Skip to content

P
pytensor
提交 4eda607b authored 作者: Tegan Maharaj's avatar Tegan Maharaj
浏览文件
操作

fixed pep and flake errors

上级 e183ab3b
隐藏空白字符变更
内嵌 并排
正在显示 包含 57 行增加 和 78 行删除
theano/tensor/nnet/tests/test_abstract_conv.py
浏览文件 @ 4eda607b
...@@ -33,9 +33,6 @@ from theano.tensor.nnet.ConvGrad3D import ConvGrad3D ...@@ -33,9 +33,6 @@ from theano.tensor.nnet.ConvGrad3D import ConvGrad3D
from theano.tensor.nnet.ConvTransp3D import ConvTransp3D from theano.tensor.nnet.ConvTransp3D import ConvTransp3D
def conv2d_corr(inputs, filters, border_mode="valid", def conv2d_corr(inputs, filters, border_mode="valid",
subsample=(1, 1), conv_mode='conv', subsample=(1, 1), conv_mode='conv',
filter_dilation=(1, 1)): filter_dilation=(1, 1)):
...@@ -1592,107 +1589,90 @@ class TestConv2dTranspose(unittest.TestCase): ...@@ -1592,107 +1589,90 @@ class TestConv2dTranspose(unittest.TestCase):
#TM copying the TestConvTypes setup, not sure exactly what unittest.TestCase is # TM copying the TestConvTypes setup, not sure exactly what unittest.TestCase is
class TestConv2dGrads(unittest.TestCase): class TestConv2dGrads(unittest.TestCase):
def __init__(self): def setUp(self):
self.random_stream = numpy.random.RandomState(unittest_tools.fetch_seed())
self.random_stream = numpy.random.RandomState(utt.fetch_seed())
self.inputs_shapes = [(8,1,12,12), (1,1,5,5), (1,1,5,6), (1,1,6,6)]
self.filters_shapes = [(5,1,2,2), (1,1,3,3)] self.inputs_shapes = [(8, 1, 12, 12), (1, 1, 5, 5), (1, 1, 5, 6), (1, 1, 6, 6)]
self.filters_shapes = [(5, 1, 2, 2), (1, 1, 3, 3)]
self.subsamples = [(1,1), (2,2)] self.subsamples = [(1, 1), (2, 2)]
self.border_modes = ["valid", "full"] self.border_modes = ["valid", "full"]
self.filter_flip = [True, False] self.filter_flip = [True, False]
self.output_grad = T.tensor4() self.output_grad = theano.tensor.tensor4()
self.output_grad_wrt = T.tensor4() self.output_grad_wrt = theano.tensor.tensor4()
self.filters = T.tensor4() self.filters = theano.tensor.tensor4()
self.x = T.tensor4('x', theano.config.floatX) #inputs self.x = theano.tensor.tensor4('x', theano.config.floatX) # inputs
self.w = T.tensor4('w', theano.config.floatX) #weights self.w = theano.tensor.tensor4('w', theano.config.floatX) # weights
#first two must be wrong; I'm not sure how to use TestCase
if unittest.TestCase == 'inputs':
test_conv2d_grad_wrt_inputs()
elif unittest.TestCase =='weights':
test_conv2d_grad_wrt_weights():
else:
self.test_conv2_grads_wrt_input_and_weights()
def test_conv2d_grad_wrt_inputs(self):
def test_conv2d_grad_wrt_inputs():
"""Compares calculated abstract grads wrt inputs with the fwd grads """Compares calculated abstract grads wrt inputs with the fwd grads
This method checks the outputs of conv2_grad_wrt_inputs against This method checks the outputs of conv2_grad_wrt_inputs against
the outputs of T.nnet.conv forward grads to make sure the the outputs of T.nnet.conv forward grads to make sure the
results are the same. results are the same.
""" """
for (in_shape, fltr_shape) in zip(self.inputs_shapes, self.filters_shapes): for (in_shape, fltr_shape) in zip(self.inputs_shapes, self.filters_shapes):
for bm in self.border_modes: for bm in self.border_modes:
for ss in self.subsamples: for ss in self.subsamples:
for ff in self.filter_flip: for ff in self.filter_flip:
if filter_flip = True: if self.filter_flip:
fltr_shape = transpose(fltr_shape) #conv2d doesn't seem to have filter_flip fltr_shape = fltr_shape[::1] # conv2d doesn't seem to have filter_flip
conv_out = T.nnet.conv.conv2d(self.x, conv_out = theano.tensor.nnet.conv.conv2d(self.x,
filters = self.filters, filters=self.filters,
border_mode = bm, border_mode=bm,
subsample = ss, subsample=ss,
image_shape = in_shape image_shape=in_shape,
filter_shape = fltr_shape filter_shape=fltr_shape
) )
conv_grad = theano.grad(wrt=[x], known_grads={conv_out: self.output_grad}) conv_grad = theano.grad(conv_out.sum(), wrt=[self.x], known_grads={conv_out: self.output_grad})
f_prime = theano.function([x, output_grad, filters], conv_grad) f_prime = theano.function([self.x, self.output_grad, self.filters], conv_grad)
conv_wrt_i_out = T.nnet.conv.abstract_conv.conv2d_grad_wrt_inputs(self.output_grad_wrt, conv_wrt_i_out = theano.tensor.nnet.conv.abstract_conv.conv2d_grad_wrt_inputs(self.output_grad_wrt,
filters = self.filters, filters=self.filters,
border_mode = bm, border_mode=bm,
subsample = ss, subsample=ss,
input_shape = in_shape, input_shape=in_shape,
filter_shape = fltr_shape, filter_shape=fltr_shape,
filter_flip = ff filter_flip=ff
) )
f = theano.function([x, output_grad_wrt, filters], conv_wrt_i_out) f = theano.function([self.x, self.output_grad_wrt, self.filters], conv_wrt_i_out)
utt.assert_allclose(f, f_prime) utt.assert_allclose(f, f_prime)
def test_conv2d_grad_wrt_weights(): def test_conv2d_grad_wrt_weights(self):
"""Compares calculated abstract grads wrt weights with the fwd grads """Compares calculated abstract grads wrt weights with the fwd grads
This method checks the outputs of conv2_grad_wrt_weights against This method checks the outputs of conv2_grad_wrt_weights against
the outputs of T.nnet.conv forward grads to make sure the the outputs of T.nnet.conv forward grads to make sure the
results are the same. results are the same.
""" """
for (in_shape, fltr_shape) in zip(self.inputs_shapes, self.filters_shapes): for (in_shape, fltr_shape) in zip(self.inputs_shapes, self.filters_shapes):
for bm in self.border_modes: for bm in self.border_modes:
for ss in self.subsamples: for ss in self.subsamples:
for ff in self.filter_flip: for ff in self.filter_flip:
if filter_flip = True: if self.filter_flip:
fltr_shape = transpose(fltr_shape) #conv2d doesn't seem to have filter_flip fltr_shape = fltr_shape[::1] # conv2d doesn't seem to have filter_flip
conv_out = T.nnet.conv.conv2d(self.w, conv_out = theano.tensor.nnet.conv.conv2d(self.w,
filters = self.filters, filters=self.filters,
border_mode = bm, border_mode=bm,
subsample = ss, subsample=ss,
image_shape = in_shape, image_shape=in_shape,
filter_shape = fltr_shape filter_shape=fltr_shape
) )
conv_grad = theano.grad(wrt=[w], known_grads={conv_out: output_grad}) conv_grad = theano.grad(conv_out.sum(), wrt=[self.w], known_grads={conv_out: self.output_grad})
f_prime = theano.function([w, output_grad, filters], conv_grad) f_prime = theano.function([self.w, self.output_grad, self.filters], conv_grad)
conv_wrt_w_out = T.nnet.conv.abstract_conv.conv2d_grad_wrt_weights(self.output_grad_wrt, conv_wrt_w_out = theano.tensor.nnet.conv.abstract_conv.conv2d_grad_wrt_weights(self.output_grad_wrt,
filters = self.filters, filters=self.filters,
border_mode = bm, border_mode=bm,
subsample = ss, subsample=ss,
input_shape = in_shape, input_shape=in_shape,
filter_shape = fltr_shape, filter_shape=fltr_shape,
filter_flip = ff filter_flip=ff
) )
f = theano.function([w, output_grad_wrt, filters], conv_wrt_w_out) f = theano.function([self.w, self.output_grad_wrt, self.filters], conv_wrt_w_out)
utt.assert_allclose(f, f_prime) utt.assert_allclose(f, f_prime)
def test_conv2_grads_wrt_input_and_weights():
test_conv2d_grad_wrt_inputs()
test_conv2d_grad_wrt_weights()
\ No newline at end of file
Markdown 格式
0% 或
您添加了 0 人 到此讨论。请谨慎行事。
请先完成此评论的编辑！
请 注册 或者 后发表评论