Skip to content

P
pytensor
提交 5c31ea6b authored 作者: Pascal Lamblin's avatar Pascal Lamblin
浏览文件
操作

Merge pull request #3885 from abergeron/faster_travis

Faster travis
上级 32d27441 62966940
隐藏空白字符变更
内嵌 并排
正在显示 包含 254 行增加 和 200 行删除
.travis.yml
浏览文件 @ 5c31ea6b
......@@ -23,19 +23,25 @@ before_install:
- export PATH=/home/travis/miniconda2/bin:$PATH
- conda update --yes conda
addons:
apt_packages:
- texlive-latex-recommended
- texlive-latex-extra
- texlive-fonts-recommended
- dvipng
install:
- if [[ $TRAVIS_PYTHON_VERSION == '2.6' ]]; then conda create --yes -q -n pyenv mkl python=2.6 numpy=1.7.1 scipy=0.11 nose=1.3.0 pyparsing=1.5 pip flake8=2.3 six=1.9.0 pep8=1.6.2 pyflakes=0.8.1; fi
- if [[ $TRAVIS_PYTHON_VERSION == '3.3' ]]; then conda create --yes -q -n pyenv mkl python=3.3 numpy=1.9.1 scipy=0.14.0 nose=1.3.4 pyparsing=1.5 pip flake8=2.3 six=1.9.0 pep8=1.6.2 pyflakes=0.8.1; fi
- if [[ $TRAVIS_PYTHON_VERSION == '2.6' ]]; then conda create --yes -q -n pyenv mkl python=2.6 numpy=1.7.1 scipy=0.11 nose=1.3.0 pyparsing=1.5 pip flake8=2.3 six=1.9.0 pep8=1.6.2 pyflakes=0.8.1 sphinx; fi
- if [[ $TRAVIS_PYTHON_VERSION == '3.3' ]]; then conda create --yes -q -n pyenv mkl python=3.3 numpy=1.9.1 scipy=0.14.0 nose=1.3.4 pyparsing=1.5 pip flake8=2.3 six=1.9.0 pep8=1.6.2 pyflakes=0.8.1 sphinx; fi
- source activate pyenv
- if [[ $TRAVIS_PYTHON_VERSION == '2.6' ]]; then pip install pydot; fi
- pip install . --no-deps
- pip install -U Sphinx
# command to run tests
env:
- PART="theano/sandbox theano/sparse theano/scalar theano/tensor/nnet theano/scan_module"
- PART="theano/tensor/tests/test_basic.py theano/tensor/signal theano/compile theano/gof theano/misc theano/tests theano/compat"
- PART="-e test_basic.py theano/tensor/tests theano/d3viz"
- PART="theano/tests/test_flake8.py" DOC=1
- PART="-e theano/tests/test_flake8.py theano/compat theano/compile theano/d3viz theano/gof theano/misc theano/sandbox theano/scalar theano/scan_module theano/sparse theano/tests theano/typed_list"
- PART="theano/tensor"
matrix:
fast_finish: true
......@@ -44,8 +50,6 @@ matrix:
env: PART="." THEANO_FLAGS="mode=FAST_COMPILE"
- python: "2.6"
env: PART="." THEANO_FLAGS="mode=FAST_COMPILE,floatX=float32"
- python: 2.6
env: PART="." DOC=1
script:
- export THEANO_FLAGS=$THEANO_FLAGS,warn.ignore_bug_before=all,on_opt_error=raise,on_shape_error=raise,gcc.cxxflags=-pipe
......@@ -60,13 +64,9 @@ script:
- cd $(python -c 'import theano; import os; print(os.path.split(theano.__file__)[0])')
- echo "$PART"
- cd -; cd Theano
- if [[ $DOC == "1" ]]; then python doc/scripts/docgen.py --nopdf; else theano-nose -v $PART; fi
- if [[ $DOC == "1" ]]; then python doc/scripts/docgen.py --test; fi
# - theano-cache list
- theano-nose -v $PART
- if [[ $DOC == "1" ]]; then python doc/scripts/docgen.py --nopdf --check; fi
# - if [[ $DOC == "1" ]]; then python doc/scripts/docgen.py --test --check; fi
#after_script:
after_failure:
- cat /home/travis/.pip/pip.log
#after_success:
cache: apt
doc/sandbox/ccodegen.txt
浏览文件 @ 5c31ea6b
......@@ -132,7 +132,7 @@ The following ResultBase represents a double (we only care about the C part).
.. code-block:: python
class Double(ResultBase):
<snip>
# <snip>
def c_declare(self):
return "double %(name)s;"
def c_init(self):
......@@ -153,7 +153,7 @@ The following ResultBase represents addition of two nonnegative doubles (we only
.. code-block:: python
class Add(Op):
<snip>
# <snip>
def c_var_names(self):
return "[['x', 'y'], ['z']]"
def c_validate_update(self):
......@@ -170,7 +170,7 @@ Generating a C function
For the example Op, the generated C function will typically look like this:
.. code-block:: python
.. code-block:: cpp
void add(PyObject* storage_x, PyObject* storage_y, PyObject* storage_z) {
PyObject* py_x = PyList_GET_ITEM(storage_x, 0); Py_XINCREF(py_x); // automatic
......@@ -232,16 +232,16 @@ Here is a sketch of the struct equivalent of the previous function:
double z; // z.c_declare
void init(PyObject* storage_x, PyObject* storage_y, PyObject* storage_z) {
<set the struct members of the same names>
<init the struct members corresponding to z>
// <set the struct members of the same names>
// <init the struct members corresponding to z>
}
void cleanup(void) {
<cleanup z>
// <cleanup z>
}
void run(void) {
<same code as before minus z's cleanup>
// <same code as before minus z's cleanup>
}
add() { this->init(); }
......
doc/scripts/docgen.py
浏览文件 @ 5c31ea6b
......@@ -16,7 +16,7 @@ if __name__ == '__main__':
opts, args = getopt.getopt(
sys.argv[1:],
'o:f:',
['rst', 'help', 'nopdf', 'cache', 'test'])
['rst', 'help', 'nopdf', 'cache', 'check', 'test'])
options.update(dict([x, y or True] for x, y in opts))
if options['--help']:
print('Usage: %s [OPTIONS] [files...]' % sys.argv[0])
......@@ -25,6 +25,7 @@ if __name__ == '__main__':
print(' --rst: only compile the doc (requires sphinx)')
print(' --nopdf: do not produce a PDF file from the doc, only HTML')
print(' --test: run all the code samples in the documentaton')
print(' --check: treat warnings as errors')
print(' --help: this help')
print('If one or more files are specified after the options then only '
'those files will be built. Otherwise the whole tree is '
......@@ -54,17 +55,21 @@ if __name__ == '__main__':
pythonpath = os.pathsep.join([throot, pythonpath])
sys.path[0:0] = [throot] # We must not use os.environ.
def call_sphinx(builder, workdir, extraopts=None):
def call_sphinx(builder, workdir):
import sphinx
if extraopts is None:
if options['--check']:
extraopts = ['-W']
else:
extraopts = []
if not options['--cache'] and files is None:
extraopts.append('-E')
docpath = os.path.join(throot, 'doc')
inopt = [docpath, workdir]
if files is not None:
inopt.extend(files)
sphinx.build_main(['', '-b', builder] + extraopts + inopt)
ret = sphinx.build_main(['', '-b', builder] + extraopts + inopt)
if ret != 0:
sys.exit(ret)
if options['--all'] or options['--rst']:
mkdir("doc")
......@@ -92,5 +97,6 @@ if __name__ == '__main__':
mkdir("doc")
sys.path[0:0] = [os.path.join(throot, 'doc')]
call_sphinx('doctest', '.')
# To go back to the original current directory.
os.chdir(currentdir)
doc/tutorial/extending_theano_c.txt
浏览文件 @ 5c31ea6b
......@@ -997,7 +997,7 @@ down computation on the GPU, but it is enabled by default on the CPU).
Then you must start Python inside GDB and in it start your Python
process (e.g. theano-nose):
.. code-block::
.. code-block:: sh
$gdb python
(gdb)r bin/theano-nose theano/
......
theano/misc/pkl_utils.py
浏览文件 @ 5c31ea6b
......@@ -124,12 +124,15 @@ if PY3:
Examples
--------
with open(fname, 'rb') as fp:
if PY3:
u = CompatUnpickler(fp, encoding="latin1")
else:
u = CompatUnpickler(fp)
mat = u.load()
::
with open(fname, 'rb') as fp:
if PY3:
u = CompatUnpickler(fp, encoding="latin1")
else:
u = CompatUnpickler(fp)
mat = u.load()
"""
pass
......@@ -145,13 +148,15 @@ else:
Examples
--------
with open(fname, 'rb') as fp:
if PY3:
u = CompatUnpickler(fp, encoding="latin1")
else:
u = CompatUnpickler(fp)
::
with open(fname, 'rb') as fp:
if PY3:
u = CompatUnpickler(fp, encoding="latin1")
else:
u = CompatUnpickler(fp)
mat = u.load()
mat = u.load()
"""
pass
......
theano/sparse/basic.py
浏览文件 @ 5c31ea6b
......@@ -790,7 +790,7 @@ shape
Returns
-------
matrix
sparse matrix
A sparse matrix having the properties specified by the inputs.
Notes
......@@ -820,7 +820,7 @@ shape
Returns
-------
matrix
sparse matrix
A sparse matrix having the properties specified by the inputs.
Notes
......@@ -1053,7 +1053,7 @@ x
Returns
-------
matrix
theano.tensor.matrix
A dense matrix, the same as `x`.
Notes
......@@ -1121,7 +1121,7 @@ x
Returns
-------
matrix
sparse matrix
The same as `x` in a sparse csr matrix format.
"""
......@@ -1137,7 +1137,7 @@ x
Returns
-------
matrix
sparse matrix
The same as `x` in a sparse csc matrix format.
"""
......@@ -1187,7 +1187,7 @@ index
Returns
-------
matrix
sparse matrix
The corresponding rows in `x`.
"""
......@@ -1284,7 +1284,7 @@ index
Returns
-------
vector
theano.tensor.vector
The corresponding elements in `x`.
"""
......@@ -1436,7 +1436,8 @@ index
Returns
-------
The corresponding slice in `x`.
sparse matrix
The corresponding slice in `x`.
Notes
......@@ -1511,7 +1512,7 @@ index
Returns
-------
scalar
theano.tensor.scalar
The corresponding item in `x`.
Notes
......@@ -1566,7 +1567,7 @@ x
Returns
-------
matrix
sparse matrix
`x` transposed.
Notes
......@@ -1617,7 +1618,7 @@ x
Returns
-------
matrix
sparse matrix
-`x`.
Notes
......@@ -1930,7 +1931,7 @@ x
Returns
-------
vector
theano.tensor.vector
A dense vector representing the diagonal elements.
Notes
......@@ -1985,7 +1986,7 @@ x
Returns
-------
matrix
sparse matrix
A sparse matrix having `x` as diagonal.
Notes
......@@ -2044,7 +2045,7 @@ x
Returns
-------
matrix
sparse matrix
The same as `x` with indices sorted.
Notes
......@@ -2854,7 +2855,7 @@ y
Returns
-------
matrix
matrix variable
`x` == `y`
Notes
......@@ -2875,7 +2876,7 @@ y
Returns
-------
matrix
matrix variable
`x` != `y`
Notes
......@@ -2896,7 +2897,7 @@ y
Returns
-------
matrix
matrix variable
`x` < `y`
Notes
......@@ -2917,7 +2918,7 @@ y
Returns
-------
matrix
matrix variable
`x` > `y`
Notes
......@@ -2937,6 +2938,7 @@ y
Returns
-------
matrix variable
`x` <= `y`
Notes
......@@ -2957,7 +2959,7 @@ y
Returns
-------
matrix
matrix variable
`x` >= `y`
Notes
......@@ -3199,7 +3201,7 @@ x
Returns
-------
matrix
sparse matrix
Exactly `x` but with a data attribute exempt of zeros.
Notes
......@@ -4112,7 +4114,7 @@ p
Returns
-------
matrix
sparse matrix
A dense matrix containing the dot product of `x` by `y`.T only
where `p` is 1.
......
theano/tensor/nnet/abstract_conv.py
浏览文件 @ 5c31ea6b
......@@ -132,68 +132,80 @@ def conv2d_grad_wrt_inputs(output_grad,
used by the convolution, such that the output_grad is upsampled
to the input shape.
:type output_grad: symbolic 4D tensor.
:param output_grad: mini-batch of feature map stacks, of shape
(batch size, input channels, input rows, input columns).
This is the tensor that will be upsampled or the output
gradient of the convolution whose gradient will be taken
with respect to the input of the convolution.
See the optional parameter ``output_grad_shape``.
:type filters: symbolic 4D tensor.
:param filters: set of filters used in CNN layer of shape
(output channels, input channels, filter rows, filter columns).
See the optional parameter ``filter_shape``.
:type output_grad_shape: None, tuple/list of len 4 of int or
Constant variable.
:param output_grad_shape: The shape of the output_grad parameter.
Optional, possibly used to choose an optimal implementation.
You can give ``None`` for any element of the list to specify that this
Parameters
----------
output_grad : symbolic 4D tensor
mini-batch of feature map stacks, of shape (batch size, input
channels, input rows, input columns). This is the tensor that
will be upsampled or the output gradient of the convolution
whose gradient will be taken with respect to the input of the
convolution. See the optional parameter
``output_grad_shape``.
filters : symbolic 4D tensor
set of filters used in CNN layer of shape (output channels,
input channels, filter rows, filter columns). See the
optional parameter ``filter_shape``.
output_grad_shape : list of 4 symbolic or real ints
The shape of the output_grad parameter. Optional, possibly
used to choose an optimal implementation. You can give
``None`` for any element of the list to specify that this
element is not known at compile time.
input_shape : list of 2 symbolic or real ints
The shape (row and column size) of the input (upsampled)
parameter. Not Optional, since given the output_grad_shape
and the subsample values, multiple input_shape may be
plausible.
filter_shape : list of 4 symbolic or real ints
The shape of the filters parameter. Optional, possibly used
to choose an optimal implementation. You can give ``None``
for any element of the list to specify that this element is
not known at compile time.
border_mode : str, int or tuple of two int
Either of the following:
:type input_shape: tuple/list of len 2 of int or Constant variable.
:param input_shape: The shape (row and column size) of the
input (upsampled) parameter.
Not Optional, since given the output_grad_shape and the subsample values,
multiple input_shape may be plausible.
``'valid'``
apply filter wherever it completely overlaps with the
input. Generates output of shape: input shape - filter
shape + 1
:type filter_shape: None, tuple/list of len 4 of int or Constant variable
:param filter_shape: The shape of the filters parameter.
Optional, possibly used to choose an optimal implementation.
You can give ``None`` for any element of the list to specify that this
element is not known at compile time.
``'full'``
apply filter wherever it partly overlaps with the input.
Generates output of shape: input shape + filter shape - 1
:type border_mode: str, int or tuple of two int
:param border_mode: Either of the following:
* ``'valid'``: apply filter wherever it completely overlaps with the
input. Generates output of shape: input shape - filter shape + 1
* ``'full'``: apply filter wherever it partly overlaps with the input.
Generates output of shape: input shape + filter shape - 1
* ``'half'``: pad input with a symmetric border of ``filter rows // 2``
rows and ``filter columns // 2`` columns, then perform a valid
convolution. For filters with an odd number of rows and columns, this
leads to the output shape being equal to the input shape.
* ``int``: pad input with a symmetric border of zeros of the given
width, then perform a valid convolution.
* ``(int1, int2)``: pad input with a symmetric border of ``int1`` rows
and ``int2`` columns, then perform a valid convolution.
:type subsample: tuple of len 2, the subsampling used in the forward pass
of the convolutional operation.
:param subsample: factor by which to subsample the output.
Also called strides elsewhere.
``'half'``
pad input with a symmetric border of ``filter rows // 2``
rows and ``filter columns // 2`` columns, then perform a
valid convolution. For filters with an odd number of rows
and columns, this leads to the output shape being equal to
the input shape.
:type filter_flip: bool
:param filter_flip: If ``True``, will flip the filter rows and columns
before sliding them over the input. This operation is normally referred
to as a convolution, and this is the default. If ``False``, the filters
are not flipped and the operation is referred to as a
cross-correlation.
``int``
pad input with a symmetric border of zeros of the given
width, then perform a valid convolution.
``(int1, int2)``
pad input with a symmetric border of ``int1`` rows and
``int2`` columns, then perform a valid convolution.
subsample : tuple of len 2
The subsampling used in the forward pass. Also called strides
elsewhere.
filter_flip : bool
If ``True``, will flip the filter rows and columns before
sliding them over the input. This operation is normally
referred to as a convolution, and this is the default. If
``False``, the filters are not flipped and the operation is
referred to as a cross-correlation.
Returns
-------
symbolic 4D tensor
set of feature maps generated by convolutional layer. Tensor
is of shape (batch size, output channels, output rows, output
columns)
:rtype: symbolic 4D tensor.
:return: set of feature maps generated by convolutional layer. Tensor is
of shape (batch size, output channels, output rows, output columns)
Notes
-----
:note: If CuDNN is available, it will be used on the
GPU. Otherwise, it is the *CorrMM* convolution that will be used
......@@ -224,71 +236,81 @@ def conv2d_grad_wrt_weights(input,
"""This function will build the symbolic graph for getting the
gradient of the output of a convolution (output_grad) w.r.t its wights.
:type input: symbolic 4D tensor.
:param input: mini-batch of feature map stacks, of shape
(batch size, input channels, input rows, input columns).
This is the input of the convolution in the forward pass.
Parameters
----------
input : symbolic 4D tensor
mini-batch of feature map stacks, of shape (batch size, input
channels, input rows, input columns). This is the input of
the convolution in the forward pass.
output_grad : symbolic 4D tensor
mini-batch of feature map stacks, of shape (batch size, input
channels, input rows, input columns). This is the gradient of
the output of convolution.
filters : symbolic 4D tensor.
set of filters used in CNN layer of shape (output channels,
input channels, filter rows, filter columns). See the
optional parameter ``filter_shape``.
output_grad_shape : list of 4 ints or Constant variables
The shape of the input parameter. Optional, possibly used to
choose an optimal implementation. You can give ``None`` for
any element of the list to specify that this element is not
known at compile time.
input_shape : list of 2 ints or Constant variables
The shape of the input parameter. This parameter indicates
the row and column size of the input in the forward pass.
Optional, possibly used to choose an optimal implementation.
You can give ``None`` for any element of the list to specify
that this element is not known at compile time.
filter_shape : list of 4 ints or Constant variables
The shape of the filters parameter. Not Optional, since given
the output_grad_shape and the input_shape, multiple
filter_shape may be plausible.
border_mode : str, int or tuple of two ints
Either of the following:
:type output_grad: symbolic 4D tensor.
:param output_grad: mini-batch of feature map stacks, of shape
(batch size, input channels, input rows, input columns).
This is the gradient of the output of convolution.
``'valid'``
apply filter wherever it completely overlaps with the
input. Generates output of shape: input shape - filter
shape + 1
:type filters: symbolic 4D tensor.
:param filters: set of filters used in CNN layer of shape
(output channels, input channels, filter rows, filter columns).
See the optional parameter ``filter_shape``.
``'full'``
apply filter wherever it partly overlaps with the input.
Generates output of shape: input shape + filter shape - 1
:type output_grad_shape: None, tuple/list of len 4 of int
or Constant variable.
:param output_grad_shape: The shape of the input parameter.
Optional, possibly used to choose an optimal implementation.
You can give ``None`` for any element of the list to specify that this
element is not known at compile time.
``'half'``
pad input with a symmetric border of ``filter rows // 2``
rows and ``filter columns // 2`` columns, then perform a
valid convolution. For filters with an odd number of rows
and columns, this leads to the output shape being equal to
the input shape.
:type input_shape: tuple/list of len 2 of int or Constant variable.
:param input_shape: The shape of the input parameter.
This parameter indicates the row and column size of the input
in the forward pass.
Optional, possibly used to choose an optimal implementation.
You can give ``None`` for any element of the list to specify that this
element is not known at compile time.
``int``
pad input with a symmetric border of zeros of the given
width, then perform a valid convolution.
:type filter_shape: None, tuple/list of len 4 of int or Constant variable.
:param filter_shape: The shape of the filters parameter.
Not Optional, since given the output_grad_shape and the input_shape,
multiple filter_shape may be plausible.
:type border_mode: str, int or tuple of two int
:param border_mode: Either of the following:
* ``'valid'``: apply filter wherever it completely overlaps with the
input. Generates output of shape: input shape - filter shape + 1
* ``'full'``: apply filter wherever it partly overlaps with the input.
Generates output of shape: input shape + filter shape - 1
* ``'half'``: pad input with a symmetric border of ``filter rows // 2``
rows and ``filter columns // 2`` columns, then perform a valid
convolution. For filters with an odd number of rows and columns, this
leads to the output shape being equal to the input shape.
* ``int``: pad input with a symmetric border of zeros of the given
width, then perform a valid convolution.
* ``(int1, int2)``: pad input with a symmetric border of ``int1`` rows
and ``int2`` columns, then perform a valid convolution.
:type subsample: tuple of len 2, the subsampling used in the forward pass
of the convolutional operation.
:param subsample: factor by which to subsample the output.
Also called strides elsewhere.
``(int1, int2)``
pad input with a symmetric border of ``int1`` rows and
``int2`` columns, then perform a valid convolution.
:type filter_flip: bool
:param filter_flip: If ``True``, will flip the filter rows and columns
before sliding them over the input. This operation is normally referred
to as a convolution, and this is the default. If ``False``, the filters
are not flipped and the operation is referred to as a
cross-correlation.
subsample : tuple of len 2
The subsampling used in the forward pass of the convolutional
operation. Also called strides elsewhere.
filter_flip : bool
If ``True``, will flip the filter rows and columns before
sliding them over the input. This operation is normally
referred to as a convolution, and this is the default. If
``False``, the filters are not flipped and the operation is
referred to as a cross-correlation.
Returns
-------
symbolic 4D tensor
set of feature maps generated by convolutional layer. Tensor
is of shape (batch size, output channels, output rows, output
columns)
:rtype: symbolic 4D tensor.
:return: set of feature maps generated by convolutional layer. Tensor is
of shape (batch size, output channels, output rows, output columns)
Notes
-----
:note: If CuDNN is available, it will be used on the
GPU. Otherwise, it is the *CorrMM* convolution that will be used
......
theano/tensor/signal/pool.py
浏览文件 @ 5c31ea6b
......@@ -30,6 +30,7 @@ def max_pool_2d_same_size(input, patch_size):
of non-overlapping patches of size (patch_size[0],patch_size[1]) to zero,
keeping only the maximum values. The output has the same dimensions as
the input.
Parameters
----------
input : 4-D theano tensor of input images
......@@ -37,6 +38,7 @@ def max_pool_2d_same_size(input, patch_size):
patch_size : tuple of length 2
Size of the patch (patch height, patch width).
(2,2) will retain only one non-zero value per patch of 4 values.
"""
output = Pool(patch_size, True)(input)
outs = MaxPoolGrad(patch_size, True)(input, output, output)
......@@ -49,6 +51,7 @@ def pool_2d(input, ds, ignore_border=None, st=None, padding=(0, 0),
Takes as input a N-D tensor, where N >= 2. It downscales the input image by
the specified factor, by keeping only the maximum value of non-overlapping
patches of size (ds[0],ds[1])
Parameters
----------
input : N-D theano tensor of input images
......@@ -64,13 +67,14 @@ def pool_2d(input, ds, ignore_border=None, st=None, padding=(0, 0),
next pool region. If st is None, it is considered equal to ds
(no overlap on pooling regions).
padding : tuple of two ints
(pad_h, pad_w), pad zeros to extend beyond four borders
of the images, pad_h is the size of the top and bottom margins,
and pad_w is the size of the left and right margins.
(pad_h, pad_w), pad zeros to extend beyond four borders of the
images, pad_h is the size of the top and bottom margins, and
pad_w is the size of the left and right margins.
mode : {'max', 'sum', 'average_inc_pad', 'average_exc_pad'}
Operation executed on each window. `max` and `sum` always exclude
the padding in the computation. `average` gives you the choice to
include or exclude it.
"""
if input.ndim < 2:
raise NotImplementedError('pool_2d requires a dimension >= 2')
......@@ -121,15 +125,17 @@ class Pool(Op):
For N-dimensional tensors, consider that the last two dimensions span
images. This Op downsamples these images by taking the max, sum or average
over different patch.
The constructor takes the max, sum or average or different input patches.
Parameters
----------
ds : list or tuple of two ints
Downsample factor over rows and column.
ds indicates the pool region size.
ignore_border : bool
If ds doesn't divide imgshape, do we include an extra row/col of partial
downsampling (False) or ignore it (True).
If ds doesn't divide imgshape, do we include an extra row/col
of partial downsampling (False) or ignore it (True).
st : list or tuple of two ints or None
Stride size, which is the number of shifts over rows/cols to get the
next pool region. If st is None, it is considered equal to ds
......@@ -141,6 +147,7 @@ class Pool(Op):
mode : {'max', 'sum', 'average_inc_pad', 'average_exc_pad'}
('average_inc_pad' excludes the padding from the count,
'average_exc_pad' include it)
"""
__props__ = ('ds', 'ignore_border', 'st', 'padding', 'mode')
......@@ -150,6 +157,7 @@ class Pool(Op):
"""
Return the shape of the output from this op, for input of given
shape and flags.
Parameters
----------
imgshape : tuple, list, or similar of integer or scalar Theano variable
......@@ -168,12 +176,14 @@ class Pool(Op):
(pad_h, pad_w), pad zeros to extend beyond four borders
of the images, pad_h is the size of the top and bottom margins,
and pad_w is the size of the left and right margins.
Returns
-------
list
The shape of the output from this op, for input of given shape.
This will have the same length as imgshape, but with last two
elements reduced as per the downsampling & ignore_border flags.
"""
if len(imgshape) < 2:
raise TypeError('imgshape must have at least two elements '
......@@ -528,27 +538,36 @@ class PoolGrad(Op):
def out_shape(imgshape, ds, ignore_border=False, st=None, padding=(0, 0)):
"""Return the shape of the output from this op, for input of given
shape and flags.
:param imgshape: the shape of a tensor of images. The last two elements
are interpreted as the number of rows, and the number of cols.
:type imgshape: tuple, list, or similar of integer or
scalar Theano variable.
:param ds: downsample factor over rows and columns
this parameter indicates the size of the pooling region
:type ds: list or tuple of two ints
:param st: the stride size. This is the distance between the pooling
regions. If it's set to None, in which case it equlas ds.
:type st: list or tuple of two ints
:param ignore_border: if ds doesn't divide imgshape, do we include an
extra row/col of partial downsampling (False) or ignore it (True).
:type ignore_border: bool
:param padding: (pad_h, pad_w), pad zeros to extend beyond four borders
of the images, pad_h is the size of the top and bottom margins,
and pad_w is the size of the left and right margins.
:type padding: tuple of two ints
:rtype: list
:returns: the shape of the output from this op, for input of given
shape. This will have the same length as imgshape, but with last
two elements reduced as per the downsampling & ignore_border flags.
Parameters
----------
imgshape : tuple of integers or scalar Theano variables
the shape of a tensor of images. The last two elements are
interpreted as the number of rows, and the number of cols.
ds : tuple of two ints
downsample factor over rows and columns this parameter
indicates the size of the pooling region
st : tuple of two ints
the stride size. This is the distance between the pooling
regions. If it's set to None, in which case it equlas ds.
ignore_border : bool
if ds doesn't divide imgshape, do we include an extra
row/col of partial downsampling (False) or ignore it
(True).
padding : tuple of two ints
(pad_h, pad_w), pad zeros to extend beyond four borders of
the images, pad_h is the size of the top and bottom
margins, and pad_w is the size of the left and right
margins.
Returns
-------
list :
the shape of the output from this op, for input of given
shape. This will have the same length as imgshape, but
with last two elements reduced as per the downsampling &
ignore_border flags.
"""
if len(imgshape) < 2:
raise TypeError('imgshape must have at least two elements '
......
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