def test_tensor():
h2o = Dense(4, 2)
model = Tensor(input_dim=3, output_dim=4, causes_dim=2, hid2output=h2o,
return_mode='both')
X = tensor.tensor3()
Y = apply_layer(model, X)
F = function([X], Y, allow_input_downcast=True)
x = np.ones((3, 10, 3))
y = F(x)
assert y.shape == (10, 3, 4+2)
def step(self, x, states):
batch_size = self._get_batch_size(x)
input_shape = (batch_size, ) + self.reshape_dim
hidden_dim = (batch_size, ) + self.output_dim
nb_filter, nb_rows, nb_cols = self.output_dim
h_tm1 = K.reshape(states[0], hidden_dim)
x_t = K.reshape(x, input_shape)
xz_t = self.conv_x_z(x_t, train=True)
xr_t = self.conv_x_r(x_t, train=True)
xh_t = self.conv_x_h(x_t, train=True)
xz_t = apply_layer(self.max_pool, xz_t)
xr_t = apply_layer(self.max_pool, xr_t)
xh_t = apply_layer(self.max_pool, xh_t)
z = self.inner_activation(xz_t + self.conv_z(h_tm1))
r = self.inner_activation(xr_t + self.conv_r(h_tm1))
hh_t = self.activation(xh_t + self.conv_h(r * h_tm1))
h_t = z * h_tm1 + (1 - z) * hh_t
h_t = K.batch_flatten(h_t)
return h_t, [h_t, ]
def step(self, x, states):
input_shape = (self.batch_size, ) + self.reshape_dim
hidden_dim = (self.batch_size, ) + self.output_dim
nb_filter, nb_rows, nb_cols = self.output_dim
h_tm1 = K.reshape(states[0], hidden_dim)
x_t = K.reshape(x, input_shape)
xz_t = apply_layer(self.conv_x_z, x_t)
xr_t = apply_layer(self.conv_x_r, x_t)
xh_t = apply_layer(self.conv_x_h, x_t)
xz_t = apply_layer(self.max_pool, xz_t)
xr_t = apply_layer(self.max_pool, xr_t)
xh_t = apply_layer(self.max_pool, xh_t)
z = self.inner_activation(xz_t + apply_layer(self.conv_z, h_tm1))
r = self.inner_activation(xr_t + apply_layer(self.conv_r, h_tm1))
hh_t = self.activation(xh_t + apply_layer(self.conv_h, r * h_tm1))
h_t = z * h_tm1 + (1 - z) * hh_t
h_t = K.flatten(h_t)
return h_t, [h_t, ]
def step(self, x, states):
input_shape = (self.batch_size, ) + self.reshape_dim
hidden_dim = (self.batch_size, ) + self.output_dim
nb_filter, nb_rows, nb_cols = self.output_dim
h_tm1 = K.reshape(states[0], hidden_dim)
x_t = K.reshape(x, input_shape)
xz_t = apply_layer(self.conv_x_z, x_t)
xr_t = apply_layer(self.conv_x_r, x_t)
xh_t = apply_layer(self.conv_x_h, x_t)
xz_t = apply_layer(self.max_pool, xz_t)
xr_t = apply_layer(self.max_pool, xr_t)
xh_t = apply_layer(self.max_pool, xh_t)
z = self.inner_activation(xz_t + apply_layer(self.conv_z, h_tm1))
r = self.inner_activation(xr_t + apply_layer(self.conv_r, h_tm1))
hh_t = self.activation(xh_t + apply_layer(self.conv_h, r * h_tm1))
h_t = z * h_tm1 + (1 - z) * hh_t
h_t = K.flatten(h_t)
return h_t, [
h_t,
]
