Skip to content

P
pytensor
提交 27c8da22 authored 作者: Chiheb Trabelsi's avatar Chiheb Trabelsi
浏览文件
操作

blas.py has been modified in order to respect the flake8 style.

blas.py do not contain long lines.
上级 ed7759fb
隐藏空白字符变更
内嵌 并排
正在显示 包含 76 行增加 和 65 行删除
theano/sandbox/cuda/blas.py
浏览文件 @ 27c8da22
from __future__ import absolute_import, print_function, division
import copy
import os
import logging
_logger = logging.getLogger(__name__)
from six import integer_types
from six.moves import StringIO, reduce
import theano
from theano import Apply
from theano import tensor
......@@ -15,6 +11,7 @@ from theano.sandbox.cuda import GpuOp
from theano.sandbox.cuda.basic_ops import (as_cuda_ndarray_variable,
gpu_contiguous)
from theano.tensor import as_tensor_variable
_logger = logging.getLogger(__name__)
class GpuBatchedDot(GpuOp):
......@@ -29,11 +26,11 @@ class GpuBatchedDot(GpuOp):
assert inp1.dtype == "float32"
assert inp2.dtype == "float32"
assert inp1.ndim == 3 # (batch, a, b)
assert inp1.ndim == 3 # (batch, a, b)
assert inp2.ndim == 3
return theano.Apply(self, [inp1, inp2],
[self.output_type(inp1, inp2)()])
[self.output_type(inp1, inp2)()])
def output_type(self, inp1, inp2):
return CudaNdarrayType(
......@@ -183,8 +180,7 @@ class GpuBatchedDot(GpuOp):
}
} else {
// copy inputs if not contiguous
""" +
("\n".join("""
""" + ("\n".join("""
if (( CudaNdarray_HOST_DIMS(%(var)s)[0] > 1 && CudaNdarray_HOST_STRIDES(%(var)s)[0] != 1
&& CudaNdarray_HOST_DIMS(%(var)s)[1] > 1 && CudaNdarray_HOST_STRIDES(%(var)s)[1] != 1
&& CudaNdarray_HOST_DIMS(%(var)s)[2] > 1 && CudaNdarray_HOST_STRIDES(%(var)s)[2] != 1)
......@@ -198,8 +194,7 @@ class GpuBatchedDot(GpuOp):
Py_XDECREF(%(var)s);
%(var)s = _copy;
}
""" % dict(var=var, fail=fail) for var in (bx, by)))
+ """
""" % dict(var=var, fail=fail) for var in (bx, by))) + """
// fail if the output is not contiguous; we can't copy it because we
// need to write to the original memory
......@@ -363,7 +358,7 @@ class GpuDot22(GpuOp):
if y.type.ndim != 2:
raise TypeError(y)
otype = CudaNdarrayType(
(x.type.broadcastable[0], y.type.broadcastable[1]))
(x.type.broadcastable[0], y.type.broadcastable[1]))
return Apply(self, [x, y], [otype()])
def c_code_cache_version(self):
......@@ -451,7 +446,7 @@ class GpuDot22Scalar(GpuOp):
if not tensor.blas._as_scalar(a):
raise TypeError(a)
otype = CudaNdarrayType(
(x.type.broadcastable[0], y.type.broadcastable[1]))
(x.type.broadcastable[0], y.type.broadcastable[1]))
return Apply(self, [x, y, a], [otype()])
def c_code_cache_version(self):
......@@ -537,8 +532,8 @@ class GpuGemm(GpuOp):
return 'GpuGemm{no_inplace}'
def __eq__(self, other):
return (type(self) == type(other)\
and self.inplace == other.inplace)
return (type(self) == type(other) and
self.inplace == other.inplace)
def __hash__(self):
return hash(type(self)) ^ hash(self.inplace)
......@@ -562,7 +557,7 @@ class GpuGemm(GpuOp):
return (4,)
def c_code(self, node, name, inputs, outputs, sub):
#z_out = alpha * dot(x,y) + beta * z_in
# z_out = alpha * dot(x,y) + beta * z_in
# inplace version, set set z_out = z_in
# not inplace version, we copy z_in to z_out.
z_in, a, x, y, b = inputs
......@@ -657,8 +652,8 @@ class GpuGemv(GpuOp):
return 'GpuGemv{no_inplace}'
def __eq__(self, other):
return (type(self) == type(other)\
and self.inplace == other.inplace)
return (type(self) == type(other) and
self.inplace == other.inplace)
def __hash__(self):
return hash(type(self)) ^ hash(self.inplace)
......@@ -682,7 +677,7 @@ class GpuGemv(GpuOp):
return (3,)
def c_code(self, node, name, inputs, outputs, sub):
#z_out = alpha * dot(x,y) + beta * z_in
# z_out = alpha * dot(x,y) + beta * z_in
# inplace version, set set z_out = z_in
# not inplace version, we copy z_in to z_out.
z_in, a, x, y, b = inputs
......@@ -757,8 +752,8 @@ class GpuGer(GpuOp):
return 'GpuGer{no_inplace}'
def __eq__(self, other):
return (type(self) == type(other)\
and self.inplace == other.inplace)
return (type(self) == type(other) and
self.inplace == other.inplace)
def __hash__(self):
return hash(type(self)) ^ hash(self.inplace)
......@@ -782,7 +777,7 @@ class GpuGer(GpuOp):
return (2,)
def c_code(self, node, name, inputs, outputs, sub):
#z_out = alpha * dot(x,y) + beta * z_in
# z_out = alpha * dot(x,y) + beta * z_in
# inplace version, set set z_out = z_in
# not inplace version, we copy z_in to z_out.
z_in, a, x, y = inputs
......@@ -934,7 +929,7 @@ class BaseGpuCorrMM(GpuOp):
# these files
files = ['corr_gemm.cu']
codes = [open(os.path.join(os.path.split(__file__)[0], f)).read()
for f in files]
for f in files]
return reduce(str.__add__, codes)
def c_code_helper(self, bottom, weights, top, direction, sub, height=None, width=None):
......@@ -947,7 +942,7 @@ class BaseGpuCorrMM(GpuOp):
Parameters
----------
bottom
bottom
Variable name of the input images in the forward pass,
or the gradient of the input images in backprop wrt. inputs
weights
......@@ -1001,7 +996,7 @@ class BaseGpuCorrMM(GpuOp):
out = bottom
else:
raise ValueError("direction must be one of 'forward', "
"'backprop weights', 'backprop inputs'")
"'backprop weights', 'backprop inputs'")
# When subsampling, we cannot unambiguously infer the height and width
# of bottom and weights from top, so we require them to be given.
# Similarly, when pad="half", we cannot infer the weight size.
......@@ -1158,7 +1153,7 @@ class GpuCorrMM(BaseGpuCorrMM):
Parameters
----------
border_mode
The width of a border of implicit zeros to pad the
The width of a border of implicit zeros to pad the
input with. Must be a tuple with 2 elements giving the numbers of rows
and columns to pad on each side, or a single integer to pad the same
on all sides, or a string shortcut setting the padding at runtime:
......@@ -1174,7 +1169,7 @@ class GpuCorrMM(BaseGpuCorrMM):
but faster.
Set to `(1, 1)` to disable subsampling.
pad
Deprecated alias for `border_mode`.
Deprecated alias for `border_mode`.
Notes
-----
......@@ -1247,8 +1242,8 @@ class GpuCorrMM_gradWeights(BaseGpuCorrMM):
"""
def __init__(self, border_mode="valid",
subsample=(1, 1),
pad=(0, 0)):
subsample=(1, 1),
pad=(0, 0)):
super(GpuCorrMM_gradWeights, self).__init__(border_mode, subsample, pad)
def make_node(self, img, topgrad, shape=None):
......@@ -1283,11 +1278,15 @@ class GpuCorrMM_gradWeights(BaseGpuCorrMM):
bottom, top = inp[:2]
weights, = grads
weights = gpu_contiguous(weights)
d_bottom = GpuCorrMM_gradInputs(self.border_mode, self.subsample)(
weights, top, bottom.shape[-2:])
d_top = GpuCorrMM(self.border_mode, self.subsample)(
bottom, weights)
d_height_width = (theano.gradient.DisconnectedType()(),) * 2 if len(inp) == 4 else ()
d_bottom = GpuCorrMM_gradInputs(
self.border_mode, self.subsample)(weights,
top,
bottom.shape[-2:])
d_top = GpuCorrMM(
self.border_mode, self.subsample)(bottom, weights)
d_height_width = (
theano.gradient.DisconnectedType()(),
) * 2 if len(inp) == 4 else ()
return (d_bottom, d_top) + d_height_width
def connection_pattern(self, node):
......@@ -1309,8 +1308,8 @@ class GpuCorrMM_gradInputs(BaseGpuCorrMM):
"""
def __init__(self, border_mode="valid",
subsample=(1, 1),
pad=(0, 0)):
subsample=(1, 1),
pad=(0, 0)):
super(GpuCorrMM_gradInputs, self).__init__(border_mode, subsample, pad)
def make_node(self, kern, topgrad, shape=None):
......@@ -1342,11 +1341,14 @@ class GpuCorrMM_gradInputs(BaseGpuCorrMM):
weights, top = inp[:2]
bottom, = grads
bottom = gpu_contiguous(bottom)
d_weights = GpuCorrMM_gradWeights(self.border_mode, self.subsample)(
d_weights = GpuCorrMM_gradWeights(
self.border_mode, self.subsample)(
bottom, top, weights.shape[-2:])
d_top = GpuCorrMM(self.border_mode, self.subsample)(
bottom, weights)
d_height_width = (theano.gradient.DisconnectedType()(),) * 2 if len(inp) == 4 else ()
d_top = GpuCorrMM(
self.border_mode, self.subsample)(bottom, weights)
d_height_width = (
theano.gradient.DisconnectedType()(),
) * 2 if len(inp) == 4 else ()
return (d_weights, d_top) + d_height_width
def connection_pattern(self, node):
......@@ -1412,7 +1414,7 @@ class BaseGpuCorr3dMM(GpuOp):
# these files
files = ['corr3d_gemm.cu']
codes = [open(os.path.join(os.path.split(__file__)[0], f)).read()
for f in files]
for f in files]
return reduce(str.__add__, codes)
def c_code_helper(self, bottom, weights,
......@@ -1459,7 +1461,7 @@ class BaseGpuCorr3dMM(GpuOp):
If self.pad == 'half', a variable giving the width of the filters
for direction="backprop weights".
Ignored otherwise.
depth
depth
If self.subsample[2] != 1, a variable giving the depth
of the filters for direction="backprop weights" or the depth of the
input images for direction="backprop inputs".
......@@ -1488,7 +1490,7 @@ class BaseGpuCorr3dMM(GpuOp):
out = bottom
else:
raise ValueError("direction must be one of 'forward', "
"'backprop weights', 'backprop inputs'")
"'backprop weights', 'backprop inputs'")
# When subsampling, we cannot unambiguously infer the height and width
# of bottom and weights from top, so we require them to be given.
# Similarly, when pad="half", we cannot infer the weight size.
......@@ -1755,10 +1757,16 @@ class GpuCorr3dMM(BaseGpuCorr3dMM):
bottom, weights = inp
top, = grads
top = gpu_contiguous(top)
d_bottom = GpuCorr3dMM_gradInputs(self.border_mode, self.subsample, self.pad)(
weights, top, bottom.shape[-3:])
d_weights = GpuCorr3dMM_gradWeights(self.border_mode, self.subsample, self.pad)(
bottom, top, weights.shape[-3:])
d_bottom = GpuCorr3dMM_gradInputs(self.border_mode,
self.subsample,
self.pad)(weights,
top,
bottom.shape[-3:])
d_weights = GpuCorr3dMM_gradWeights(self.border_mode,
self.subsample,
self.pad)(bottom,
top,
weights.shape[-3:])
return d_bottom, d_weights
......@@ -1863,11 +1871,14 @@ class GpuCorr3dMM_gradInputs(BaseGpuCorr3dMM):
weights, top = inp[:2]
bottom, = grads
bottom = gpu_contiguous(bottom)
d_weights = GpuCorr3dMM_gradWeights(self.border_mode, self.subsample, self.pad)(
bottom, top, weights.shape[-3:])
d_top = GpuCorr3dMM(self.border_mode, self.subsample, self.pad)(
d_weights = GpuCorr3dMM_gradWeights(
self.border_mode, self.subsample, self.pad)(
bottom, top, weights.shape[-3:])
d_top = GpuCorr3dMM(
self.border_mode, self.subsample, self.pad)(
bottom, weights)
d_height_width_depth = (theano.gradient.DisconnectedType()(),) * 3 if len(inp) == 5 else ()
d_height_width_depth = (theano.gradient.DisconnectedType()(),)\
* 3 if len(inp) == 5 else ()
return (d_weights, d_top) + d_height_width_depth
def connection_pattern(self, node):
......@@ -1938,19 +1949,19 @@ class GpuConv(GpuOp):
raise ValueError(mode)
def __init__(self, border_mode,
subsample=(1, 1),
logical_img_hw=None,
logical_kern_hw=None,
logical_kern_align_top=True,
version=-1,
direction_hint=None,
verbose=0,
kshp=None,
imshp=None,
max_threads_dim0=None,
nkern=None,
bsize=None,
fft_opt=True):
subsample=(1, 1),
logical_img_hw=None,
logical_kern_hw=None,
logical_kern_align_top=True,
version=-1,
direction_hint=None,
verbose=0,
kshp=None,
imshp=None,
max_threads_dim0=None,
nkern=None,
bsize=None,
fft_opt=True):
self.border_mode = border_mode
if version != -1:
raise Exception(
......@@ -2107,7 +2118,7 @@ class GpuConv(GpuOp):
# these files
files = ['conv_kernel.cu', 'conv_full_kernel.cu', 'conv.cu']
codes = [open(os.path.join(os.path.split(__file__)[0], f)).read()
for f in files]
for f in files]
return reduce(str.__add__, codes)
def c_code(self, node, nodename, inp, out_, sub):
......@@ -2186,7 +2197,7 @@ class GpuDownsampleFactorMax(GpuOp):
return Apply(self, [x], [x.type()])
# def perform(self, node, input_storage, output_storage):
#raise NotImplementedError('only C is implemented')
# raise NotImplementedError('only C is implemented')
def c_code_cache_version(self):
return (6)
......
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