def setUp(self):
self.func = PReLU()
self.func.W = numpy.random.uniform(-1, 1, self.func.W.shape).astype(
numpy.float32)
self.func.gW.fill(0)
self.W = self.func.W.copy() # fixed on CPU
# Avoid unstability of numerical gradient
self.x = numpy.random.uniform(.5, 1, (4, 3, 2)).astype(numpy.float32)
self.x *= numpy.random.randint(2, size=(4, 3, 2)) * 2 - 1
self.gy = numpy.random.uniform(-1, 1, (4, 3, 2)).astype(numpy.float32)
class TestPReLUSingle(TestCase):
def setUp(self):
self.func = PReLU()
self.func.W = numpy.random.uniform(
-1, 1, self.func.W.shape).astype(numpy.float32)
self.func.gW.fill(0)
self.W = self.func.W.copy() # fixed on CPU
# Avoid unstability of numerical gradient
self.x = numpy.random.uniform(.5, 1, (4, 3, 2)).astype(numpy.float32)
self.x *= numpy.random.randint(2, size=(4, 3, 2)) * 2 - 1
self.gy = numpy.random.uniform(-1, 1, (4, 3, 2)).astype(numpy.float32)
def check_forward(self, x_data):
x = Variable(x_data)
y = self.func(x)
y_expect = self.x.copy()
for i in numpy.ndindex(self.x.shape):
if self.x[i] < 0:
y_expect[i] *= self.W
assert_allclose(y_expect, y.data)
def test_forward_cpu(self):
self.check_forward(self.x)
@attr.gpu
def test_forward_gpu(self):
self.func.to_gpu()
self.check_forward(cuda.to_gpu(self.x))
def check_backward(self, x_data, y_grad):
x = Variable(x_data)
y = self.func(x)
y.grad = y_grad
y.backward()
func = y.creator
f = lambda: func.forward((x.data,))
gx, gW = numerical_grad(f, (x.data, func.W), (y.grad,))
assert_allclose(gx, x.grad)
assert_allclose(gW, func.gW, atol=1e-4)
def test_backward_cpu(self):
self.check_backward(self.x, self.gy)
@attr.gpu
def test_backward_gpu(self):
self.func.to_gpu()
self.check_backward(cuda.to_gpu(self.x), cuda.to_gpu(self.gy))
class TestPReLUSingle(TestCase):
def setUp(self):
self.func = PReLU()
self.func.W = numpy.random.uniform(-1, 1, self.func.W.shape).astype(
numpy.float32)
self.func.gW.fill(0)
self.W = self.func.W.copy() # fixed on CPU
# Avoid unstability of numerical gradient
self.x = numpy.random.uniform(.5, 1, (4, 3, 2)).astype(numpy.float32)
self.x *= numpy.random.randint(2, size=(4, 3, 2)) * 2 - 1
self.gy = numpy.random.uniform(-1, 1, (4, 3, 2)).astype(numpy.float32)
def check_forward(self, x_data):
x = Variable(x_data)
y = self.func(x)
y_expect = self.x.copy()
for i in numpy.ndindex(self.x.shape):
if self.x[i] < 0:
y_expect[i] *= self.W
assert_allclose(y_expect, y.data)
def test_forward_cpu(self):
self.check_forward(self.x)
@attr.gpu
def test_forward_gpu(self):
self.func.to_gpu()
self.check_forward(cuda.to_gpu(self.x))
def check_backward(self, x_data, y_grad):
x = Variable(x_data)
y = self.func(x)
y.grad = y_grad
y.backward()
func = y.creator
f = lambda: func.forward((x.data, ))
gx, gW = numerical_grad(f, (x.data, func.W), (y.grad, ))
assert_allclose(gx, x.grad)
assert_allclose(gW, func.gW, atol=1e-4)
def test_backward_cpu(self):
self.check_backward(self.x, self.gy)
@attr.gpu
def test_backward_gpu(self):
self.func.to_gpu()
self.check_backward(cuda.to_gpu(self.x), cuda.to_gpu(self.gy))
def setUp(self):
self.func = PReLU()
self.func.W = numpy.random.uniform(
-1, 1, self.func.W.shape).astype(numpy.float32)
self.func.gW.fill(0)
self.W = self.func.W.copy() # fixed on CPU
# Avoid unstability of numerical gradient
self.x = numpy.random.uniform(.5, 1, (4, 3, 2)).astype(numpy.float32)
self.x *= numpy.random.randint(2, size=(4, 3, 2)) * 2 - 1
self.gy = numpy.random.uniform(-1, 1, (4, 3, 2)).astype(numpy.float32)