def __init__(
self,
features,
hidden_features,
context_features=None,
num_blocks=2,
use_residual_blocks=True,
random_mask=False,
activation=F.relu,
dropout_probability=0.0,
use_batch_norm=False,
):
self.features = features
made = made_module.MADE(
features=features,
hidden_features=hidden_features,
context_features=context_features,
num_blocks=num_blocks,
output_multiplier=self._output_dim_multiplier(),
use_residual_blocks=use_residual_blocks,
random_mask=random_mask,
activation=activation,
dropout_probability=dropout_probability,
use_batch_norm=use_batch_norm,
)
self._epsilon = 1e-3
super(MaskedAffineAutoregressiveTransform, self).__init__(made)
def __init__(
self,
num_bins,
features,
hidden_features,
context_features=None,
num_blocks=2,
use_residual_blocks=True,
random_mask=False,
activation=F.relu,
dropout_probability=0.0,
use_batch_norm=False,
):
self.num_bins = num_bins
self.features = features
made = made_module.MADE(
features=features,
hidden_features=hidden_features,
context_features=context_features,
num_blocks=num_blocks,
output_multiplier=self._output_dim_multiplier(),
use_residual_blocks=use_residual_blocks,
random_mask=random_mask,
activation=activation,
dropout_probability=dropout_probability,
use_batch_norm=use_batch_norm,
)
super().__init__(made)
def test_gradients(self):
features = 10
hidden_features = 256
num_blocks = 20
output_multiplier = 3
for use_residual_blocks, random_mask in [
(False, False),
(False, True),
(True, False),
]:
with self.subTest(use_residual_blocks=use_residual_blocks,
random_mask=random_mask):
model = made.MADE(
features=features,
hidden_features=hidden_features,
num_blocks=num_blocks,
output_multiplier=output_multiplier,
use_residual_blocks=use_residual_blocks,
random_mask=random_mask,
)
inputs = torch.randn(1, features)
inputs.requires_grad = True
for k in range(features * output_multiplier):
outputs = model(inputs)
outputs[0, k].backward()
depends = inputs.grad.data[0] != 0.0
dim = k // output_multiplier
self.assertEqual(torch.all(depends[dim:] == 0), 1)
def test_unconditional(self):
features = 100
hidden_features = 200
num_blocks = 5
output_multiplier = 3
batch_size = 16
inputs = torch.randn(batch_size, features)
for use_residual_blocks, random_mask in [
(False, False),
(False, True),
(True, False),
]:
with self.subTest(use_residual_blocks=use_residual_blocks,
random_mask=random_mask):
model = made.MADE(
features=features,
hidden_features=hidden_features,
num_blocks=num_blocks,
output_multiplier=output_multiplier,
context_features=None,
use_residual_blocks=use_residual_blocks,
random_mask=random_mask,
)
outputs = model(inputs)
self.assertEqual(outputs.dim(), 2)
self.assertEqual(outputs.shape[0], batch_size)
self.assertEqual(outputs.shape[1],
output_multiplier * features)
def test_total_mask_sequential(self):
features = 10
hidden_features = 50
num_blocks = 5
output_multiplier = 1
for use_residual_blocks in [True, False]:
with self.subTest(use_residual_blocks=use_residual_blocks):
model = made.MADE(
features=features,
hidden_features=hidden_features,
num_blocks=num_blocks,
output_multiplier=output_multiplier,
use_residual_blocks=use_residual_blocks,
random_mask=False,
)
total_mask = model.initial_layer.mask
for block in model.blocks:
if use_residual_blocks:
self.assertIsInstance(block, made.MaskedResidualBlock)
total_mask = block.linear_layers[0].mask @ total_mask
total_mask = block.linear_layers[1].mask @ total_mask
else:
self.assertIsInstance(block,
made.MaskedFeedforwardBlock)
total_mask = block.linear.mask @ total_mask
total_mask = model.final_layer.mask @ total_mask
total_mask = (total_mask > 0).float()
reference = torch.tril(torch.ones([features, features]), -1)
self.assertEqual(total_mask, reference)
def __init__(self,
features,
hidden_features,
lower_bounds,
upper_bounds,
context_features=None,
num_bins=10,
tails=None,
tail_bound=1.0,
num_blocks=2,
use_residual_blocks=True,
random_mask=False,
activation=F.relu,
dropout_probability=0.0,
use_batch_norm=False,
min_bin_width=rational_quadratic.DEFAULT_MIN_BIN_WIDTH,
min_bin_height=rational_quadratic.DEFAULT_MIN_BIN_HEIGHT,
min_derivative=rational_quadratic.DEFAULT_MIN_DERIVATIVE,
permutation=0):
self.num_bins = num_bins
self.min_bin_width = min_bin_width
self.min_bin_height = min_bin_height
self.min_derivative = min_derivative
self.tails = tails
self.tail_bound = tail_bound
self.lower_bounds = lower_bounds
self.upper_bounds = upper_bounds
self.permutation = permutation
autoregressive_net = made_module.MADE(
features=2 * features,
hidden_features=hidden_features,
autoregressive_features=features,
context_features=context_features,
num_blocks=num_blocks,
output_multiplier=self._output_dim_multiplier(),
use_residual_blocks=use_residual_blocks,
random_mask=random_mask,
activation=activation,
dropout_probability=dropout_probability,
use_batch_norm=use_batch_norm,
)
super().__init__(autoregressive_net)
self.phase = torch.nn.Parameter(torch.zeros_like(self.lower_bounds))
def test_total_mask_random(self):
features = 10
hidden_features = 50
num_blocks = 5
output_multiplier = 1
model = made.MADE(
features=features,
hidden_features=hidden_features,
num_blocks=num_blocks,
output_multiplier=output_multiplier,
use_residual_blocks=False,
random_mask=True,
)
total_mask = model.initial_layer.mask
for block in model.blocks:
self.assertIsInstance(block, made.MaskedFeedforwardBlock)
total_mask = block.linear.mask @ total_mask
total_mask = model.final_layer.mask @ total_mask
total_mask = (total_mask > 0).float()
self.assertEqual(torch.triu(total_mask),
torch.zeros([features, features]))
def __init__(
self,
features,
hidden_features,
context_features=None,
num_bins=10,
num_blocks=2,
tails=None,
tail_bound=1.0,
use_residual_blocks=True,
random_mask=False,
activation=F.relu,
dropout_probability=0.0,
use_batch_norm=False,
min_bin_width=rational_quadratic.DEFAULT_MIN_BIN_WIDTH,
min_bin_height=rational_quadratic.DEFAULT_MIN_BIN_HEIGHT,
min_derivative=rational_quadratic.DEFAULT_MIN_DERIVATIVE,
):
self.num_bins = num_bins
self.min_bin_width = min_bin_width
self.min_bin_height = min_bin_height
self.min_derivative = min_derivative
self.tails = tails
self.tail_bound = tail_bound
self.features = features
made = made_module.MADE(
features=features,
hidden_features=hidden_features,
context_features=context_features,
num_blocks=num_blocks,
output_multiplier=self._output_dim_multiplier(),
use_residual_blocks=use_residual_blocks,
random_mask=random_mask,
activation=activation,
dropout_probability=dropout_probability,
use_batch_norm=use_batch_norm,
)
super().__init__(made)
