def __init__(
self,
input,
n_visible=784,
n_hidden=500,
W=None,
hbias=None,
vbias=None,
numpy_rng=None,
transpose=False,
theano_rng=None,
weight_decay=0.0002,
):
RBM.__init__(
self,
input=input,
n_visible=n_visible,
n_hidden=n_hidden,
W=W,
hbias=hbias,
vbias=vbias,
numpy_rng=numpy_rng,
theano_rng=theano_rng,
weight_decay=weight_decay,
)
def __init__(self, input=None, n_visible=784, n_hidden=500,
W=None, h_bias=None, v_bias=None, numpy_rng=None, theano_rng=None):
"""
GBRBM constructor. Defines the parameters of the model along with
basic operations for inferring hidden from visible (and vice-versa).
It initialize parent class (RBM).
:param input: None for standalone RBMs or symbolic variable if RBM is part of a larger graph.
:param n_visible: number of visible units
:param n_hidden: number of hidden units
:param W: None for standalone RBMs or symbolic variable pointing to a
shared weight matrix in case RBM is part of a DBN network; in a DBN,
the weights are shared between RBMs and layers of a MLP
:param h_bias: None for standalone RBMs or symbolic variable pointing
to a shared hidden units bias vector in case RBM is part of a
different network
:param v_bias: None for standalone RBMs or a symbolic variable
pointing to a shared visible units bias
"""
RBM.__init__(
self,
input=input,
n_visible=n_visible,
n_hidden=n_hidden,
W=W, h_bias=h_bias,
v_bias=v_bias,
numpy_rng=numpy_rng,
theano_rng=theano_rng)
def __init__(
self,
input,
n_in=784,
n_hidden=500,
W=None,
hbias=None,
vbias=None,
numpy_rng=None,
transpose=False,
activation=T.nnet.sigmoid,
theano_rng=None,
name="grbm",
W_r=None,
dropout=0,
dropconnect=0,
):
# initialize parent class (RBM)
RBM.__init__(
self,
input=input,
n_visible=n_in,
n_hidden=n_hidden,
W=W,
hbias=hbias,
vbias=vbias,
numpy_rng=numpy_rng,
theano_rng=theano_rng,
)
def __init__(self,
input,
n_in,
n_out,
rng=None,
W=None,
vbias=None,
hbias=None):
self.input = input
self.n_in = n_in
self.n_out = n_out
RBM.__init__(self,
self.input,
n_visible=self.n_in,
n_hidden=self.n_out,
W=W,
vbias=vbias,
hbias=hbias)
if rng is None:
numpy_rng = np.random.RandomState(1234)
rng = MRG_RandomStreams(numpy_rng.randint(2**30))
self.theano_rng = rng
self.output = self.sample_h_given_v(self.input)[2]
self.feedbackward = self.gibbs_vhv(self.input)[2]
def __init__(self, input, n_in=784, n_hidden=500, \
W=None, hbias=None, vbias=None, numpy_rng=None, transpose=False, activation=T.nnet.sigmoid,
theano_rng=None, name='grbm', W_r=None, dropout=0, dropconnect=0):
# initialize parent class (RBM)
RBM.__init__(self, input=input, n_visible=n_in, n_hidden=n_hidden, \
W=W, hbias=hbias, vbias=vbias, numpy_rng=numpy_rng,
theano_rng=theano_rng)
def __init__(self, input, n_visible=784, n_hidden=500, W=None, hbias=None, vbias=None, numpy_rng=None,
theano_rng=None):
# initialize parent class (RBM)
# RBM.__init__(self, input=input, n_visible=n_in, n_hidden=n_hidden, activation=activation,
# W=W, hbias=hbias, vbias=vbias, transpose=transpose, numpy_rng=numpy_rng,
# theano_rng=theano_rng, name=name, dropout=dropout, dropconnect=dropconnect)
RBM.__init__(self, input=input, n_visible=n_visible, n_hidden=n_hidden, W=W, hbias=hbias, vbias=vbias,
numpy_rng=numpy_rng, theano_rng=theano_rng)
def __init__(self,
n_visible,
n_hidden,
sample_visible=False,
sigma=1,
**kwargs):
self.sample_visible = sample_visible
self.sigma = sigma
RBM.__init__(self, n_visible, n_hidden, **kwargs)
def __init__(self,
input,
n_visible=16,
n_hidden=20,
W=None, hbias=None, vbias=None,
numpy_rng=None, theano_rng=None):
# initialize parent class (RBM)
RBM.__init__(self,
input=input,
n_visible=n_visible,
n_hidden=n_hidden,
W=W, hbias=hbias, vbias=vbias,
numpy_rng=numpy_rng, theano_rng=theano_rng)
def __init__(self,
input,
n_hid,
n_vis,
Wp=None,
W=None,
hbias=None,
vbias=None):
# build parameters of Shifted RBM
RBM.__init__(self,
input,
n_visible=n_vis,
n_hidden=n_hid,
W=W,
hbias=hbias,
vbias=vbias)
def __init__(self,
n_visible,
n_hidden,
sample_visible=False,
sigma=1,
l1=0.0,
**kwargs):
# self.train_sigma = train_sigma
# if train_sigma:
# self.sigma = tf.Variable(sigma*tf.ones([n_visible]), dtype=tf.float32)
# self.delta_sigma = tf.Variable(tf.zeros([n_visible]), dtype=tf.float32)
# else:
# self.sigma = sigma
self.l1 = l1
self.lr_penalty = 1
RBM.__init__(self, n_visible, n_hidden, **kwargs)
def __init__(self, *args, **kwargs):
RBM.__init__(self, *args, **kwargs)
self.temp = 1
def __init__(self, input, n_visible=784, n_hidden=500, \
W=None, hbias=None, vbias=None, numpy_rng=None, transpose=False,
theano_rng=None, weight_decay=0.0002):
RBM.__init__(self, input=input, n_visible=n_visible, n_hidden=n_hidden, \
W=W, hbias=hbias, vbias=vbias, numpy_rng=numpy_rng,
theano_rng=theano_rng, weight_decay=weight_decay)
def __init__(self, *args, **kwargs):
RBM.__init__(self, *args, **kwargs)
def __init__(self, input, n_visible, n_hidden):
RBM.__init__(self, input=input, n_visible=n_visible,
n_hidden=n_hidden)
def __init__(self, input, n_hid, n_vis, Wp=None,
W=None, hbias=None, vbias=None):
# build parameters of Shifted RBM
RBM.__init__(self, input, n_visible=n_vis, n_hidden=n_hid,
W=W, hbias=hbias, vbias=vbias)
