Skip to content

P
pytensor
提交 d36faf23 authored 作者: --global's avatar --global
浏览文件
操作

Flake8 on dnn.py

上级 d4d3ae5b
隐藏空白字符变更
内嵌 并排
正在显示 包含 52 行增加 和 43 行删除
theano/sandbox/cuda/dnn.py
浏览文件 @ d36faf23
......@@ -3,7 +3,7 @@ import numpy
import warnings
import theano
from theano import Apply, gof, tensor, config, Variable
from theano import Apply, tensor, config, Variable
from theano.scalar import as_scalar, constant, Log
from theano.gradient import DisconnectedType, grad_not_implemented
from theano.gof import Optimizer, local_optimizer, COp
......@@ -129,6 +129,7 @@ if (%(err)s != CUDNN_STATUS_SUCCESS) {
""" % dict(var=var, err=err, desc=desc, fail=fail)
class DnnBase(GpuOp, COp):
"""
Creates a handle for cudnn and pulls in the cudnn libraries and headers.
......@@ -385,6 +386,7 @@ def ensure_float(val, default, name):
raise TypeError("%s: type is not float32" % (name,))
return val
def ensure_int(val, default, name):
if val is None:
return default.clone()
......@@ -424,7 +426,7 @@ class GpuDnnConv(DnnBase, COp):
if workmem is not None:
warnings.warn(("GpuDnnConv: parameter 'workmem' is deprecated. "
"Use 'algo' instead."), stacklevel=3)
assert algo == None
assert algo is None
self.algo = workmem
else:
if algo is None:
......@@ -446,7 +448,8 @@ class GpuDnnConv(DnnBase, COp):
"implementation based on heuristics "
"requires CuDNN v3")
elif self.algo in ['time_once', 'time_on_shape_change']:
raise RuntimeError("CuDNN convolution timing requires CuDNN v3")
raise RuntimeError("CuDNN convolution timing requires CuDNN "
"v3")
assert self.algo in ['none', 'small', 'large', 'fft', 'guess_once',
'guess_on_shape_change', 'time_once',
......@@ -533,8 +536,10 @@ class GpuDnnConv(DnnBase, COp):
top = gpu_contiguous(top)
d_img = GpuDnnConvGradI()(kerns, top, gpu_alloc_empty(*img.shape), desc)
d_kerns = GpuDnnConvGradW()(img, top, gpu_alloc_empty(*kerns.shape), desc)
d_img = GpuDnnConvGradI()(kerns, top, gpu_alloc_empty(*img.shape),
desc)
d_kerns = GpuDnnConvGradW()(img, top, gpu_alloc_empty(*kerns.shape),
desc)
d_alpha = grad_not_implemented(self, 4, alpha)
d_beta = grad_not_implemented(self, 5, beta)
......@@ -580,7 +585,6 @@ class GpuDnnConv(DnnBase, COp):
return [shape[2]]
class GpuDnnConv3d(GpuDnnConv):
"""
The forward convolution.
......@@ -603,7 +607,7 @@ class GpuDnnConv3d(GpuDnnConv):
if workmem is not None:
warnings.warn(("GpuDnnConv3d: parameter 'workmem' is deprecated. "
"Use 'algo' instead."), stacklevel=3)
assert algo == None
assert algo is None
algo = workmem
super(GpuDnnConv3d, self).__init__(inplace=inplace, algo='none')
......@@ -636,8 +640,10 @@ class GpuDnnConv3d(GpuDnnConv):
top = gpu_contiguous(top)
d_img = GpuDnnConv3dGradI()(kerns, top, gpu_alloc_empty(*img.shape), desc)
d_kerns = GpuDnnConv3dGradW()(img, top, gpu_alloc_empty(*kerns.shape), desc)
d_img = GpuDnnConv3dGradI()(kerns, top, gpu_alloc_empty(*img.shape),
desc)
d_kerns = GpuDnnConv3dGradW()(img, top, gpu_alloc_empty(*kerns.shape),
desc)
d_alpha = grad_not_implemented(self, 4, alpha)
d_beta = grad_not_implemented(self, 5, beta)
......@@ -706,7 +712,7 @@ class GpuDnnConvGradW(DnnBase, COp):
if workmem is not None:
warnings.warn(("GpuDnnConvGradW: parameter 'workmem' is "
"deprecated. Use 'algo' instead."), stacklevel=3)
assert algo == None
assert algo is None
self.algo = workmem
else:
if algo is None:
......@@ -736,7 +742,8 @@ class GpuDnnConvGradW(DnnBase, COp):
kerns = gpu_contiguous(kerns)
d_img = GpuDnnConvGradI()(kerns, top, gpu_alloc_empty(*img.shape), desc)
d_img = GpuDnnConvGradI()(kerns, top, gpu_alloc_empty(*img.shape),
desc)
d_top = GpuDnnConv()(img, kerns, gpu_alloc_empty(*top.shape), desc)
d_alpha = grad_not_implemented(self, 4, alpha)
d_beta = grad_not_implemented(self, 5, beta)
......@@ -838,7 +845,7 @@ class GpuDnnConv3dGradW(GpuDnnConvGradW):
if workmem is not None:
warnings.warn(("GpuDnnConv3dGradW: parameter 'workmem' is "
"deprecated. Use 'algo' instead."), stacklevel=3)
assert algo == None
assert algo is None
algo = workmem
super(GpuDnnConv3dGradW, self).__init__(inplace=inplace,
......@@ -852,7 +859,8 @@ class GpuDnnConv3dGradW(GpuDnnConvGradW):
kerns = gpu_contiguous(kerns)
d_img = GpuDnnConv3dGradI()(kerns, top, gpu_alloc_empty(*img.shape), desc)
d_img = GpuDnnConv3dGradI()(kerns, top, gpu_alloc_empty(*img.shape),
desc)
d_top = GpuDnnConv3d()(img, kerns, gpu_alloc_empty(*top.shape), desc)
d_alpha = grad_not_implemented(self, 4, alpha)
d_beta = grad_not_implemented(self, 5, beta)
......@@ -871,7 +879,6 @@ class GpuDnnConv3dGradW(GpuDnnConvGradW):
if output.type.ndim != 5:
raise TypeError('output must be 5D tensor')
if not isinstance(desc.type, CDataType) \
or desc.type.ctype != 'cudnnConvolutionDescriptor_t':
raise TypeError('desc must be cudnnConvolutionDescriptor_t')
......@@ -893,7 +900,8 @@ class GpuDnnConvGradI(DnnBase, COp):
"""
__props__ = ('algo', 'inplace',)
__input_name__ = ('kernel', 'grad', 'output', 'descriptor', 'alpha', 'beta')
__input_name__ = ('kernel', 'grad', 'output', 'descriptor', 'alpha',
'beta')
def __init__(self, inplace=False, workmem=None, algo=None):
"""
......@@ -908,7 +916,7 @@ class GpuDnnConvGradI(DnnBase, COp):
if workmem is not None:
warnings.warn(("GpuDnnConvGradI: parameter 'workmem' is "
"deprecated. Use 'algo' instead."), stacklevel=3)
assert algo == None
assert algo is None
self.algo = workmem
else:
if algo is None:
......@@ -938,7 +946,8 @@ class GpuDnnConvGradI(DnnBase, COp):
img = gpu_contiguous(img)
d_kerns = GpuDnnConvGradW()(img, top, gpu_alloc_empty(*kerns.shape), desc)
d_kerns = GpuDnnConvGradW()(img, top, gpu_alloc_empty(*kerns.shape),
desc)
d_top = GpuDnnConv()(img, kerns, gpu_alloc_empty(*top.shape), desc)
d_alpha = grad_not_implemented(self, 4, alpha)
d_beta = grad_not_implemented(self, 5, beta)
......@@ -1018,7 +1027,6 @@ class GpuDnnConvGradI(DnnBase, COp):
return [shape[2]]
class GpuDnnConv3dGradI(GpuDnnConvGradI):
"""
The convolution gradient with respect to the inputs.
......@@ -1029,7 +1037,8 @@ class GpuDnnConv3dGradI(GpuDnnConvGradI):
"""
__props__ = ('algo', 'inplace',)
__input_name__ = ('kernel', 'grad', 'output', 'descriptor', 'alpha', 'beta')
__input_name__ = ('kernel', 'grad', 'output', 'descriptor', 'alpha',
'beta')
def __init__(self, inplace=False, workmem=None, algo=None):
"""
......@@ -1041,7 +1050,7 @@ class GpuDnnConv3dGradI(GpuDnnConvGradI):
if workmem is not None:
warnings.warn(("GpuDnnConv3dGradI: parameter 'workmem' is "
"deprecated. Use 'algo' instead."), stacklevel=3)
assert algo == None
assert algo is None
algo = workmem
super(GpuDnnConv3dGradI, self).__init__(inplace=inplace,
......@@ -1049,14 +1058,14 @@ class GpuDnnConv3dGradI(GpuDnnConvGradI):
assert self.algo in ['none', 'guess_once', 'guess_on_shape_change',
'time_once', 'time_on_shape_change']
def grad(self, inp, grads):
kerns, top, output, desc, alpha, beta = inp
img, = grads
img = gpu_contiguous(img)
d_kerns = GpuDnnConv3dGradW()(img, top, gpu_alloc_empty(*kerns.shape), desc)
d_kerns = GpuDnnConv3dGradW()(img, top, gpu_alloc_empty(*kerns.shape),
desc)
d_top = GpuDnnConv3d()(img, kerns, gpu_alloc_empty(*top.shape), desc)
d_alpha = grad_not_implemented(self, 4, alpha)
d_beta = grad_not_implemented(self, 5, beta)
......@@ -1086,7 +1095,6 @@ class GpuDnnConv3dGradI(GpuDnnConvGradI):
[output.type()])
def dnn_conv(img, kerns, border_mode='valid', subsample=(1, 1),
conv_mode='conv', direction_hint=None, workmem=None, algo=None):
"""
......@@ -1126,7 +1134,7 @@ def dnn_conv(img, kerns, border_mode='valid', subsample=(1, 1),
if workmem is not None:
warnings.warn(("dnn_conv: parameter 'workmem' is deprecated. Use "
"'algo' instead."), stacklevel=3)
assert algo == None
assert algo is None
algo = workmem
# Ensure the value of direction_hint is supported
......@@ -1134,7 +1142,7 @@ def dnn_conv(img, kerns, border_mode='valid', subsample=(1, 1),
fgraph = getattr(img, 'fgraph', None) or getattr(kerns, 'fgraph', None)
if (border_mode == 'valid' and subsample == (1, 1) and
direction_hint == 'bprop weights'):
direction_hint == 'bprop weights'):
# Special case: We are asked to use GpuDnnConvGradW. We need to set
# up a suitable 'fake' convolution to compute the gradient for.
img = gpu_contiguous(img.dimshuffle(1, 0, 2, 3))
......@@ -1146,14 +1154,14 @@ def dnn_conv(img, kerns, border_mode='valid', subsample=(1, 1),
shape2 = shape_i(img, 2, fgraph) - shape_i(kerns, 2, fgraph) + 1
shape3 = shape_i(img, 3, fgraph) - shape_i(kerns, 3, fgraph) + 1
out = gpu_alloc_empty(shape_i(kerns, 1, fgraph),
shape_i(img, 1, fgraph), shape2, shape3)
shape_i(img, 1, fgraph), shape2, shape3)
desc = GpuDnnConvDesc(border_mode='valid', subsample=(1, 1),
conv_mode='cross')(img.shape, out.shape)
conv = GpuDnnConvGradW()(img, kerns, out, desc)
return as_cuda_ndarray_variable(conv.dimshuffle(1, 0, 2, 3))
elif (border_mode == 'full' and subsample == (1, 1) and
direction_hint != 'forward!' and version() == -1):
direction_hint != 'forward!' and version() == -1):
# Special case: In CuDNN v1, we can be faster by using GpuDnnConvGradI
# to compute the full convolution as the backward pass of a valid
# convolution. We just need to set up a suitable 'fake' valid
......@@ -1164,7 +1172,7 @@ def dnn_conv(img, kerns, border_mode='valid', subsample=(1, 1),
shape2 = shape_i(img, 2, fgraph) + shape_i(kerns, 2, fgraph) - 1
shape3 = shape_i(img, 3, fgraph) + shape_i(kerns, 3, fgraph) - 1
out = gpu_alloc_empty(shape_i(img, 0, fgraph),
shape_i(kerns, 1, fgraph), shape2, shape3)
shape_i(kerns, 1, fgraph), shape2, shape3)
desc = GpuDnnConvDesc(border_mode='valid', subsample=(1, 1),
conv_mode=conv_mode)(out.shape, kerns.shape)
return GpuDnnConvGradI()(kerns, img, out, desc)
......@@ -1198,8 +1206,8 @@ def dnn_conv3d(img, kerns, border_mode='valid', subsample=(1, 1, 1),
:param border_mode: one of 'valid', 'full'; additionally, the padding size
could be directly specified by an integer or a pair of integers
:param subsample: perform subsampling of the output (default: (1, 1, 1))
:param conv_mode: perform convolution (kernels flipped) or cross-correlation.
One of 'conv', 'cross'. (default: 'conv')
:param conv_mode: perform convolution (kernels flipped) or
cross-correlation. One of 'conv', 'cross'. (default: 'conv')
:param direction_hint: Used by graph optimizers to change algorithm choice.
By default, GpuDnnConv will be used to carry out the convolution.
If border_mode is 'valid', subsample is (1,1,1) and direction_hint is
......@@ -1222,7 +1230,7 @@ def dnn_conv3d(img, kerns, border_mode='valid', subsample=(1, 1, 1),
if workmem is not None:
warnings.warn(("dnn_conv3d: parameter 'workmem' is deprecated. Use "
"'algo' instead."), stacklevel=3)
assert algo == None
assert algo is None
algo = workmem
# Ensure the value of direction_hint is supported
......@@ -1230,7 +1238,7 @@ def dnn_conv3d(img, kerns, border_mode='valid', subsample=(1, 1, 1),
fgraph = getattr(img, 'fgraph', None) or getattr(kerns, 'fgraph', None)
if (border_mode == 'valid' and subsample == (1, 1, 1) and
direction_hint == 'bprop weights'):
direction_hint == 'bprop weights'):
# Special case: We are asked to use GpuDnnConvGradW. We need to set
# up a suitable 'fake' convolution to compute the gradient for.
img = gpu_contiguous(img.dimshuffle(1, 0, 2, 3, 4))
......@@ -1243,7 +1251,7 @@ def dnn_conv3d(img, kerns, border_mode='valid', subsample=(1, 1, 1),
shape3 = shape_i(img, 3, fgraph) - shape_i(kerns, 3, fgraph) + 1
shape4 = shape_i(img, 4, fgraph) - shape_i(kerns, 4, fgraph) + 1
out = gpu_alloc_empty(shape_i(kerns, 1, fgraph),
shape_i(img, 1, fgraph), shape2, shape3, shape3)
shape_i(img, 1, fgraph), shape2, shape3, shape4)
desc = GpuDnnConvDesc(border_mode='valid', subsample=(1, 1, 1),
conv_mode='cross')(img.shape, out.shape)
conv = GpuDnnConv3dGradW()(img, kerns, out, desc)
......@@ -1258,8 +1266,8 @@ def dnn_conv3d(img, kerns, border_mode='valid', subsample=(1, 1, 1),
conv_mode=conv_mode)(img.shape, kerns.shape)
desc_op = desc.owner.op
out_shp = GpuDnnConv3d.get_out_shape(img.shape, kerns.shape,
desc_op.border_mode,
desc_op.subsample)
desc_op.border_mode,
desc_op.subsample)
out = gpu_alloc_empty(*out_shp)
return GpuDnnConv3d(algo=algo)(img, kerns, out, desc)
......@@ -1536,7 +1544,8 @@ class GpuDnnPoolGrad(DnnBase):
:param inp: the input of the pooling.
:param out: the output of the pooling in the forward.
:param inp_grad: same size as out, but is the corresponding gradient information.
:param inp_grad: same size as out, but is the corresponding gradient
information.
:param desc: The pooling descriptor.
"""
__props__ = ()
......@@ -2226,14 +2235,14 @@ if True:
@register_opt('cudnn')
@local_optimizer([GpuElemwise])
def local_log_softmax_dnn(node):
# The log-softmax implementation is only available starting at CuDNN V3.
# The log-softmax implementation is only available starting at CuDNN V3
if not dnn_available() or version() < (3000, 3000):
return
if (isinstance(node.op, GpuElemwise) and
isinstance(node.op.scalar_op, Log) and
node.inputs[0].owner and
isinstance(node.inputs[0].owner.op, GpuDnnSoftmax) and
len(node.inputs[0].owner.out.clients) == 1):
isinstance(node.op.scalar_op, Log) and
node.inputs[0].owner and
isinstance(node.inputs[0].owner.op, GpuDnnSoftmax) and
len(node.inputs[0].owner.out.clients) == 1):
log_input = node.inputs[0]
softmax_node = log_input.owner
......@@ -2260,8 +2269,8 @@ if True:
def local_softmax_dnn_grad(node):
if (isinstance(node.op, SoftmaxGrad) and
((node.inputs[0].owner and
isinstance(node.inputs[0].owner.op, HostFromGpu))
or (node.inputs[1].owner and
isinstance(node.inputs[0].owner.op, HostFromGpu)) or
(node.inputs[1].owner and
isinstance(node.inputs[1].owner.op, HostFromGpu)))):
if not dnn_available():
return
......
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