def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
i = sharedX(0., 'counter')
i_t = i + 1.
b1_t = self.b1**i_t
b2_t = self.b2**i_t
lr_t = self.lr * T.sqrt(1. - b2_t) / (1 - b1_t)
#b1 = 1 - self.b1 * self.lambd**i
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
m = sharedX(p.get_value() * 0.)
v = sharedX(p.get_value() * 0.)
#m_t = b1 * m + (1 - b1) * g
m_t = self.b1 * m + (1 - self.b1) * g
v_t = self.b2 * v + (1 - self.b2) * g**2
g_t = m_t / (T.sqrt(v_t) + self.eps)
p_t = p - lr_scaler * lr_t * g_t
updates[m] = m_t
updates[v] = v_t
updates[p] = p_t
updates[i] = i_t
return updates
def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
i = sharedX(0., 'counter')
i_t = i + 1.
b1_t = self.b1**i_t
b2_t = self.b2**i_t
lr_t = self.lr * T.sqrt(1. - b2_t) / (1 - b1_t)
#b1 = 1 - self.b1 * self.lambd**i
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
m = sharedX(p.get_value() * 0.)
v = sharedX(p.get_value() * 0.)
#m_t = b1 * m + (1 - b1) * g
m_t = self.b1 * m + (1 - self.b1) * g
v_t = self.b2 * v + (1 - self.b2) * g**2
g_t = m_t / (T.sqrt(v_t) + self.eps)
p_t = p - lr_scaler * lr_t * g_t
updates[m] = m_t
updates[v] = v_t
updates[p] = p_t
updates[i] = i_t
return updates
def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
cnt = sharedX(0, 'counter')
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
m = sharedX(p.get_value() * 0.)
v = sharedX(p.get_value() * 0.)
b1 = self.b1 * self.lambd**cnt
m_t = b1 * m + (1 - b1) * g
v_t = self.b2 * v + (1 - self.b2) * g**2
m_t_hat = m_t / (1. - self.b1**(cnt + 1))
v_t_hat = v_t / (1. - self.b2**(cnt + 1))
g_t = m_t_hat / (T.sqrt(v_t_hat) + self.e)
p_t = p - lr_scaler * self.lr * g_t
updates[m] = m_t
updates[v] = v_t
updates[p] = p_t
updates[cnt] = cnt + 1
return updates
def __init__(self, rho=0.5, eps=1e-6, **kwargs):
super(BatchNormLayer, self).__init__(**kwargs)
self.rho = rho
self.eps = eps
self.mu = sharedX(InitCell('zeros').get(self.nout),
name='mu_' + self.name)
self.sigma = sharedX(InitCell('ones').get(self.nout),
name='sigma_' + self.name)
def __init__(self, lr, lr_scalers=None):
"""
.. todo::
WRITEME
"""
self.lr = sharedX(lr)
if lr_scalers is not None:
self.lr_scalers = lr_scalers
else:
self.lr_scalers = OrderedDict()
def __init__(self, lr, lr_scalers=None):
"""
.. todo::
WRITEME
"""
self.lr = sharedX(lr)
if lr_scalers is not None:
self.lr_scalers = lr_scalers
else:
self.lr_scalers = OrderedDict()
def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
u = sharedX(p.get_value() * 0.)
avg_grad = sharedX(p.get_value() * 0.)
sqr_grad = sharedX(p.get_value() * 0.)
avg_grad_t = self.sec_mom * avg_grad + (1 - self.sec_mom) * g
sqr_grad_t = self.sec_mom * sqr_grad + (1 - self.sec_mom) * g**2
g_t = g / T.sqrt(sqr_grad_t - avg_grad_t**2 + self.e)
u_t = self.mom * u - lr_scaler * self.lr * g_t
p_t = p + u_t
updates[avg_grad] = avg_grad_t
updates[sqr_grad] = sqr_grad_t
updates[u] = u_t
updates[p] = p_t
return updates
def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
u = sharedX(p.get_value() * 0.)
avg_grad = sharedX(p.get_value() * 0.)
sqr_grad = sharedX(p.get_value() * 0.)
avg_grad_t = self.sec_mom * avg_grad + (1 - self.sec_mom) * g
sqr_grad_t = self.sec_mom * sqr_grad + (1 - self.sec_mom) * g**2
g_t = g / T.sqrt(sqr_grad_t - avg_grad_t**2 + self.e)
u_t = self.mom * u - lr_scaler * self.lr * g_t
p_t = p + u_t
updates[avg_grad] = avg_grad_t
updates[sqr_grad] = sqr_grad_t
updates[u] = u_t
updates[p] = p_t
return updates
def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
g_tt = OrderedDict()
cnt = sharedX(0, 'counter')
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
m = sharedX(p.get_value() * 0.)
v = sharedX(p.get_value() * 0.)
b1 = self.b1 * self.lambd**cnt
m_t = b1 * m + (1 - b1) * g
v_t = self.b2 * v + (1 - self.b2) * g**2
m_t_hat = m_t / (1. - self.b1**(cnt + 1))
v_t_hat = v_t / (1. - self.b2**(cnt + 1))
g_t = m_t_hat / (T.sqrt(v_t_hat) + self.e)
p_t = p - lr_scaler * self.lr * g_t
g_tt[p] = g_t
updates[m] = m_t
updates[v] = v_t
updates[p] = p_t
if self.post_clip:
g_norm = sum([T.sqr(x/self.batch_size).sum()
for x in g_tt.values()])
not_finite = T.or_(T.isnan(g_norm), T.isinf(g_norm))
g_norm = T.sqrt(g_norm)
scaler = self.scaler / T.maximum(self.scaler, g_norm)
for p, g in g_tt.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
p_t = p - lr_scaler * self.lr * g * scaler
updates[p] = p_t
updates[cnt] = cnt + 1
return updates
def __init__(self, lr, lr_scalers=None, post_clip=0,
scaler=5, batch_size=1):
"""
.. todo::
WRITEME
"""
self.lr = sharedX(lr)
if lr_scalers is not None:
self.lr_scalers = lr_scalers
else:
self.lr_scalers = OrderedDict()
self.post_clip = post_clip
self.scaler = scaler
self.batch_size = batch_size
def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
cnt = sharedX(0, 'counter')
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
m = sharedX(p.get_value() * 0.)
v = sharedX(p.get_value() * 0.)
b1 = self.b1 * self.lambd**cnt
m_t = b1 * m + (1 - b1) * g
v_t = self.b2 * v + (1 - self.b2) * g**2
m_t_hat = m_t / (1. - self.b1**(cnt + 1))
v_t_hat = v_t / (1. - self.b2**(cnt + 1))
g_t = m_t_hat / (T.sqrt(v_t_hat) + self.e)
p_t = p - lr_scaler * self.lr * g_t
updates[m] = m_t
updates[v] = v_t
updates[p] = p_t
updates[cnt] = cnt + 1
return updates
def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
u = sharedX(p.get_value() * 0.)
u_t = self.mom * u - self.lr * g
if self.nesterov:
u_t = self.mom * u_t - lr_scaler * self.lr * g
p_t = p + u_t
updates[u] = u_t
updates[p] = p_t
return updates
def get_updates(self, grads):
"""
.. todo::
WRITEME
"""
updates = OrderedDict()
for p, g in grads.items():
lr_scaler = self.lr_scalers.get(str(p), 1.)
u = sharedX(p.get_value() * 0.)
u_t = self.mom * u - self.lr * g
if self.nesterov:
u_t = self.mom * u_t - lr_scaler * self.lr * g
p_t = p + u_t
updates[u] = u_t
updates[p] = p_t
return updates
init_W=init_W,
init_b=init_b_sig)
nodes = [lstm_1, lstm_2, lstm_3, prior, kl,
x_1, x_2, x_3, x_4,
z_1, z_2, z_3, z_4,
phi_1, phi_2, phi_3, phi_4, phi_mu, phi_sig,
prior_1, prior_2, prior_3, prior_4, prior_mu, prior_sig,
theta_1, theta_2, theta_3, theta_4, theta_mu, theta_sig]
for node in nodes:
node.initialize()
params = flatten([node.get_params().values() for node in nodes])
step_count = sharedX(0, name='step_count')
last_lstm_1 = np.zeros((batch_size, lstm_1_dim*2), dtype=theano.config.floatX)
last_lstm_2 = np.zeros((batch_size, lstm_2_dim*2), dtype=theano.config.floatX)
last_lstm_3 = np.zeros((batch_size, lstm_3_dim*2), dtype=theano.config.floatX)
lstm_1_tm1 = sharedX(last_lstm_1, name='lstm_1_tm1')
lstm_2_tm1 = sharedX(last_lstm_2, name='lstm_2_tm1')
lstm_3_tm1 = sharedX(last_lstm_3, name='lstm_3_tm1')
update_list = [step_count, lstm_1_tm1, lstm_2_tm1, lstm_3_tm1]
step_count = T.switch(T.le(step_count, reset_freq), step_count + 1, 0)
s_1_0 = T.switch(T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(lstm_1_tm1), 0.), 'int32')),
lstm_1.get_init_state(batch_size), lstm_1_tm1)
s_2_0 = T.switch(T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(lstm_2_tm1), 0.), 'int32')),
lstm_2.get_init_state(batch_size), lstm_2_tm1)
def setX(self, x, name=None):
return sharedX(x, name)
def getX(self, shape, name=None):
return sharedX(self.init_param(shape), name)
nout=k,
unit='softmax',
init_W=init_W,
init_b=init_b)
nodes = [rnn,
x_1, x_2, x_3, x_4,
theta_1, theta_2, theta_3, theta_4, theta_mu, theta_sig, coeff]
params = OrderedDict()
for node in nodes:
if node.initialize() is not None:
params.update(node.initialize())
params = init_tparams(params)
step_count = sharedX(0, name='step_count')
last_rnn = np.zeros((batch_size, rnn_dim*2), dtype=theano.config.floatX)
rnn_tm1 = sharedX(last_rnn, name='rnn_tm1')
shared_updates = OrderedDict()
shared_updates[step_count] = step_count + 1
s_0 = T.switch(T.eq(T.mod(step_count, reset_freq), 0),
rnn.get_init_state(batch_size), rnn_tm1)
x_1_temp = x_1.fprop([x], params)
x_2_temp = x_2.fprop([x_1_temp], params)
x_3_temp = x_3.fprop([x_2_temp], params)
x_4_temp = x_4.fprop([x_3_temp], params)
def inner_fn(x_t, s_tm1):
phi_mu,
phi_sig,
theta_1,
theta_2,
theta_3,
theta_4,
theta_mu,
theta_sig,
]
for node in nodes:
node.initialize()
params = flatten([node.get_params().values() for node in nodes])
step_count = sharedX(0, name="step_count")
last_encoder = np.zeros((batch_size, encoder_dim * 2), dtype=theano.config.floatX)
last_decoder = np.zeros((batch_size, decoder_dim * 2), dtype=theano.config.floatX)
encoder_tm1 = sharedX(last_encoder, name="encoder_tm1")
decoder_tm1 = sharedX(last_decoder, name="decoder_tm1")
update_list = [step_count, encoder_tm1, decoder_tm1]
step_count = T.switch(T.le(step_count, reset_freq), step_count + 1, 0)
enc_0 = T.switch(
T.or_(T.cast(T.eq(step_count, 0), "int32"), T.cast(T.eq(T.sum(encoder_tm1), 0.0), "int32")),
encoder.get_init_state(batch_size),
encoder_tm1,
)
dec_0 = T.switch(
T.or_(T.cast(T.eq(step_count, 0), "int32"), T.cast(T.eq(T.sum(decoder_tm1), 0.0), "int32")),
decoder.get_init_state(batch_size),
def main(args):
trial = int(args['trial'])
pkl_name = 'vrnn_gauss_%d' % trial
channel_name = 'valid_nll_upper_bound'
data_path = args['data_path']
save_path = args['save_path']
monitoring_freq = int(args['monitoring_freq'])
force_saving_freq = int(args['force_saving_freq'])
reset_freq = int(args['reset_freq'])
epoch = int(args['epoch'])
batch_size = int(args['batch_size'])
m_batch_size = int(args['m_batch_size'])
x_dim = int(args['x_dim'])
z_dim = int(args['z_dim'])
rnn_dim = int(args['rnn_dim'])
lr = float(args['lr'])
debug = int(args['debug'])
print "trial no. %d" % trial
print "batch size %d" % batch_size
print "learning rate %f" % lr
print "saving pkl file '%s'" % pkl_name
print "to the save path '%s'" % save_path
q_z_dim = 500
p_z_dim = 500
p_x_dim = 600
x2s_dim = 600
z2s_dim = 500
target_dim = x_dim
file_name = 'blizzard_unseg_tbptt'
normal_params = np.load(data_path + file_name + '_normal.npz')
X_mean = normal_params['X_mean']
X_std = normal_params['X_std']
model = Model()
train_data = Blizzard_tbptt(name='train',
path=data_path,
frame_size=x_dim,
file_name=file_name,
X_mean=X_mean,
X_std=X_std)
valid_data = Blizzard_tbptt(name='valid',
path=data_path,
frame_size=x_dim,
file_name=file_name,
X_mean=X_mean,
X_std=X_std)
x = train_data.theano_vars()
m_x = valid_data.theano_vars()
if debug:
x.tag.test_value = np.zeros((15, batch_size, x_dim),
dtype=theano.config.floatX)
m_x.tag.test_value = np.zeros((15, m_batch_size, x_dim),
dtype=theano.config.floatX)
init_W = InitCell('rand')
init_U = InitCell('ortho')
init_b = InitCell('zeros')
init_b_sig = InitCell('const', mean=0.6)
x_1 = FullyConnectedLayer(name='x_1',
parent=['x_t'],
parent_dim=[x_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_2 = FullyConnectedLayer(name='x_2',
parent=['x_1'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_3 = FullyConnectedLayer(name='x_3',
parent=['x_2'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_4 = FullyConnectedLayer(name='x_4',
parent=['x_3'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
z_1 = FullyConnectedLayer(name='z_1',
parent=['z_t'],
parent_dim=[z_dim],
nout=z2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
z_2 = FullyConnectedLayer(name='z_2',
parent=['z_1'],
parent_dim=[z2s_dim],
nout=z2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
z_3 = FullyConnectedLayer(name='z_3',
parent=['z_2'],
parent_dim=[z2s_dim],
nout=z2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
z_4 = FullyConnectedLayer(name='z_4',
parent=['z_3'],
parent_dim=[z2s_dim],
nout=z2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
rnn = LSTM(name='rnn',
parent=['x_4', 'z_4'],
parent_dim=[x2s_dim, z2s_dim],
nout=rnn_dim,
unit='tanh',
init_W=init_W,
init_U=init_U,
init_b=init_b)
phi_1 = FullyConnectedLayer(name='phi_1',
parent=['x_4', 's_tm1'],
parent_dim=[x2s_dim, rnn_dim],
nout=q_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
phi_2 = FullyConnectedLayer(name='phi_2',
parent=['phi_1'],
parent_dim=[q_z_dim],
nout=q_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
phi_3 = FullyConnectedLayer(name='phi_3',
parent=['phi_2'],
parent_dim=[q_z_dim],
nout=q_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
phi_4 = FullyConnectedLayer(name='phi_4',
parent=['phi_3'],
parent_dim=[q_z_dim],
nout=q_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
phi_mu = FullyConnectedLayer(name='phi_mu',
parent=['phi_4'],
parent_dim=[q_z_dim],
nout=z_dim,
unit='linear',
init_W=init_W,
init_b=init_b)
phi_sig = FullyConnectedLayer(name='phi_sig',
parent=['phi_4'],
parent_dim=[q_z_dim],
nout=z_dim,
unit='softplus',
cons=1e-4,
init_W=init_W,
init_b=init_b_sig)
prior_1 = FullyConnectedLayer(name='prior_1',
parent=['s_tm1'],
parent_dim=[rnn_dim],
nout=p_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
prior_2 = FullyConnectedLayer(name='prior_2',
parent=['prior_1'],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
prior_3 = FullyConnectedLayer(name='prior_3',
parent=['prior_2'],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
prior_4 = FullyConnectedLayer(name='prior_4',
parent=['prior_3'],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
prior_mu = FullyConnectedLayer(name='prior_mu',
parent=['prior_4'],
parent_dim=[p_z_dim],
nout=z_dim,
unit='linear',
init_W=init_W,
init_b=init_b)
prior_sig = FullyConnectedLayer(name='prior_sig',
parent=['prior_4'],
parent_dim=[p_z_dim],
nout=z_dim,
unit='softplus',
cons=1e-4,
init_W=init_W,
init_b=init_b_sig)
theta_1 = FullyConnectedLayer(name='theta_1',
parent=['z_4', 's_tm1'],
parent_dim=[z2s_dim, rnn_dim],
nout=p_x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_2 = FullyConnectedLayer(name='theta_2',
parent=['theta_1'],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_3 = FullyConnectedLayer(name='theta_3',
parent=['theta_2'],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_4 = FullyConnectedLayer(name='theta_4',
parent=['theta_3'],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_mu = FullyConnectedLayer(name='theta_mu',
parent=['theta_4'],
parent_dim=[p_x_dim],
nout=target_dim,
unit='linear',
init_W=init_W,
init_b=init_b)
theta_sig = FullyConnectedLayer(name='theta_sig',
parent=['theta_4'],
parent_dim=[p_x_dim],
nout=target_dim,
unit='softplus',
cons=1e-4,
init_W=init_W,
init_b=init_b_sig)
nodes = [
rnn, x_1, x_2, x_3, x_4, z_1, z_2, z_3, z_4, phi_1, phi_2, phi_3,
phi_4, phi_mu, phi_sig, prior_1, prior_2, prior_3, prior_4, prior_mu,
prior_sig, theta_1, theta_2, theta_3, theta_4, theta_mu, theta_sig
]
params = OrderedDict()
for node in nodes:
if node.initialize() is not None:
params.update(node.initialize())
params = init_tparams(params)
step_count = sharedX(0, name='step_count')
last_rnn = np.zeros((batch_size, rnn_dim * 2), dtype=theano.config.floatX)
rnn_tm1 = sharedX(last_rnn, name='rnn_tm1')
shared_updates = OrderedDict()
shared_updates[step_count] = step_count + 1
# Resets / Initializes the cell-state or the memory-state of each LSTM to
# zero.
s_0 = T.switch(T.eq(T.mod(step_count, reset_freq), 0),
rnn.get_init_state(batch_size), rnn_tm1)
# Forward Propagate the input to get more complex features for
# every time step.
x_1_temp = x_1.fprop([x], params)
x_2_temp = x_2.fprop([x_1_temp], params)
x_3_temp = x_3.fprop([x_2_temp], params)
x_4_temp = x_4.fprop([x_3_temp], params)
def inner_fn(x_t, s_tm1):
# Generate the mean and standard deviation of the
# latent variables Z_t | X_t for every time-step of the LSTM.
# This is a function of the input and the hidden state of the previous
# time step.
phi_1_t = phi_1.fprop([x_t, s_tm1], params)
phi_2_t = phi_2.fprop([phi_1_t], params)
phi_3_t = phi_3.fprop([phi_2_t], params)
phi_4_t = phi_4.fprop([phi_3_t], params)
phi_mu_t = phi_mu.fprop([phi_4_t], params)
phi_sig_t = phi_sig.fprop([phi_4_t], params)
# Prior on the latent variables at every time-step
# Dependent only on the hidden-step.
prior_1_t = prior_1.fprop([s_tm1], params)
prior_2_t = prior_2.fprop([prior_1_t], params)
prior_3_t = prior_3.fprop([prior_2_t], params)
prior_4_t = prior_4.fprop([prior_3_t], params)
prior_mu_t = prior_mu.fprop([prior_4_t], params)
prior_sig_t = prior_sig.fprop([prior_4_t], params)
# Sample from the latent distibution with mean phi_mu_t
# and std phi_sig_t
z_t = Gaussian_sample(phi_mu_t, phi_sig_t)
# h_t = f(h_(t-1)), z_t, x_t)
z_1_t = z_1.fprop([z_t], params)
z_2_t = z_2.fprop([z_1_t], params)
z_3_t = z_3.fprop([z_2_t], params)
z_4_t = z_4.fprop([z_3_t], params)
s_t = rnn.fprop([[x_t, z_4_t], [s_tm1]], params)
return s_t, phi_mu_t, phi_sig_t, prior_mu_t, prior_sig_t, z_4_t
# Iterate over every time-step
((s_temp, phi_mu_temp, phi_sig_temp, prior_mu_temp, prior_sig_temp, z_4_temp), updates) =\
theano.scan(fn=inner_fn,
sequences=[x_4_temp],
outputs_info=[s_0, None, None, None, None, None])
for k, v in updates.iteritems():
k.default_update = v
shared_updates[rnn_tm1] = s_temp[-1]
s_temp = concatenate([s_0[None, :, :], s_temp[:-1]], axis=0)
# Generate the output distribution at every time-step.
# This is as a function of the latent variables and the hidden-state at
# every time-step.
theta_1_temp = theta_1.fprop([z_4_temp, s_temp], params)
theta_2_temp = theta_2.fprop([theta_1_temp], params)
theta_3_temp = theta_3.fprop([theta_2_temp], params)
theta_4_temp = theta_4.fprop([theta_3_temp], params)
theta_mu_temp = theta_mu.fprop([theta_4_temp], params)
theta_sig_temp = theta_sig.fprop([theta_4_temp], params)
kl_temp = KLGaussianGaussian(phi_mu_temp, phi_sig_temp, prior_mu_temp,
prior_sig_temp)
recon = Gaussian(x, theta_mu_temp, theta_sig_temp)
recon_term = recon.mean()
kl_term = kl_temp.mean()
nll_upper_bound = recon_term + kl_term
nll_upper_bound.name = 'nll_upper_bound'
# Forward-propagation of the validation data.
m_x_1_temp = x_1.fprop([m_x], params)
m_x_2_temp = x_2.fprop([m_x_1_temp], params)
m_x_3_temp = x_3.fprop([m_x_2_temp], params)
m_x_4_temp = x_4.fprop([m_x_3_temp], params)
m_s_0 = rnn.get_init_state(m_batch_size)
# Get the hidden-states, conditional mean, standard deviation, prior mean
# and prior standard deviation of the latent variables at every time-step.
((m_s_temp, m_phi_mu_temp, m_phi_sig_temp, m_prior_mu_temp, m_prior_sig_temp, m_z_4_temp), m_updates) =\
theano.scan(fn=inner_fn,
sequences=[m_x_4_temp],
outputs_info=[m_s_0, None, None, None, None, None])
for k, v in m_updates.iteritems():
k.default_update = v
# Get the inferred mean (X_t | Z_t) at every time-step of the validation
# data.
m_s_temp = concatenate([m_s_0[None, :, :], m_s_temp[:-1]], axis=0)
m_theta_1_temp = theta_1.fprop([m_z_4_temp, m_s_temp], params)
m_theta_2_temp = theta_2.fprop([m_theta_1_temp], params)
m_theta_3_temp = theta_3.fprop([m_theta_2_temp], params)
m_theta_4_temp = theta_4.fprop([m_theta_3_temp], params)
m_theta_mu_temp = theta_mu.fprop([m_theta_4_temp], params)
m_theta_sig_temp = theta_sig.fprop([m_theta_4_temp], params)
# Compute the data log-likelihood + KL-divergence on the validation data.
m_kl_temp = KLGaussianGaussian(m_phi_mu_temp, m_phi_sig_temp,
m_prior_mu_temp, m_prior_sig_temp)
m_recon = Gaussian(m_x, m_theta_mu_temp, m_theta_sig_temp)
m_recon_term = m_recon.mean()
m_kl_term = m_kl_temp.mean()
m_nll_upper_bound = m_recon_term + m_kl_term
m_nll_upper_bound.name = 'nll_upper_bound'
m_recon_term.name = 'recon_term'
m_kl_term.name = 'kl_term'
max_x = m_x.max()
mean_x = m_x.mean()
min_x = m_x.min()
max_x.name = 'max_x'
mean_x.name = 'mean_x'
min_x.name = 'min_x'
max_theta_mu = m_theta_mu_temp.max()
mean_theta_mu = m_theta_mu_temp.mean()
min_theta_mu = m_theta_mu_temp.min()
max_theta_mu.name = 'max_theta_mu'
mean_theta_mu.name = 'mean_theta_mu'
min_theta_mu.name = 'min_theta_mu'
max_theta_sig = m_theta_sig_temp.max()
mean_theta_sig = m_theta_sig_temp.mean()
min_theta_sig = m_theta_sig_temp.min()
max_theta_sig.name = 'max_theta_sig'
mean_theta_sig.name = 'mean_theta_sig'
min_theta_sig.name = 'min_theta_sig'
max_phi_sig = m_phi_sig_temp.max()
mean_phi_sig = m_phi_sig_temp.mean()
min_phi_sig = m_phi_sig_temp.min()
max_phi_sig.name = 'max_phi_sig'
mean_phi_sig.name = 'mean_phi_sig'
min_phi_sig.name = 'min_phi_sig'
max_prior_sig = m_prior_sig_temp.max()
mean_prior_sig = m_prior_sig_temp.mean()
min_prior_sig = m_prior_sig_temp.min()
max_prior_sig.name = 'max_prior_sig'
mean_prior_sig.name = 'mean_prior_sig'
min_prior_sig.name = 'min_prior_sig'
model.inputs = [x]
model.params = params
model.nodes = nodes
model.set_updates(shared_updates)
optimizer = Adam(lr=lr)
monitor_fn = theano.function(
inputs=[m_x],
outputs=[
m_nll_upper_bound, m_recon_term, m_kl_term, max_phi_sig,
mean_phi_sig, min_phi_sig, max_prior_sig, mean_prior_sig,
min_prior_sig, max_theta_sig, mean_theta_sig, min_theta_sig, max_x,
mean_x, min_x, max_theta_mu, mean_theta_mu, min_theta_mu
],
on_unused_input='ignore')
extension = [
GradientClipping(batch_size=batch_size, check_nan=1),
EpochCount(epoch),
Monitoring(freq=monitoring_freq,
monitor_fn=monitor_fn,
ddout=[
m_nll_upper_bound, m_recon_term, m_kl_term, max_phi_sig,
mean_phi_sig, min_phi_sig, max_prior_sig,
mean_prior_sig, min_prior_sig, max_theta_sig,
mean_theta_sig, min_theta_sig, max_x, mean_x, min_x,
max_theta_mu, mean_theta_mu, min_theta_mu
],
data=[
Iterator(train_data, m_batch_size, start=0, end=112640),
Iterator(valid_data,
m_batch_size,
start=2040064,
end=2152704)
]),
Picklize(freq=monitoring_freq,
force_save_freq=force_saving_freq,
path=save_path),
EarlyStopping(freq=monitoring_freq,
force_save_freq=force_saving_freq,
path=save_path,
channel=channel_name),
WeightNorm()
]
mainloop = Training(name=pkl_name,
data=Iterator(train_data,
batch_size,
start=0,
end=2040064),
model=model,
optimizer=optimizer,
cost=nll_upper_bound,
outputs=[nll_upper_bound],
extension=extension)
mainloop.run()
def setX(self, x, name=None):
return sharedX(x, name)
parent_dim=[decoder_dim],
nout=target_size,
unit='softplus',
cons=1e-4,
init_W=init_W,
init_b=init_b_sig)
nodes = [encoder, decoder, prior, kl,
phi_mu, phi_sig, theta_mu, theta_sig]
for node in nodes:
node.initialize()
params = flatten([node.get_params().values() for node in nodes])
step_count = sharedX(0, name='step_count')
last_encoder = np.zeros((batch_size, encoder_dim*2), dtype=theano.config.floatX)
last_decoder = np.zeros((batch_size, decoder_dim*2), dtype=theano.config.floatX)
encoder_tm1 = sharedX(last_encoder, name='encoder_tm1')
decoder_tm1 = sharedX(last_decoder, name='decoder_tm1')
update_list = [step_count, encoder_tm1, decoder_tm1]
step_count = T.switch(T.le(step_count, reset_freq), step_count + 1, 0)
enc_0 = T.switch(T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(encoder_tm1), 0.), 'int32')),
encoder.get_init_state(batch_size), encoder_tm1)
dec_0 = T.switch(T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(decoder_tm1), 0.), 'int32')),
decoder.get_init_state(batch_size), decoder_tm1)
x_shape = x.shape
def main(args):
trial = int(args['trial'])
pkl_name = 'rnn_gauss_%d' % trial
channel_name = 'valid_nll'
data_path = args['data_path']
save_path = args['save_path']
monitoring_freq = int(args['monitoring_freq'])
force_saving_freq = int(args['force_saving_freq'])
reset_freq = int(args['reset_freq'])
epoch = int(args['epoch'])
batch_size = int(args['batch_size'])
m_batch_size = int(args['m_batch_size'])
x_dim = int(args['x_dim'])
z_dim = int(args['z_dim'])
rnn_dim = int(args['rnn_dim'])
lr = float(args['lr'])
debug = int(args['debug'])
print "trial no. %d" % trial
print "batch size %d" % batch_size
print "learning rate %f" % lr
print "saving pkl file '%s'" % pkl_name
print "to the save path '%s'" % save_path
x2s_dim = 800
s2x_dim = 800
target_dim = x_dim
file_name = 'blizzard_unseg_tbptt'
normal_params = np.load(data_path + file_name + '_normal.npz')
X_mean = normal_params['X_mean']
X_std = normal_params['X_std']
model = Model()
train_data = Blizzard_tbptt(name='train',
path=data_path,
frame_size=x_dim,
file_name=file_name,
X_mean=X_mean,
X_std=X_std)
valid_data = Blizzard_tbptt(name='valid',
path=data_path,
frame_size=x_dim,
file_name=file_name,
X_mean=X_mean,
X_std=X_std)
x = train_data.theano_vars()
m_x = valid_data.theano_vars()
if debug:
x.tag.test_value = np.zeros((15, batch_size, x_dim),
dtype=theano.config.floatX)
m_x.tag.test_value = np.zeros((15, m_batch_size, x_dim),
dtype=theano.config.floatX)
init_W = InitCell('rand')
init_U = InitCell('ortho')
init_b = InitCell('zeros')
init_b_sig = InitCell('const', mean=0.6)
x_1 = FullyConnectedLayer(name='x_1',
parent=['x_t'],
parent_dim=[x_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_2 = FullyConnectedLayer(name='x_2',
parent=['x_1'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_3 = FullyConnectedLayer(name='x_3',
parent=['x_2'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_4 = FullyConnectedLayer(name='x_4',
parent=['x_3'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
rnn = LSTM(name='rnn',
parent=['x_4'],
parent_dim=[x2s_dim],
nout=rnn_dim,
unit='tanh',
init_W=init_W,
init_U=init_U,
init_b=init_b)
theta_1 = FullyConnectedLayer(name='theta_1',
parent=['s_tm1'],
parent_dim=[rnn_dim],
nout=s2x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_2 = FullyConnectedLayer(name='theta_2',
parent=['theta_1'],
parent_dim=[s2x_dim],
nout=s2x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_3 = FullyConnectedLayer(name='theta_3',
parent=['theta_2'],
parent_dim=[s2x_dim],
nout=s2x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_4 = FullyConnectedLayer(name='theta_4',
parent=['theta_3'],
parent_dim=[s2x_dim],
nout=s2x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_mu = FullyConnectedLayer(name='theta_mu',
parent=['theta_4'],
parent_dim=[s2x_dim],
nout=target_dim,
unit='linear',
init_W=init_W,
init_b=init_b)
theta_sig = FullyConnectedLayer(name='theta_sig',
parent=['theta_4'],
parent_dim=[s2x_dim],
nout=target_dim,
unit='softplus',
cons=1e-4,
init_W=init_W,
init_b=init_b_sig)
nodes = [
rnn, x_1, x_2, x_3, x_4, theta_1, theta_2, theta_3, theta_4, theta_mu,
theta_sig
]
params = OrderedDict()
for node in nodes:
if node.initialize() is not None:
params.update(node.initialize())
params = init_tparams(params)
step_count = sharedX(0, name='step_count')
last_rnn = np.zeros((batch_size, rnn_dim * 2), dtype=theano.config.floatX)
rnn_tm1 = sharedX(last_rnn, name='rnn_tm1')
shared_updates = OrderedDict()
shared_updates[step_count] = step_count + 1
s_0 = T.switch(T.eq(T.mod(step_count, reset_freq), 0),
rnn.get_init_state(batch_size), rnn_tm1)
x_1_temp = x_1.fprop([x], params)
x_2_temp = x_2.fprop([x_1_temp], params)
x_3_temp = x_3.fprop([x_2_temp], params)
x_4_temp = x_4.fprop([x_3_temp], params)
def inner_fn(x_t, s_tm1):
s_t = rnn.fprop([[x_t], [s_tm1]], params)
return s_t
(s_temp, updates) = theano.scan(fn=inner_fn,
sequences=[x_4_temp],
outputs_info=[s_0])
for k, v in updates.iteritems():
k.default_update = v
shared_updates[rnn_tm1] = s_temp[-1]
s_temp = concatenate([s_0[None, :, :], s_temp[:-1]], axis=0)
theta_1_temp = theta_1.fprop([s_temp], params)
theta_2_temp = theta_2.fprop([theta_1_temp], params)
theta_3_temp = theta_3.fprop([theta_2_temp], params)
theta_4_temp = theta_4.fprop([theta_3_temp], params)
theta_mu_temp = theta_mu.fprop([theta_4_temp], params)
theta_sig_temp = theta_sig.fprop([theta_4_temp], params)
recon = Gaussian(x, theta_mu_temp, theta_sig_temp)
recon_term = recon.mean()
recon_term.name = 'nll'
m_x_1_temp = x_1.fprop([m_x], params)
m_x_2_temp = x_2.fprop([m_x_1_temp], params)
m_x_3_temp = x_3.fprop([m_x_2_temp], params)
m_x_4_temp = x_4.fprop([m_x_3_temp], params)
m_s_0 = rnn.get_init_state(m_batch_size)
(m_s_temp, m_updates) = theano.scan(fn=inner_fn,
sequences=[m_x_4_temp],
outputs_info=[m_s_0])
for k, v in m_updates.iteritems():
k.default_update = v
m_s_temp = concatenate([m_s_0[None, :, :], m_s_temp[:-1]], axis=0)
m_theta_1_temp = theta_1.fprop([m_s_temp], params)
m_theta_2_temp = theta_2.fprop([m_theta_1_temp], params)
m_theta_3_temp = theta_3.fprop([m_theta_2_temp], params)
m_theta_4_temp = theta_4.fprop([m_theta_3_temp], params)
m_theta_mu_temp = theta_mu.fprop([m_theta_4_temp], params)
m_theta_sig_temp = theta_sig.fprop([m_theta_4_temp], params)
m_recon = Gaussian(m_x, m_theta_mu_temp, m_theta_sig_temp)
m_recon_term = m_recon.mean()
m_recon_term.name = 'nll'
max_x = m_x.max()
mean_x = m_x.mean()
min_x = m_x.min()
max_x.name = 'max_x'
mean_x.name = 'mean_x'
min_x.name = 'min_x'
max_theta_mu = m_theta_mu_temp.max()
mean_theta_mu = m_theta_mu_temp.mean()
min_theta_mu = m_theta_mu_temp.min()
max_theta_mu.name = 'max_theta_mu'
mean_theta_mu.name = 'mean_theta_mu'
min_theta_mu.name = 'min_theta_mu'
max_theta_sig = m_theta_sig_temp.max()
mean_theta_sig = m_theta_sig_temp.mean()
min_theta_sig = m_theta_sig_temp.min()
max_theta_sig.name = 'max_theta_sig'
mean_theta_sig.name = 'mean_theta_sig'
min_theta_sig.name = 'min_theta_sig'
model.inputs = [x]
model.params = params
model.nodes = nodes
model.set_updates(shared_updates)
optimizer = Adam(lr=lr)
monitor_fn = theano.function(inputs=[m_x],
outputs=[
m_recon_term, max_theta_sig,
mean_theta_sig, min_theta_sig, max_x,
mean_x, min_x, max_theta_mu,
mean_theta_mu, min_theta_mu
],
on_unused_input='ignore')
extension = [
GradientClipping(batch_size=batch_size, check_nan=1),
EpochCount(epoch),
Monitoring(freq=monitoring_freq,
monitor_fn=monitor_fn,
ddout=[
m_recon_term, max_theta_sig, mean_theta_sig,
min_theta_sig, max_x, mean_x, min_x, max_theta_mu,
mean_theta_mu, min_theta_mu
],
data=[
Iterator(train_data, m_batch_size, start=0, end=112640),
Iterator(valid_data,
m_batch_size,
start=2040064,
end=2152704)
]),
Picklize(freq=monitoring_freq,
force_save_freq=force_saving_freq,
path=save_path),
EarlyStopping(freq=monitoring_freq,
force_save_freq=force_saving_freq,
path=save_path,
channel=channel_name),
WeightNorm()
]
mainloop = Training(name=pkl_name,
data=Iterator(train_data,
batch_size,
start=0,
end=2040064),
model=model,
optimizer=optimizer,
cost=recon_term,
outputs=[recon_term],
extension=extension)
mainloop.run()
init_W=init_W,
init_b=init_b_sig)
nodes = [main_lstm, prior, kl,
x_1, x_2, x_3, x_4,
z_1, z_2, z_3, z_4,
phi_1, phi_2, phi_3, phi_4, phi_mu, phi_sig,
prior_1, prior_2, prior_3, prior_4, prior_mu, prior_sig,
theta_1, theta_2, theta_3, theta_4, theta_mu, theta_sig]
for node in nodes:
node.initialize()
params = flatten([node.get_params().values() for node in nodes])
step_count = sharedX(0, name='step_count')
last_main_lstm = np.zeros((batch_size, main_lstm_dim*2), dtype=theano.config.floatX)
main_lstm_tm1 = sharedX(last_main_lstm, name='main_lstm_tm1')
update_list = [step_count, main_lstm_tm1]
step_count = T.switch(T.le(step_count, reset_freq), step_count + 1, 0)
s_0 = T.switch(T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(main_lstm_tm1), 0.), 'int32')),
main_lstm.get_init_state(batch_size), main_lstm_tm1)
x_shape = x.shape
x_in = x.reshape((x_shape[0]*x_shape[1], -1))
x_1_in = x_1.fprop([x_in])
x_2_in = x_2.fprop([x_1_in])
x_3_in = x_3.fprop([x_2_in])
x_4_in = x_4.fprop([x_3_in])
def main(args):
trial = int(args['trial'])
pkl_name = 'vrnn_gmm_%d' % trial
channel_name = 'valid_nll_upper_bound'
data_path = args['data_path']
save_path = args['save_path']
monitoring_freq = int(args['monitoring_freq'])
force_saving_freq = int(args['force_saving_freq'])
reset_freq = int(args['reset_freq'])
epoch = int(args['epoch'])
batch_size = int(args['batch_size'])
m_batch_size = int(args['m_batch_size'])
x_dim = int(args['x_dim'])
z_dim = int(args['z_dim'])
rnn_dim = int(args['rnn_dim'])
k = int(args['num_k'])
lr = float(args['lr'])
debug = int(args['debug'])
print "trial no. %d" % trial
print "batch size %d" % batch_size
print "learning rate %f" % lr
print "saving pkl file '%s'" % pkl_name
print "to the save path '%s'" % save_path
q_z_dim = 500
p_z_dim = 500
p_x_dim = 500
x2s_dim = 500
z2s_dim = 500
target_dim = x_dim * k
file_name = 'blizzard_unseg_tbptt'
normal_params = np.load(data_path + file_name + '_normal.npz')
X_mean = normal_params['X_mean']
X_std = normal_params['X_std']
model = Model()
train_data = Blizzard_tbptt(name='train',
path=data_path,
frame_size=x_dim,
file_name=file_name,
X_mean=X_mean,
X_std=X_std)
valid_data = Blizzard_tbptt(name='valid',
path=data_path,
frame_size=x_dim,
file_name=file_name,
X_mean=X_mean,
X_std=X_std)
x = train_data.theano_vars()
m_x = valid_data.theano_vars()
if debug:
x.tag.test_value = np.zeros((15, batch_size, x_dim), dtype=theano.config.floatX)
m_x.tag.test_value = np.zeros((15, m_batch_size, x_dim), dtype=theano.config.floatX)
init_W = InitCell('rand')
init_U = InitCell('ortho')
init_b = InitCell('zeros')
init_b_sig = InitCell('const', mean=0.6)
x_1 = FullyConnectedLayer(name='x_1',
parent=['x_t'],
parent_dim=[x_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_2 = FullyConnectedLayer(name='x_2',
parent=['x_1'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_3 = FullyConnectedLayer(name='x_3',
parent=['x_2'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_4 = FullyConnectedLayer(name='x_4',
parent=['x_3'],
parent_dim=[x2s_dim],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
z_1 = FullyConnectedLayer(name='z_1',
parent=['z_t'],
parent_dim=[z_dim],
nout=z2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
z_2 = FullyConnectedLayer(name='z_2',
parent=['z_1'],
parent_dim=[z2s_dim],
nout=z2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
z_3 = FullyConnectedLayer(name='z_3',
parent=['z_2'],
parent_dim=[z2s_dim],
nout=z2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
z_4 = FullyConnectedLayer(name='z_4',
parent=['z_3'],
parent_dim=[z2s_dim],
nout=z2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
rnn = LSTM(name='rnn',
parent=['x_4', 'z_4'],
parent_dim=[x2s_dim, z2s_dim],
nout=rnn_dim,
unit='tanh',
init_W=init_W,
init_U=init_U,
init_b=init_b)
phi_1 = FullyConnectedLayer(name='phi_1',
parent=['x_4', 's_tm1'],
parent_dim=[x2s_dim, rnn_dim],
nout=q_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
phi_2 = FullyConnectedLayer(name='phi_2',
parent=['phi_1'],
parent_dim=[q_z_dim],
nout=q_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
phi_3 = FullyConnectedLayer(name='phi_3',
parent=['phi_2'],
parent_dim=[q_z_dim],
nout=q_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
phi_4 = FullyConnectedLayer(name='phi_4',
parent=['phi_3'],
parent_dim=[q_z_dim],
nout=q_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
phi_mu = FullyConnectedLayer(name='phi_mu',
parent=['phi_4'],
parent_dim=[q_z_dim],
nout=z_dim,
unit='linear',
init_W=init_W,
init_b=init_b)
phi_sig = FullyConnectedLayer(name='phi_sig',
parent=['phi_4'],
parent_dim=[q_z_dim],
nout=z_dim,
unit='softplus',
cons=1e-4,
init_W=init_W,
init_b=init_b_sig)
prior_1 = FullyConnectedLayer(name='prior_1',
parent=['s_tm1'],
parent_dim=[rnn_dim],
nout=p_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
prior_2 = FullyConnectedLayer(name='prior_2',
parent=['prior_1'],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
prior_3 = FullyConnectedLayer(name='prior_3',
parent=['prior_2'],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
prior_4 = FullyConnectedLayer(name='prior_4',
parent=['prior_3'],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
prior_mu = FullyConnectedLayer(name='prior_mu',
parent=['prior_4'],
parent_dim=[p_z_dim],
nout=z_dim,
unit='linear',
init_W=init_W,
init_b=init_b)
prior_sig = FullyConnectedLayer(name='prior_sig',
parent=['prior_4'],
parent_dim=[p_z_dim],
nout=z_dim,
unit='softplus',
cons=1e-4,
init_W=init_W,
init_b=init_b_sig)
theta_1 = FullyConnectedLayer(name='theta_1',
parent=['z_4', 's_tm1'],
parent_dim=[z2s_dim, rnn_dim],
nout=p_x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_2 = FullyConnectedLayer(name='theta_2',
parent=['theta_1'],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_3 = FullyConnectedLayer(name='theta_3',
parent=['theta_2'],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_4 = FullyConnectedLayer(name='theta_4',
parent=['theta_3'],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
theta_mu = FullyConnectedLayer(name='theta_mu',
parent=['theta_4'],
parent_dim=[p_x_dim],
nout=target_dim,
unit='linear',
init_W=init_W,
init_b=init_b)
theta_sig = FullyConnectedLayer(name='theta_sig',
parent=['theta_4'],
parent_dim=[p_x_dim],
nout=target_dim,
unit='softplus',
cons=1e-4,
init_W=init_W,
init_b=init_b_sig)
coeff = FullyConnectedLayer(name='coeff',
parent=['theta_4'],
parent_dim=[p_x_dim],
nout=k,
unit='softmax',
init_W=init_W,
init_b=init_b)
nodes = [rnn,
x_1, x_2, x_3, x_4,
z_1, z_2, z_3, z_4,
phi_1, phi_2, phi_3, phi_4, phi_mu, phi_sig,
prior_1, prior_2, prior_3, prior_4, prior_mu, prior_sig,
theta_1, theta_2, theta_3, theta_4, theta_mu, theta_sig, coeff]
params = OrderedDict()
for node in nodes:
if node.initialize() is not None:
params.update(node.initialize())
params = init_tparams(params)
step_count = sharedX(0, name='step_count')
last_rnn = np.zeros((batch_size, rnn_dim*2), dtype=theano.config.floatX)
rnn_tm1 = sharedX(last_rnn, name='rnn_tm1')
shared_updates = OrderedDict()
shared_updates[step_count] = step_count + 1
s_0 = T.switch(T.eq(T.mod(step_count, reset_freq), 0),
rnn.get_init_state(batch_size), rnn_tm1)
x_shape = x.shape
x_in = x.reshape((x_shape[0]*x_shape[1], -1))
x_1_in = x_1.fprop([x_in], params)
x_2_in = x_2.fprop([x_1_in], params)
x_3_in = x_3.fprop([x_2_in], params)
x_4_in = x_4.fprop([x_3_in], params)
x_4_in = x_4_in.reshape((x_shape[0], x_shape[1], -1))
def inner_fn(x_t, s_tm1):
phi_1_t = phi_1.fprop([x_t, s_tm1], params)
phi_2_t = phi_2.fprop([phi_1_t], params)
phi_3_t = phi_3.fprop([phi_2_t], params)
phi_4_t = phi_4.fprop([phi_3_t], params)
phi_mu_t = phi_mu.fprop([phi_4_t], params)
phi_sig_t = phi_sig.fprop([phi_4_t], params)
prior_1_t = prior_1.fprop([s_tm1], params)
prior_2_t = prior_2.fprop([prior_1_t], params)
prior_3_t = prior_3.fprop([prior_2_t], params)
prior_4_t = prior_4.fprop([prior_3_t], params)
prior_mu_t = prior_mu.fprop([prior_4_t], params)
prior_sig_t = prior_sig.fprop([prior_4_t], params)
z_t = Gaussian_sample(phi_mu_t, phi_sig_t)
z_1_t = z_1.fprop([z_t], params)
z_2_t = z_2.fprop([z_1_t], params)
z_3_t = z_3.fprop([z_2_t], params)
z_4_t = z_4.fprop([z_3_t], params)
s_t = rnn.fprop([[x_t, z_4_t], [s_tm1]], params)
return s_t, phi_mu_t, phi_sig_t, prior_mu_t, prior_sig_t, z_4_t
((s_temp, phi_mu_temp, phi_sig_temp, prior_mu_temp, prior_sig_temp, z_4_temp), updates) =\
theano.scan(fn=inner_fn,
sequences=[x_4_in],
outputs_info=[s_0, None, None, None, None, None])
for k, v in updates.iteritems():
k.default_update = v
shared_updates[rnn_tm1] = s_temp[-1]
s_temp = concatenate([s_0[None, :, :], s_temp[:-1]], axis=0)
theta_1_temp = theta_1.fprop([z_4_temp, s_temp], params)
theta_2_temp = theta_2.fprop([theta_1_temp], params)
theta_3_temp = theta_3.fprop([theta_2_temp], params)
theta_4_temp = theta_4.fprop([theta_3_temp], params)
theta_mu_temp = theta_mu.fprop([theta_4_temp], params)
theta_sig_temp = theta_sig.fprop([theta_4_temp], params)
coeff_temp = coeff.fprop([theta_4_temp], params)
kl_temp = KLGaussianGaussian(phi_mu_temp, phi_sig_temp, prior_mu_temp, prior_sig_temp)
x_shape = x.shape
x_in = x.reshape((x_shape[0]*x_shape[1], -1))
theta_mu_in = theta_mu_temp.reshape((x_shape[0]*x_shape[1], -1))
theta_sig_in = theta_sig_temp.reshape((x_shape[0]*x_shape[1], -1))
coeff_in = coeff_temp.reshape((x_shape[0]*x_shape[1], -1))
recon = GMM(x_in, theta_mu_in, theta_sig_in, coeff_in)
recon_term = recon.mean()
kl_term = kl_temp.mean()
nll_upper_bound = recon_term + kl_term
nll_upper_bound.name = 'nll_upper_bound'
m_x_1_temp = x_1.fprop([m_x], params)
m_x_2_temp = x_2.fprop([m_x_1_temp], params)
m_x_3_temp = x_3.fprop([m_x_2_temp], params)
m_x_4_temp = x_4.fprop([m_x_3_temp], params)
m_s_0 = rnn.get_init_state(m_batch_size)
((m_s_temp, m_phi_mu_temp, m_phi_sig_temp, m_prior_mu_temp, m_prior_sig_temp, m_z_4_temp), m_updates) =\
theano.scan(fn=inner_fn,
sequences=[m_x_4_temp],
outputs_info=[m_s_0, None, None, None, None, None])
for k, v in m_updates.iteritems():
k.default_update = v
m_s_temp = concatenate([m_s_0[None, :, :], m_s_temp[:-1]], axis=0)
m_theta_1_temp = theta_1.fprop([m_z_4_temp, m_s_temp], params)
m_theta_2_temp = theta_2.fprop([m_theta_1_temp], params)
m_theta_3_temp = theta_3.fprop([m_theta_2_temp], params)
m_theta_4_temp = theta_4.fprop([m_theta_3_temp], params)
m_theta_mu_temp = theta_mu.fprop([m_theta_4_temp], params)
m_theta_sig_temp = theta_sig.fprop([m_theta_4_temp], params)
m_coeff_temp = coeff.fprop([m_theta_4_temp], params)
m_kl_temp = KLGaussianGaussian(m_phi_mu_temp, m_phi_sig_temp, m_prior_mu_temp, m_prior_sig_temp)
m_x_shape = m_x.shape
m_x_in = m_x.reshape((m_x_shape[0]*m_x_shape[1], -1))
m_theta_mu_in = m_theta_mu_temp.reshape((m_x_shape[0]*m_x_shape[1], -1))
m_theta_sig_in = m_theta_sig_temp.reshape((m_x_shape[0]*m_x_shape[1], -1))
m_coeff_in = m_coeff_temp.reshape((m_x_shape[0]*m_x_shape[1], -1))
m_recon = GMM(m_x_in, m_theta_mu_in, m_theta_sig_in, m_coeff_in)
m_recon_term = m_recon.mean()
m_kl_term = m_kl_temp.mean()
m_nll_upper_bound = m_recon_term + m_kl_term
m_nll_upper_bound.name = 'nll_upper_bound'
m_recon_term.name = 'recon_term'
m_kl_term.name = 'kl_term'
max_x = m_x.max()
mean_x = m_x.mean()
min_x = m_x.min()
max_x.name = 'max_x'
mean_x.name = 'mean_x'
min_x.name = 'min_x'
max_theta_mu = m_theta_mu_in.max()
mean_theta_mu = m_theta_mu_in.mean()
min_theta_mu = m_theta_mu_in.min()
max_theta_mu.name = 'max_theta_mu'
mean_theta_mu.name = 'mean_theta_mu'
min_theta_mu.name = 'min_theta_mu'
max_theta_sig = m_theta_sig_in.max()
mean_theta_sig = m_theta_sig_in.mean()
min_theta_sig = m_theta_sig_in.min()
max_theta_sig.name = 'max_theta_sig'
mean_theta_sig.name = 'mean_theta_sig'
min_theta_sig.name = 'min_theta_sig'
max_phi_sig = m_phi_sig_temp.max()
mean_phi_sig = m_phi_sig_temp.mean()
min_phi_sig = m_phi_sig_temp.min()
max_phi_sig.name = 'max_phi_sig'
mean_phi_sig.name = 'mean_phi_sig'
min_phi_sig.name = 'min_phi_sig'
max_prior_sig = m_prior_sig_temp.max()
mean_prior_sig = m_prior_sig_temp.mean()
min_prior_sig = m_prior_sig_temp.min()
max_prior_sig.name = 'max_prior_sig'
mean_prior_sig.name = 'mean_prior_sig'
min_prior_sig.name = 'min_prior_sig'
model.inputs = [x]
model.params = params
model.nodes = nodes
model.set_updates(shared_updates)
optimizer = Adam(
lr=lr
)
monitor_fn = theano.function(inputs=[m_x],
outputs=[m_nll_upper_bound, m_recon_term, m_kl_term,
max_phi_sig, mean_phi_sig, min_phi_sig,
max_prior_sig, mean_prior_sig, min_prior_sig,
max_theta_sig, mean_theta_sig, min_theta_sig,
max_x, mean_x, min_x,
max_theta_mu, mean_theta_mu, min_theta_mu],
on_unused_input='ignore')
extension = [
GradientClipping(batch_size=batch_size, check_nan=1),
EpochCount(epoch),
Monitoring(freq=monitoring_freq,
monitor_fn=monitor_fn,
ddout=[m_nll_upper_bound, m_recon_term, m_kl_term,
max_phi_sig, mean_phi_sig, min_phi_sig,
max_prior_sig, mean_prior_sig, min_prior_sig,
max_theta_sig, mean_theta_sig, min_theta_sig,
max_x, mean_x, min_x,
max_theta_mu, mean_theta_mu, min_theta_mu],
data=[Iterator(train_data, m_batch_size, start=0, end=112640),
Iterator(valid_data, m_batch_size, start=2040064, end=2152704)]),
Picklize(freq=monitoring_freq, force_save_freq=force_saving_freq, path=save_path),
EarlyStopping(freq=monitoring_freq, force_save_freq=force_saving_freq, path=save_path, channel=channel_name),
WeightNorm()
]
mainloop = Training(
name=pkl_name,
data=Iterator(train_data, batch_size, start=0, end=2040064),
model=model,
optimizer=optimizer,
cost=nll_upper_bound,
outputs=[nll_upper_bound],
extension=extension
)
mainloop.run()
def __init__(self, rho=0.5, eps=1e-6, **kwargs):
super(BatchNormLayer, self).__init__(**kwargs)
self.rho = rho
self.eps = eps
self.mu = sharedX(InitCell('zeros').get(self.nout), name='mu_'+self.name)
self.sigma = sharedX(InitCell('ones').get(self.nout), name='sigma_'+self.name)
def getX(self, shape, name=None):
return sharedX(self.init_param(shape), name)
output = FullyConnectedLayer(name='output',
parent=['h1', 'h2', 'h3'],
parent_dim=[200, 200, 200],
nout=205,
unit='softmax',
init_W=init_W,
init_b=init_b)
nodes = [h1, h2, h3, output]
for node in nodes:
node.initialize()
params = flatten([node.get_params().values() for node in nodes])
step_count = sharedX(0, name='step_count')
last_h = np.zeros((batch_size, 400), dtype=np.float32)
h1_tm1 = sharedX(last_h, name='h1_tm1')
h2_tm1 = sharedX(last_h, name='h2_tm1')
h3_tm1 = sharedX(last_h, name='h3_tm1')
update_list = [step_count, h1_tm1, h2_tm1, h3_tm1]
step_count = T.switch(T.le(step_count, reset_freq), step_count + 1, 0)
s1_0 = T.switch(
T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(h1_tm1), 0.), 'int32')), h1.get_init_state(),
h1_tm1)
s2_0 = T.switch(
T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(h2_tm1), 0.), 'int32')), h2.get_init_state(),
output = FullyConnectedLayer(name='output',
parent=['h1', 'h2', 'h3'],
parent_dim=[200, 200, 200],
nout=205,
unit='softmax',
init_W=init_W,
init_b=init_b)
nodes = [h1, h2, h3, output]
for node in nodes:
node.initialize()
params = flatten([node.get_params().values() for node in nodes])
step_count = sharedX(0, name='step_count')
last_h = np.zeros((batch_size, 400), dtype=np.float32)
h1_tm1 = sharedX(last_h, name='h1_tm1')
h2_tm1 = sharedX(last_h, name='h2_tm1')
h3_tm1 = sharedX(last_h, name='h3_tm1')
update_list = [step_count, h1_tm1, h2_tm1, h3_tm1]
step_count = T.switch(T.le(step_count, reset_freq),
step_count + 1, 0)
s1_0 = T.switch(T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(h1_tm1), 0.), 'int32')),
h1.get_init_state(), h1_tm1)
s2_0 = T.switch(T.or_(T.cast(T.eq(step_count, 0), 'int32'),
T.cast(T.eq(T.sum(h2_tm1), 0.), 'int32')),
h2.get_init_state(), h2_tm1)
def main(args):
trial = int(args["trial"])
pkl_name = "vrnn_gauss_%d" % trial
channel_name = "valid_nll_upper_bound"
data_path = args["data_path"]
save_path = args["save_path"]
data_path = os.path.expanduser(args["data_path"])
save_path = os.path.expanduser(args["save_path"])
monitoring_freq = int(args["monitoring_freq"])
force_saving_freq = int(args["force_saving_freq"])
reset_freq = int(args["reset_freq"])
epoch = int(args["epoch"])
batch_size = int(args["batch_size"])
m_batch_size = int(args["m_batch_size"])
x_dim = int(args["x_dim"])
z_dim = int(args["z_dim"])
rnn_dim = int(args["rnn_dim"])
lr = float(args["lr"])
debug = int(args["debug"])
print "trial no. %d" % trial
print "batch size %d" % batch_size
print "learning rate %f" % lr
print "saving pkl file '%s'" % pkl_name
print "to the save path '%s'" % save_path
q_z_dim = 500
p_z_dim = 500
p_x_dim = 600
x2s_dim = 600
z2s_dim = 500
target_dim = x_dim
file_name = "blizzard_tbptt"
normal_params = np.load(data_path + file_name + "_normal.npz")
X_mean = normal_params["X_mean"]
X_std = normal_params["X_std"]
model = Model()
train_data = Blizzard_tbptt(
name="train", path=data_path, frame_size=x_dim, file_name=file_name, X_mean=X_mean, X_std=X_std
)
valid_data = Blizzard_tbptt(
name="valid", path=data_path, frame_size=x_dim, file_name=file_name, X_mean=X_mean, X_std=X_std
)
x = train_data.theano_vars()
m_x = valid_data.theano_vars()
if debug:
x.tag.test_value = np.zeros((15, batch_size, x_dim), dtype=theano.config.floatX)
m_x.tag.test_value = np.zeros((15, m_batch_size, x_dim), dtype=theano.config.floatX)
init_W = InitCell("rand")
init_U = InitCell("ortho")
init_b = InitCell("zeros")
init_b_sig = InitCell("const", mean=0.6)
x_1 = FullyConnectedLayer(
name="x_1", parent=["x_t"], parent_dim=[x_dim], nout=x2s_dim, unit="relu", init_W=init_W, init_b=init_b
)
x_2 = FullyConnectedLayer(
name="x_2", parent=["x_1"], parent_dim=[x2s_dim], nout=x2s_dim, unit="relu", init_W=init_W, init_b=init_b
)
x_3 = FullyConnectedLayer(
name="x_3", parent=["x_2"], parent_dim=[x2s_dim], nout=x2s_dim, unit="relu", init_W=init_W, init_b=init_b
)
x_4 = FullyConnectedLayer(
name="x_4", parent=["x_3"], parent_dim=[x2s_dim], nout=x2s_dim, unit="relu", init_W=init_W, init_b=init_b
)
z_1 = FullyConnectedLayer(
name="z_1", parent=["z_t"], parent_dim=[z_dim], nout=z2s_dim, unit="relu", init_W=init_W, init_b=init_b
)
z_2 = FullyConnectedLayer(
name="z_2", parent=["z_1"], parent_dim=[z2s_dim], nout=z2s_dim, unit="relu", init_W=init_W, init_b=init_b
)
z_3 = FullyConnectedLayer(
name="z_3", parent=["z_2"], parent_dim=[z2s_dim], nout=z2s_dim, unit="relu", init_W=init_W, init_b=init_b
)
z_4 = FullyConnectedLayer(
name="z_4", parent=["z_3"], parent_dim=[z2s_dim], nout=z2s_dim, unit="relu", init_W=init_W, init_b=init_b
)
rnn = LSTM(
name="rnn",
parent=["x_4", "z_4"],
parent_dim=[x2s_dim, z2s_dim],
nout=rnn_dim,
unit="tanh",
init_W=init_W,
init_U=init_U,
init_b=init_b,
)
phi_1 = FullyConnectedLayer(
name="phi_1",
parent=["x_4", "s_tm1"],
parent_dim=[x2s_dim, rnn_dim],
nout=q_z_dim,
unit="relu",
init_W=init_W,
init_b=init_b,
)
phi_2 = FullyConnectedLayer(
name="phi_2", parent=["phi_1"], parent_dim=[q_z_dim], nout=q_z_dim, unit="relu", init_W=init_W, init_b=init_b
)
phi_3 = FullyConnectedLayer(
name="phi_3", parent=["phi_2"], parent_dim=[q_z_dim], nout=q_z_dim, unit="relu", init_W=init_W, init_b=init_b
)
phi_4 = FullyConnectedLayer(
name="phi_4", parent=["phi_3"], parent_dim=[q_z_dim], nout=q_z_dim, unit="relu", init_W=init_W, init_b=init_b
)
phi_mu = FullyConnectedLayer(
name="phi_mu", parent=["phi_4"], parent_dim=[q_z_dim], nout=z_dim, unit="linear", init_W=init_W, init_b=init_b
)
phi_sig = FullyConnectedLayer(
name="phi_sig",
parent=["phi_4"],
parent_dim=[q_z_dim],
nout=z_dim,
unit="softplus",
cons=1e-4,
init_W=init_W,
init_b=init_b_sig,
)
prior_1 = FullyConnectedLayer(
name="prior_1", parent=["s_tm1"], parent_dim=[rnn_dim], nout=p_z_dim, unit="relu", init_W=init_W, init_b=init_b
)
prior_2 = FullyConnectedLayer(
name="prior_2",
parent=["prior_1"],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit="relu",
init_W=init_W,
init_b=init_b,
)
prior_3 = FullyConnectedLayer(
name="prior_3",
parent=["prior_2"],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit="relu",
init_W=init_W,
init_b=init_b,
)
prior_4 = FullyConnectedLayer(
name="prior_4",
parent=["prior_3"],
parent_dim=[p_z_dim],
nout=p_z_dim,
unit="relu",
init_W=init_W,
init_b=init_b,
)
prior_mu = FullyConnectedLayer(
name="prior_mu",
parent=["prior_4"],
parent_dim=[p_z_dim],
nout=z_dim,
unit="linear",
init_W=init_W,
init_b=init_b,
)
prior_sig = FullyConnectedLayer(
name="prior_sig",
parent=["prior_4"],
parent_dim=[p_z_dim],
nout=z_dim,
unit="softplus",
cons=1e-4,
init_W=init_W,
init_b=init_b_sig,
)
theta_1 = FullyConnectedLayer(
name="theta_1",
parent=["z_4", "s_tm1"],
parent_dim=[z2s_dim, rnn_dim],
nout=p_x_dim,
unit="relu",
init_W=init_W,
init_b=init_b,
)
theta_2 = FullyConnectedLayer(
name="theta_2",
parent=["theta_1"],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit="relu",
init_W=init_W,
init_b=init_b,
)
theta_3 = FullyConnectedLayer(
name="theta_3",
parent=["theta_2"],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit="relu",
init_W=init_W,
init_b=init_b,
)
theta_4 = FullyConnectedLayer(
name="theta_4",
parent=["theta_3"],
parent_dim=[p_x_dim],
nout=p_x_dim,
unit="relu",
init_W=init_W,
init_b=init_b,
)
theta_mu = FullyConnectedLayer(
name="theta_mu",
parent=["theta_4"],
parent_dim=[p_x_dim],
nout=target_dim,
unit="linear",
init_W=init_W,
init_b=init_b,
)
theta_sig = FullyConnectedLayer(
name="theta_sig",
parent=["theta_4"],
parent_dim=[p_x_dim],
nout=target_dim,
unit="softplus",
cons=1e-4,
init_W=init_W,
init_b=init_b_sig,
)
nodes = [
rnn,
x_1,
x_2,
x_3,
x_4,
z_1,
z_2,
z_3,
z_4,
phi_1,
phi_2,
phi_3,
phi_4,
phi_mu,
phi_sig,
prior_1,
prior_2,
prior_3,
prior_4,
prior_mu,
prior_sig,
theta_1,
theta_2,
theta_3,
theta_4,
theta_mu,
theta_sig,
]
params = OrderedDict()
for node in nodes:
if node.initialize() is not None:
params.update(node.initialize())
params = init_tparams(params)
step_count = sharedX(0, name="step_count")
last_rnn = np.zeros((batch_size, rnn_dim * 2), dtype=theano.config.floatX)
rnn_tm1 = sharedX(last_rnn, name="rnn_tm1")
shared_updates = OrderedDict()
shared_updates[step_count] = step_count + 1
s_0 = T.switch(T.eq(T.mod(step_count, reset_freq), 0), rnn.get_init_state(batch_size), rnn_tm1)
x_1_temp = x_1.fprop([x], params)
x_2_temp = x_2.fprop([x_1_temp], params)
x_3_temp = x_3.fprop([x_2_temp], params)
x_4_temp = x_4.fprop([x_3_temp], params)
def inner_fn(x_t, s_tm1):
phi_1_t = phi_1.fprop([x_t, s_tm1], params)
phi_2_t = phi_2.fprop([phi_1_t], params)
phi_3_t = phi_3.fprop([phi_2_t], params)
phi_4_t = phi_4.fprop([phi_3_t], params)
phi_mu_t = phi_mu.fprop([phi_4_t], params)
phi_sig_t = phi_sig.fprop([phi_4_t], params)
prior_1_t = prior_1.fprop([s_tm1], params)
prior_2_t = prior_2.fprop([prior_1_t], params)
prior_3_t = prior_3.fprop([prior_2_t], params)
prior_4_t = prior_4.fprop([prior_3_t], params)
prior_mu_t = prior_mu.fprop([prior_4_t], params)
prior_sig_t = prior_sig.fprop([prior_4_t], params)
z_t = Gaussian_sample(phi_mu_t, phi_sig_t)
z_1_t = z_1.fprop([z_t], params)
z_2_t = z_2.fprop([z_1_t], params)
z_3_t = z_3.fprop([z_2_t], params)
z_4_t = z_4.fprop([z_3_t], params)
s_t = rnn.fprop([[x_t, z_4_t], [s_tm1]], params)
return s_t, phi_mu_t, phi_sig_t, prior_mu_t, prior_sig_t, z_4_t, z_t
((s_temp, phi_mu_temp, phi_sig_temp, prior_mu_temp, prior_sig_temp, z_4_temp, z_t), updates) = theano.scan(
fn=inner_fn, sequences=[x_4_temp], outputs_info=[s_0, None, None, None, None, None, None]
)
for k, v in updates.iteritems():
k.default_update = v
shared_updates[rnn_tm1] = s_temp[-1]
s_temp = concatenate([s_0[None, :, :], s_temp[:-1]], axis=0)
theta_1_temp = theta_1.fprop([z_4_temp, s_temp], params)
theta_2_temp = theta_2.fprop([theta_1_temp], params)
theta_3_temp = theta_3.fprop([theta_2_temp], params)
theta_4_temp = theta_4.fprop([theta_3_temp], params)
theta_mu_temp = theta_mu.fprop([theta_4_temp], params)
theta_sig_temp = theta_sig.fprop([theta_4_temp], params)
kl_temp = KLGaussianGaussian(phi_mu_temp, phi_sig_temp, prior_mu_temp, prior_sig_temp)
#### recon = Gaussian(x, theta_mu_temp, theta_sig_temp)
recon = Gaussian(x, theta_mu_temp, theta_sig_temp) - Gaussian(z_t, phi_mu_temp, phi_sig_temp)
recon += Gaussian(z_t, prior_mu_temp, prior_sig_temp)
recon_term = recon.mean() / 5.0
kl_term = kl_temp.mean()
##### nll_upper_bound = recon_term + kl_term
nll_upper_bound = recon_term
nll_upper_bound.name = "nll_upper_bound"
m_x_1_temp = x_1.fprop([m_x], params)
m_x_2_temp = x_2.fprop([m_x_1_temp], params)
m_x_3_temp = x_3.fprop([m_x_2_temp], params)
m_x_4_temp = x_4.fprop([m_x_3_temp], params)
m_s_0 = rnn.get_init_state(m_batch_size)
(
(m_s_temp, m_phi_mu_temp, m_phi_sig_temp, m_prior_mu_temp, m_prior_sig_temp, m_z_4_temp, m_z_t),
m_updates,
) = theano.scan(fn=inner_fn, sequences=[m_x_4_temp], outputs_info=[m_s_0, None, None, None, None, None, None])
for k, v in m_updates.iteritems():
k.default_update = v
m_s_temp = concatenate([m_s_0[None, :, :], m_s_temp[:-1]], axis=0)
m_theta_1_temp = theta_1.fprop([m_z_4_temp, m_s_temp], params)
m_theta_2_temp = theta_2.fprop([m_theta_1_temp], params)
m_theta_3_temp = theta_3.fprop([m_theta_2_temp], params)
m_theta_4_temp = theta_4.fprop([m_theta_3_temp], params)
m_theta_mu_temp = theta_mu.fprop([m_theta_4_temp], params)
m_theta_sig_temp = theta_sig.fprop([m_theta_4_temp], params)
m_kl_temp = KLGaussianGaussian(m_phi_mu_temp, m_phi_sig_temp, m_prior_mu_temp, m_prior_sig_temp)
m_recon = Gaussian(m_x, m_theta_mu_temp, m_theta_sig_temp)
m_recon_term = m_recon.mean()
m_kl_term = m_kl_temp.mean()
m_nll_upper_bound = m_recon_term + m_kl_term
m_nll_upper_bound.name = "nll_upper_bound"
m_recon_term.name = "recon_term"
m_kl_term.name = "kl_term"
max_x = m_x.max()
mean_x = m_x.mean()
min_x = m_x.min()
max_x.name = "max_x"
mean_x.name = "mean_x"
min_x.name = "min_x"
max_theta_mu = m_theta_mu_temp.max()
mean_theta_mu = m_theta_mu_temp.mean()
min_theta_mu = m_theta_mu_temp.min()
max_theta_mu.name = "max_theta_mu"
mean_theta_mu.name = "mean_theta_mu"
min_theta_mu.name = "min_theta_mu"
max_theta_sig = m_theta_sig_temp.max()
mean_theta_sig = m_theta_sig_temp.mean()
min_theta_sig = m_theta_sig_temp.min()
max_theta_sig.name = "max_theta_sig"
mean_theta_sig.name = "mean_theta_sig"
min_theta_sig.name = "min_theta_sig"
max_phi_sig = m_phi_sig_temp.max()
mean_phi_sig = m_phi_sig_temp.mean()
min_phi_sig = m_phi_sig_temp.min()
max_phi_sig.name = "max_phi_sig"
mean_phi_sig.name = "mean_phi_sig"
min_phi_sig.name = "min_phi_sig"
max_prior_sig = m_prior_sig_temp.max()
mean_prior_sig = m_prior_sig_temp.mean()
min_prior_sig = m_prior_sig_temp.min()
max_prior_sig.name = "max_prior_sig"
mean_prior_sig.name = "mean_prior_sig"
min_prior_sig.name = "min_prior_sig"
model.inputs = [x]
model.params = params
model.nodes = nodes
model.set_updates(shared_updates)
optimizer = Adam(lr=lr)
monitor_fn = theano.function(
inputs=[m_x],
outputs=[
m_nll_upper_bound,
m_recon_term,
m_kl_term,
max_phi_sig,
mean_phi_sig,
min_phi_sig,
max_prior_sig,
mean_prior_sig,
min_prior_sig,
max_theta_sig,
mean_theta_sig,
min_theta_sig,
max_x,
mean_x,
min_x,
max_theta_mu,
mean_theta_mu,
min_theta_mu,
],
on_unused_input="ignore",
)
extension = [
GradientClipping(batch_size=batch_size, check_nan=1),
EpochCount(epoch),
Monitoring(
freq=monitoring_freq,
monitor_fn=monitor_fn,
ddout=[
m_nll_upper_bound,
m_recon_term,
m_kl_term,
max_phi_sig,
mean_phi_sig,
min_phi_sig,
max_prior_sig,
mean_prior_sig,
min_prior_sig,
max_theta_sig,
mean_theta_sig,
min_theta_sig,
max_x,
mean_x,
min_x,
max_theta_mu,
mean_theta_mu,
min_theta_mu,
],
data=[
Iterator(train_data, m_batch_size, start=0, end=112640),
Iterator(valid_data, m_batch_size, start=2040064, end=2152704),
],
),
Picklize(freq=monitoring_freq, force_save_freq=force_saving_freq, path=save_path),
EarlyStopping(freq=monitoring_freq, force_save_freq=force_saving_freq, path=save_path, channel=channel_name),
WeightNorm(),
]
mainloop = Training(
name=pkl_name,
data=KIter(train_data, batch_size, start=0, end=2040064),
model=model,
optimizer=optimizer,
cost=nll_upper_bound,
outputs=[nll_upper_bound],
extension=extension,
)
mainloop.run()