def test_discriminative_rbm_labels_sampling():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
samples = torch.ones(1, 128)
probs, labels = new_discriminative_rbm.labels_sampling(samples)
assert probs.size(1) == 1
assert labels.size(0) == 1
def test_discriminative_rbm_c_setter():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
try:
new_discriminative_rbm.c = 1
except:
new_discriminative_rbm.c = torch.nn.Parameter(torch.zeros(10))
assert new_discriminative_rbm.c.size(0) == 10
def test_discriminative_rbm_loss_setter():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
try:
new_discriminative_rbm.loss = 'LOSS'
except:
new_discriminative_rbm.loss = torch.nn.CrossEntropyLoss()
assert type(new_discriminative_rbm.loss).__name__ == 'CrossEntropyLoss'
def test_discriminative_rbm_U_setter():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
try:
new_discriminative_rbm.U = 1
except:
new_discriminative_rbm.U = torch.nn.Parameter(
torch.randn(10, 128) * 0.01)
assert new_discriminative_rbm.U.size(0) == 10
assert new_discriminative_rbm.U.size(1) == 128
def test_discriminative_rbm_n_classes_setter():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
try:
new_discriminative_rbm.n_classes = 'a'
except:
new_discriminative_rbm.n_classes = 1
assert new_discriminative_rbm.n_classes == 1
try:
new_discriminative_rbm.n_classes = 0
except:
new_discriminative_rbm.n_classes = 1
assert new_discriminative_rbm.n_classes == 1
def test_discriminative_rbm_fit():
train = torchvision.datasets.MNIST(
root='./data',
train=True,
download=True,
transform=torchvision.transforms.ToTensor())
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM(
n_visible=784,
n_hidden=128,
n_classes=10,
learning_rate=0.1,
momentum=0,
decay=0,
use_gpu=False)
loss, acc = new_discriminative_rbm.fit(train, batch_size=128, epochs=1)
assert loss >= 0
assert acc >= 0
def test_discriminative_rbm_predict():
test = torchvision.datasets.MNIST(
root='./data',
train=False,
download=True,
transform=torchvision.transforms.ToTensor())
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM(
n_visible=784,
n_hidden=128,
n_classes=10,
learning_rate=0.1,
momentum=0,
decay=0,
use_gpu=False)
acc, probs, labels = new_discriminative_rbm.predict(test)
assert acc >= 0
assert probs.size(0) == 10000
assert labels.size(0) == 10000
def test_discriminative_rbm_loss():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
assert type(new_discriminative_rbm.loss).__name__ == 'CrossEntropyLoss'
def test_discriminative_rbm_n_classes():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
assert new_discriminative_rbm.n_classes == 1
def test_discriminative_rbm_c():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
assert new_discriminative_rbm.c.size(0) == 1
def test_discriminative_rbm_U():
new_discriminative_rbm = discriminative_rbm.DiscriminativeRBM()
assert new_discriminative_rbm.U.size(0) == 1
assert new_discriminative_rbm.U.size(1) == 128