X_std = train_data.X_std
valid_data = TIMIT(name='valid',
path=data_path,
frame_size=frame_size,
shuffle=0,
use_n_gram=1,
X_mean=X_mean,
X_std=X_std)
init_W = InitCell('rand')
init_U = InitCell('ortho')
init_b = InitCell('zeros')
init_b_sig = InitCell('const', mean=0.6)
x, x_mask = train_data.theano_vars()
if debug:
x.tag.test_value = np.zeros((15, batch_size, frame_size), dtype=np.float32)
temp = np.ones((15, batch_size), dtype=np.float32)
temp[:, -2:] = 0.
x_mask.tag.test_value = temp
x_1 = FullyConnectedLayer(name='x_1',
parent=['x_t'],
parent_dim=[frame_size],
nout=x2s_dim,
unit='relu',
init_W=init_W,
init_b=init_b)
x_2 = FullyConnectedLayer(name='x_2',
shuffle=0,
use_n_gram=1,
X_mean=X_mean,
X_std=X_std)
exp = unpickle(exp_path + exp_name + '_best.pkl')
nodes = exp.model.nodes
names = [node.name for node in nodes]
output = GaussianLayer(name='output',
parent=['theta_mu',
'theta_sig'],
use_sample=1,
nout=frame_size)
x, y, spk_info, mask = train_data.theano_vars()
if debug:
x.tag.test_value = np.zeros((15, batch_size, frame_size), dtype=np.float32)
y.tag.test_value = np.zeros((15, batch_size, label_size), dtype=np.float32)
temp = np.ones((15, batch_size), dtype=np.float32)
temp[:, -2:] = 0.
mask.tag.test_value = temp
spk_info.tag.test_value = np.zeros((batch_size, 630), dtype=np.float32)
[main_lstm, prior, kl,
x_1, x_2, x_3, x_4,
z_1, z_2, z_3, z_4,
y_1, y_2, y_3, y_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] = nodes