def _prepare_generator(self, generator, noise_space,
condition_distribution, new_W_irange, input_source):
noise_dim = noise_space.get_total_dimension()
condition_dim = self.condition_space.get_total_dimension()
first_layer = generator.mlp.layers[0]
pretrain_W, _ = first_layer.get_param_values()
rng = generator.mlp.rng
new_W = np.vstack((pretrain_W,
rng.uniform(-new_W_irange, new_W_irange,
(condition_dim, pretrain_W.shape[1]))))
new_W = sharedX(new_W)
new_W.name = first_layer.get_params()[0].name + '_retrain'
first_layer.transformer = MatrixMul(new_W)
first_layer.input_space = CompositeSpace(
components=[noise_space, self.condition_space])
generator.mlp.input_space = first_layer.input_space
# HACK!
generator.mlp._input_source = input_source
return ConditionalGenerator(
generator.mlp,
input_condition_space=self.condition_space,
condition_distribution=condition_distribution,
noise_dim=noise_dim)
def setUp(self):
self.noise_dim = 10
self.num_labels = 10
self.condition_dtype = 'uint8'
self.condition_space = VectorSpace(dim=self.num_labels,
dtype=self.condition_dtype)
self.condition_formatter = OneHotFormatter(self.num_labels,
dtype=self.condition_dtype)
self.condition_distribution = OneHotDistribution(self.condition_space)
# TODO this nvis stuff is dirty. The ConditionalGenerator should handle it
self.mlp_nvis = self.noise_dim + self.num_labels
self.mlp_nout = 1
# Set up model
self.mlp = MLP(nvis=self.mlp_nvis,
layers=[Linear(self.mlp_nout, 'out', irange=0.1)])
self.G = ConditionalGenerator(
input_condition_space=self.condition_space,
condition_distribution=self.condition_distribution,
noise_dim=self.noise_dim,
mlp=self.mlp)
