def test_lasagne_gradient(num_classes):
bounds = (0, 255)
channels = num_classes
def mean_brightness_net(images):
logits = GlobalPoolLayer(images)
return logits
images_var = T.tensor4('images', dtype='float32')
images = InputLayer((None, channels, 5, 5), images_var)
logits = mean_brightness_net(images)
preprocessing = (np.arange(num_classes)[None, None],
np.random.uniform(size=(5, 5, channels)) + 1)
model = LasagneModel(images,
logits,
preprocessing=preprocessing,
bounds=bounds)
epsilon = 1e-2
np.random.seed(23)
test_image = np.random.rand(channels, 5, 5).astype(np.float32)
test_label = 7
_, g1 = model.predictions_and_gradient(test_image, test_label)
l1 = model._loss_fn(test_image[None] - epsilon / 2 * g1, [test_label])[0]
l2 = model._loss_fn(test_image[None] + epsilon / 2 * g1, [test_label])[0]
# make sure that gradient is numerically correct
np.testing.assert_array_almost_equal(1e4 * (l2 - l1),
1e4 * epsilon * np.linalg.norm(g1)**2,
decimal=1)
def test_lasagne_gradient(num_classes):
bounds = (0, 255)
channels = num_classes
def mean_brightness_net(images):
logits = GlobalPoolLayer(images)
return logits
images_var = T.tensor4('images', dtype='float32')
images = InputLayer((None, channels, 5, 5), images_var)
logits = mean_brightness_net(images)
preprocessing = (np.arange(num_classes)[:, None, None],
np.random.uniform(size=(channels, 5, 5)) + 1)
model = LasagneModel(
images,
logits,
preprocessing=preprocessing,
bounds=bounds)
# theano and lasagne calculate the cross-entropy from the probbilities
# rather than combining softmax and cross-entropy calculation; they
# therefore have lower numerical accuracy
epsilon = 1e-3
np.random.seed(23)
test_image = np.random.rand(channels, 5, 5).astype(np.float32)
test_label = 7
_, g1 = model.predictions_and_gradient(test_image, test_label)
l1 = model._loss_fn(test_image[None] - epsilon / 2 * g1, [test_label])[0]
l2 = model._loss_fn(test_image[None] + epsilon / 2 * g1, [test_label])[0]
assert 1e5 * (l2 - l1) > 1
# make sure that gradient is numerically correct
np.testing.assert_array_almost_equal(
1e5 * (l2 - l1),
1e5 * epsilon * np.linalg.norm(g1)**2,
decimal=1)
def test_lasagne_gradient(num_classes):
bounds = (0, 255)
channels = num_classes
def mean_brightness_net(images):
logits = GlobalPoolLayer(images)
return logits
images_var = T.tensor4('images', dtype='float32')
images = InputLayer((None, channels, 5, 5), images_var)
logits = mean_brightness_net(images)
preprocessing = (np.arange(num_classes)[None, None],
np.random.uniform(size=(5, 5, channels)) + 1)
model = LasagneModel(
images,
logits,
preprocessing=preprocessing,
bounds=bounds)
epsilon = 1e-2
np.random.seed(23)
test_image = np.random.rand(channels, 5, 5).astype(np.float32)
test_label = 7
_, g1 = model.predictions_and_gradient(test_image, test_label)
l1 = model._loss_fn(test_image[None] - epsilon / 2 * g1, [test_label])[0]
l2 = model._loss_fn(test_image[None] + epsilon / 2 * g1, [test_label])[0]
# make sure that gradient is numerically correct
np.testing.assert_array_almost_equal(
1e4 * (l2 - l1),
1e4 * epsilon * np.linalg.norm(g1)**2,
decimal=1)