def test_e_dropout_rbm_total_energy():
new_e_dropout_rbm = e_dropout_rbm.EDropoutRBM()
h = torch.ones(1, 128)
v = torch.ones(1, 128)
energy = new_e_dropout_rbm.total_energy(h, v)
assert energy != 0
def test_e_dropout_rbm_M_setter():
new_e_dropout_rbm = e_dropout_rbm.EDropoutRBM()
try:
new_e_dropout_rbm.M = 1
except:
new_e_dropout_rbm.M = torch.Tensor()
assert isinstance(new_e_dropout_rbm.M, torch.Tensor)
def test_e_dropout_rbm_hidden_sampling():
new_e_dropout_rbm = e_dropout_rbm.EDropoutRBM()
new_e_dropout_rbm.M = torch.ones((1, 128))
v = torch.ones(1, 128)
probs, states = new_e_dropout_rbm.hidden_sampling(v)
assert probs.size(1) == 128
assert states.size(1) == 128
def test_e_dropout_rbm_energy_dropout():
new_e_dropout_rbm = e_dropout_rbm.EDropoutRBM()
new_e_dropout_rbm.M = torch.ones((1, 128))
e = torch.ones(1, 128)
p_prob = torch.ones(1, 128)
n_prob = torch.ones(1, 128)
new_e_dropout_rbm.energy_dropout(e, p_prob, n_prob)
pass
def test_e_dropout_rbm_reconstruct():
test = torchvision.datasets.KMNIST(
root='./data',
train=False,
download=True,
transform=torchvision.transforms.ToTensor())
new_e_dropout_rbm = e_dropout_rbm.EDropoutRBM(n_visible=784,
n_hidden=128,
steps=1,
learning_rate=0.1,
momentum=0,
decay=0,
temperature=1,
use_gpu=False)
e, v = new_e_dropout_rbm.reconstruct(test)
assert e >= 0
assert v.size(1) == 784
def test_e_dropout_rbm_fit():
train = torchvision.datasets.KMNIST(
root='./data',
train=True,
download=True,
transform=torchvision.transforms.ToTensor())
new_e_dropout_rbm = e_dropout_rbm.EDropoutRBM(n_visible=784,
n_hidden=128,
steps=1,
learning_rate=0.1,
momentum=0,
decay=0,
temperature=1,
use_gpu=False)
e, pl = new_e_dropout_rbm.fit(train, batch_size=128, epochs=1)
assert e >= 0
assert pl <= 0
def test_e_dropout_rbm_M():
new_e_dropout_rbm = e_dropout_rbm.EDropoutRBM()
assert isinstance(new_e_dropout_rbm.M, torch.Tensor)
