def test_average_pool_invalid(
device, x_shape, ksize, stride, pad, pad_mode, float_dtype):
with pytest.raises(chainerx.DimensionError):
chainerx.average_pool(
**_create_average_pool_args(
chainerx, device, x_shape, ksize, stride, pad, pad_mode,
float_dtype))
def test_average_pool_invalid(
device, x_shape, ksize, stride, pad, pad_mode, float_dtype):
x = numpy.random.uniform(-1, 1, x_shape).astype(float_dtype)
x = chainerx.array(x)
pad_mode_kwargs = _get_pad_mode_kwargs(pad_mode, True)
with pytest.raises(chainerx.DimensionError):
chainerx.average_pool(
x, ksize=ksize, stride=stride, pad=pad, **pad_mode_kwargs)
def test_average_pool_invalid(device, x_shape, ksize, stride, pad, pad_mode,
float_dtype):
x = numpy.random.uniform(-1, 1, x_shape).astype(float_dtype)
x = chainerx.array(x)
pad_mode_kwargs = _get_pad_mode_kwargs(pad_mode, True)
with pytest.raises(chainerx.DimensionError):
chainerx.average_pool(x,
ksize=ksize,
stride=stride,
pad=pad,
**pad_mode_kwargs)
def forward_chainerx(self, inputs):
x, = inputs
pad_mode_kwargs = _get_pad_mode_kwargs(self.pad_mode, True)
y = chainerx.average_pool(
x, ksize=self.ksize, stride=self.stride, pad=self.pad,
**pad_mode_kwargs)
return y,
def forward_chainerx(self, inputs):
x, = inputs
pad_mode_kwargs = _get_pad_mode_kwargs(self.pad_mode, True)
y = chainerx.average_pool(x,
ksize=self.ksize,
stride=self.stride,
pad=self.pad,
**pad_mode_kwargs)
return y,
def __call__(self, x):
h = self.bn1(self.conv1(x))
h = chx.max_pool(chx.maximum(0, h), 3, stride=2)
h = self.res2(h)
h = self.res3(h)
h = self.res4(h)
h = self.res5(h)
h = chx.average_pool(h, 7, stride=1)
h = self.fc(h)
return h
def __call__(self, x):
h = self.bn1(self.conv1(x))
h = chx.max_pool(chx.maximum(0, h), 3, stride=2)
h = self.res2(h)
h = self.res3(h)
h = self.res4(h)
h = self.res5(h)
h = chx.average_pool(h, 7, stride=1)
h = self.fc(h)
return h
def test_average_pool(
device, x_shape, ksize, stride, pad, pad_mode, float_dtype):
if device.backend.name == 'cuda' and len(ksize) != 2 and len(ksize) != 3:
# cuDNN supports only 2 and 3 spatial dimensions.
return chainerx.testing.ignore()
def create_args(xp):
return _create_average_pool_args(
xp, device, x_shape, ksize, stride, pad, pad_mode, float_dtype)
chainerx.testing.assert_allclose(
chainerx.average_pool(**create_args(chainerx)),
chainer.functions.average_pooling_nd(**create_args(numpy)).data)
def test_average_pool(device, x_shape, ksize, stride, pad, pad_mode,
float_dtype):
if device.backend.name == 'cuda' and len(ksize) != 2 and len(ksize) != 3:
# cuDNN supports only 2 and 3 spatial dimensions.
return chainerx.testing.ignore()
def create_args(xp):
return _create_average_pool_args(xp, device, x_shape, ksize, stride,
pad, pad_mode, float_dtype)
chainerx.testing.assert_allclose(
chainerx.average_pool(**create_args(chainerx)),
chainer.functions.average_pooling_nd(**create_args(numpy)).data)
def forward_chainerx(self, inputs):
x, = inputs
if x.device.backend.name == 'cuda' and self.ndim not in (2, 3):
return chainer.Fallback
if self.pad_value == 0:
pad_mode = 'zero'
elif self.pad_value is None:
pad_mode = 'ignore'
else:
assert False
return chainerx.average_pool(
x, self.ksize, self.stride, self.pad, pad_mode),
def forward_chainerx(self, inputs):
x, = inputs
if x.device.backend.name == 'cuda' and self.ndim not in (2, 3):
return chainer.Fallback
if self.pad_value == 0:
pad_mode = 'zero'
elif self.pad_value is None:
pad_mode = 'ignore'
else:
assert False
return chainerx.average_pool(x, self.ksize, self.stride, self.pad,
pad_mode),
def test_average_pool_invalid(device, x_shape, ksize, stride, pad, pad_mode,
float_dtype):
with pytest.raises(chainerx.DimensionError):
chainerx.average_pool(
**_create_average_pool_args(chainerx, device, x_shape, ksize,
stride, pad, pad_mode, float_dtype))
