def create_model(num_blocks=3,
block_size=48,
sym_recon_grad=False,
actnorm=False,
split_prior=False,
recon_loss_weight=1.0):
current_size = (3, 32, 32)
alpha = 1e-6
layers = [
Dequantization(UniformDistribution(size=current_size)),
Normalization(translation=0, scale=256),
Normalization(translation=-alpha, scale=1 / (1 - 2 * alpha)),
LogitTransform(),
]
for l in range(num_blocks):
layers.append(Squeeze())
current_size = (current_size[0] * 4, current_size[1] // 2,
current_size[2] // 2)
for k in range(block_size):
if actnorm:
layers.append(ActNorm(current_size[0]))
layers.append(Conv1x1(current_size[0]))
layers.append(Coupling(current_size))
if split_prior and l < num_blocks - 1:
layers.append(SplitPrior(current_size, NegativeGaussianLoss))
current_size = (current_size[0] // 2, current_size[1],
current_size[2])
return FlowSequential(NegativeGaussianLoss(size=current_size), *layers)
def create_model(num_blocks=2, block_size=16, sym_recon_grad=False,
actnorm=False, split_prior=False, recon_loss_weight=100.0):
alpha = 1e-6
layers = [
Dequantization(UniformDistribution(size=(1, 28, 28))),
Normalization(translation=0, scale=256),
Normalization(translation=-alpha, scale=1 / (1 - 2 * alpha)),
LogitTransform(),
]
current_size = (1, 28, 28)
for l in range(num_blocks):
layers.append(Squeeze())
current_size = (current_size[0]*4, current_size[1]//2, current_size[2]//2)
for k in range(block_size):
if actnorm:
layers.append(ActNorm(current_size[0]))
layers.append(SelfNormConv(current_size[0], current_size[0], (1, 1),
bias=True, stride=1, padding=0,
sym_recon_grad=sym_recon_grad,
recon_loss_weight=recon_loss_weight))
layers.append(Coupling(current_size))
if split_prior and l < num_blocks - 1:
layers.append(SplitPrior(current_size, NegativeGaussianLoss))
current_size = (current_size[0] // 2, current_size[1], current_size[2])
return FlowSequential(NegativeGaussianLoss(size=current_size),
*layers)
def test_logdet(input_size=(12, 4, 8, 8)):
check_logdet(LearnableLeakyRelu().to('cuda'), input_size)
check_logdet(SplineActivation(input_size).to('cuda'), input_size)
check_logdet(SmoothLeakyRelu().to('cuda'), input_size)
check_logdet(SmoothTanh().to('cuda'), input_size)
check_logdet(Identity().to('cuda'), input_size)
check_logdet(ActNorm(input_size[1]).to('cuda'), input_size)
check_logdet(Coupling(input_size[1:]).to('cuda'), input_size)
check_logdet(
Normalization(translation=-1e-6, scale=1 / (1 - 2 * 1e-6)).to('cuda'),
input_size)
check_logdet(Squeeze().to('cuda'), input_size)
check_logdet(UnSqueeze().to('cuda'), input_size)
test_snf_logdet(input_size)
print("All log-det tests passed")
def test_inverses(input_size=(12, 4, 16, 16)):
check_inverse(LearnableLeakyRelu().to('cuda'), input_size)
check_inverse(SplineActivation(input_size).to('cuda'), input_size)
check_inverse(SmoothLeakyRelu().to('cuda'), input_size)
check_inverse(SmoothTanh().to('cuda'), input_size)
check_inverse(Identity().to('cuda'), input_size)
check_inverse(ActNorm(input_size[1]).to('cuda'), input_size)
check_inverse(Conv1x1(input_size[1]).to('cuda'), input_size)
check_inverse(Conv1x1Householder(input_size[1], 10).to('cuda'), input_size)
check_inverse(Coupling(input_size[1:]).to('cuda'), input_size)
check_inverse(
Normalization(translation=-1e-6, scale=1 / (1 - 2 * 1e-6)).to('cuda'),
input_size)
check_inverse(Squeeze().to('cuda'), input_size)
check_inverse(UnSqueeze().to('cuda'), input_size)
test_snf_layer_inverses(input_size)
print("All inverse tests passed")