str),
summary_dir=os.path.join(self.tf_summary_dir,
rbm_str),
visible_unit_type=expanded_args['noise'][l],
stddev=stddev,
num_hidden=expanded_args['layers'][l],
main_dir=self.main_dir,
learning_rate=expanded_args['learning_rate'][l],
gibbs_sampling_steps=expanded_args['gibbs_k'][l],
num_epochs=expanded_args['num_epochs'][l],
batch_size=expanded_args['batch_size'][l],
verbose=self.verbose,
regtype=expanded_args['regtype'][l])
self.rbms.append(new_rbm)
self.rbm_graphs.append(tf.Graph())
def __init__(self,
layers,
model_name='sdae',
main_dir='sdae/',
models_dir='models/',
data_dir='data/',
summary_dir='logs/',
enc_act_func=[tf.nn.tanh],
dec_act_func=[None],
loss_func=['cross_entropy'],
num_epochs=[10],
batch_size=[10],
dataset='mnist',
opt=['gradient_descent'],
regtype=['none'],
learning_rate=[0.01],
momentum=0.5,
finetune_dropout=1,
corr_type=['none'],
finetune_regtype='none',
corr_frac=[0.],
verbose=1,
finetune_loss_func='cross_entropy',
finetune_enc_act_func=[tf.nn.relu],
tied_weights=True,
finetune_dec_act_func=[tf.nn.sigmoid],
l2reg=[5e-4],
finetune_batch_size=20,
do_pretrain=False,
finetune_opt='gradient_descent',
finetune_learning_rate=0.001,
finetune_num_epochs=10):
"""Constructor.
:param layers: list containing the hidden units for each layer
:param enc_act_func: Activation function for the encoder.
[tf.nn.tanh, tf.nn.sigmoid]
:param dec_act_func: Activation function for the decoder.
[tf.nn.tanh, tf.nn.sigmoid, None]
:param regtype: regularization type, can be 'l2','l1' and 'none'
:param finetune_regtype: regularization type for finetuning
:param finetune_loss_func: Loss function for the softmax layer.
string, default ['cross_entropy', 'mean_squared']
:param finetune_dropout: dropout parameter
:param finetune_learning_rate: learning rate for the finetuning.
float, default 0.001
:param finetune_enc_act_func: activation function for the encoder
finetuning phase
:param finetune_dec_act_func: activation function for the decoder
finetuning phase
:param finetune_opt: optimizer for the finetuning phase
:param finetune_num_epochs: Number of epochs for the finetuning.
int, default 20
:param finetune_batch_size: Size of each mini-batch for the finetuning.
int, default 20
:param tied_weights: if True, the decoder layers weights are
constrained to be the transpose of the encoder layers
:param corr_type: Type of input corruption. string, default 'none'.
["none", "masking", "salt_and_pepper"]
:param corr_frac: Fraction of the input to corrupt. float, default 0.0
:param verbose: Level of verbosity. 0 - silent, 1 - print accuracy.
int, default 0
:param do_pretrain: True: uses variables from pretraining,
False: initialize new variables.
"""
# WARNING! This must be the first expression in the function or else it
# will send other variables to expanded_args()
# This function takes all the passed parameters that are lists and
# expands them to the number of layers, if the number
# of layers is more than the list of the parameter.
expanded_args = utilities.expand_args(**locals())
UnsupervisedModel.__init__(self, model_name, main_dir, models_dir,
data_dir, summary_dir)
self._initialize_training_parameters(
loss_func=finetune_loss_func,
learning_rate=finetune_learning_rate,
num_epochs=finetune_num_epochs,
batch_size=finetune_batch_size,
l2reg=l2reg,
regtype=finetune_regtype,
dropout=finetune_dropout,
dataset=dataset,
opt=finetune_opt,
momentum=momentum)
self.do_pretrain = do_pretrain
self.layers = layers
self.tied_weights = tied_weights
self.verbose = verbose
self.finetune_enc_act_func = expanded_args['finetune_enc_act_func']
self.finetune_dec_act_func = expanded_args['finetune_dec_act_func']
self.input_ref = None
# Model parameters
self.encoding_w_ = [] # list of matrices of encoding weights per layer
self.encoding_b_ = [] # list of arrays of encoding biases per layer
self.decoding_w = [] # list of matrices of decoding weights per layer
self.decoding_b = [] # list of arrays of decoding biases per layer
self.reconstruction = None
self.autoencoders = []
self.autoencoder_graphs = []
for l, layer in enumerate(layers):
dae_str = 'dae-' + str(l + 1)
self.autoencoders.append(
denoising_autoencoder.DenoisingAutoencoder(
n_components=layer,
main_dir=self.main_dir,
model_name=self.model_name + '-' + dae_str,
models_dir=os.path.join(self.models_dir, dae_str),
data_dir=os.path.join(self.data_dir, dae_str),
summary_dir=os.path.join(self.tf_summary_dir, dae_str),
enc_act_func=expanded_args['enc_act_func'][l],
dec_act_func=expanded_args['dec_act_func'][l],
loss_func=expanded_args['loss_func'][l],
regtype=expanded_args['regtype'][l],
opt=expanded_args['opt'][l],
learning_rate=expanded_args['learning_rate'][l],
l2reg=expanded_args['l2reg'],
dataset=self.dataset,
momentum=self.momentum,
verbose=self.verbose,
corr_type=expanded_args['corr_type'][l],
corr_frac=expanded_args['corr_frac'][l],
num_epochs=expanded_args['num_epochs'][l],
batch_size=expanded_args['batch_size'][l]))
self.autoencoder_graphs.append(tf.Graph())
def __init__(
self, layers, model_name='sdae', main_dir='sdae/',
models_dir='models/', data_dir='data/', summary_dir='logs/',
enc_act_func=[tf.nn.tanh], dec_act_func=[None],
loss_func=['cross_entropy'], num_epochs=[10], batch_size=[10],
dataset='mnist', opt=['gradient_descent'], regtype=['none'],
learning_rate=[0.01], momentum=0.5, finetune_dropout=1,
corr_type=['none'], finetune_regtype='none', corr_frac=[0.], verbose=1,
finetune_loss_func='cross_entropy', finetune_enc_act_func=[tf.nn.relu],
tied_weights=True, finetune_dec_act_func=[tf.nn.sigmoid], l2reg=[5e-4],
finetune_batch_size=20, do_pretrain=False,
finetune_opt='gradient_descent', finetune_learning_rate=0.001,
finetune_num_epochs=10):
"""Constructor.
:param layers: list containing the hidden units for each layer
:param enc_act_func: Activation function for the encoder.
[tf.nn.tanh, tf.nn.sigmoid]
:param dec_act_func: Activation function for the decoder.
[tf.nn.tanh, tf.nn.sigmoid, None]
:param regtype: regularization type, can be 'l2','l1' and 'none'
:param finetune_regtype: regularization type for finetuning
:param finetune_loss_func: Loss function for the softmax layer.
string, default ['cross_entropy', 'mean_squared']
:param finetune_dropout: dropout parameter
:param finetune_learning_rate: learning rate for the finetuning.
float, default 0.001
:param finetune_enc_act_func: activation function for the encoder
finetuning phase
:param finetune_dec_act_func: activation function for the decoder
finetuning phase
:param finetune_opt: optimizer for the finetuning phase
:param finetune_num_epochs: Number of epochs for the finetuning.
int, default 20
:param finetune_batch_size: Size of each mini-batch for the finetuning.
int, default 20
:param tied_weights: if True, the decoder layers weights are
constrained to be the transpose of the encoder layers
:param corr_type: Type of input corruption. string, default 'none'.
["none", "masking", "salt_and_pepper"]
:param corr_frac: Fraction of the input to corrupt. float, default 0.0
:param verbose: Level of verbosity. 0 - silent, 1 - print accuracy.
int, default 0
:param do_pretrain: True: uses variables from pretraining,
False: initialize new variables.
"""
# WARNING! This must be the first expression in the function or else it
# will send other variables to expanded_args()
# This function takes all the passed parameters that are lists and
# expands them to the number of layers, if the number
# of layers is more than the list of the parameter.
expanded_args = utilities.expand_args(**locals())
UnsupervisedModel.__init__(
self, model_name, main_dir, models_dir, data_dir, summary_dir)
self._initialize_training_parameters(
loss_func=finetune_loss_func, learning_rate=finetune_learning_rate,
num_epochs=finetune_num_epochs, batch_size=finetune_batch_size,
l2reg=l2reg, regtype=finetune_regtype, dropout=finetune_dropout,
dataset=dataset, opt=finetune_opt, momentum=momentum)
self.do_pretrain = do_pretrain
self.layers = layers
self.tied_weights = tied_weights
self.verbose = verbose
self.finetune_enc_act_func = expanded_args['finetune_enc_act_func']
self.finetune_dec_act_func = expanded_args['finetune_dec_act_func']
self.input_ref = None
# Model parameters
self.encoding_w_ = [] # list of matrices of encoding weights per layer
self.encoding_b_ = [] # list of arrays of encoding biases per layer
self.decoding_w = [] # list of matrices of decoding weights per layer
self.decoding_b = [] # list of arrays of decoding biases per layer
self.reconstruction = None
self.autoencoders = []
self.autoencoder_graphs = []
for l, layer in enumerate(layers):
dae_str = 'dae-' + str(l + 1)
self.autoencoders.append(
denoising_autoencoder.DenoisingAutoencoder(
n_components=layer, main_dir=self.main_dir,
model_name=self.model_name + '-' + dae_str,
models_dir=os.path.join(self.models_dir, dae_str),
data_dir=os.path.join(self.data_dir, dae_str),
summary_dir=os.path.join(self.tf_summary_dir, dae_str),
enc_act_func=expanded_args['enc_act_func'][l],
dec_act_func=expanded_args['dec_act_func'][l],
loss_func=expanded_args['loss_func'][l],
regtype=expanded_args['regtype'][l],
opt=expanded_args['opt'][l],
learning_rate=expanded_args['learning_rate'][l],
l2reg=expanded_args['l2reg'], dataset=self.dataset,
momentum=self.momentum, verbose=self.verbose,
corr_type=expanded_args['corr_type'][l],
corr_frac=expanded_args['corr_frac'][l],
num_epochs=expanded_args['num_epochs'][l],
batch_size=expanded_args['batch_size'][l]))
self.autoencoder_graphs.append(tf.Graph())
