Skip to content
项目
群组
代码片段
帮助
当前项目
正在载入...
登录 / 注册
切换导航面板
P
pytensor
项目
项目
详情
活动
周期分析
仓库
仓库
文件
提交
分支
标签
贡献者
图表
比较
统计图
议题
0
议题
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
CI / CD
CI / CD
流水线
作业
日程
统计图
Wiki
Wiki
代码片段
代码片段
成员
成员
折叠边栏
关闭边栏
活动
图像
聊天
创建新问题
作业
提交
问题看板
Open sidebar
testgroup
pytensor
Commits
627bd58e
提交
627bd58e
authored
5月 19, 2011
作者:
Frederic Bastien
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
Removed old function CudaNdarray_new_null() as it is deprecated. Now use CudaNdarray_New()
上级
b5bdebb8
隐藏空白字符变更
内嵌
并排
正在显示
6 个修改的文件
包含
32 行增加
和
43 行删除
+32
-43
basic_ops.py
theano/sandbox/cuda/basic_ops.py
+2
-2
blas.py
theano/sandbox/cuda/blas.py
+8
-8
cuda_ndarray.cu
theano/sandbox/cuda/cuda_ndarray.cu
+12
-18
cuda_ndarray.cuh
theano/sandbox/cuda/cuda_ndarray.cuh
+2
-7
elemwise.py
theano/sandbox/cuda/elemwise.py
+2
-2
nnet.py
theano/sandbox/cuda/nnet.py
+6
-6
没有找到文件。
theano/sandbox/cuda/basic_ops.py
浏览文件 @
627bd58e
...
...
@@ -1934,7 +1934,7 @@ class GpuAlloc(Op):
str
+=
"||CudaNdarray_HOST_DIMS(
%(out)
s)[
%(idx)
s]!=dims[
%(idx)
s]"
%
locals
()
str
+=
"""){
Py_XDECREF(
%(out)
s);
%(out)
s= (CudaNdarray*)CudaNdarray_
new_null
();
%(out)
s= (CudaNdarray*)CudaNdarray_
New
();
CudaNdarray_alloc_contiguous(
%(out)
s,
%(nd)
s, dims);
}
if (CudaNdarray_CopyFromCudaNdarray(
%(out)
s,
%(value)
s, true))
...
...
@@ -1952,7 +1952,7 @@ class GpuAlloc(Op):
return
[
None
for
i
in
inputs
]
def
c_code_cache_version
(
self
):
return
(
2
,)
return
(
3
,)
gpu_alloc
=
GpuAlloc
()
...
...
theano/sandbox/cuda/blas.py
浏览文件 @
627bd58e
...
...
@@ -22,7 +22,7 @@ class GpuDot22(Op):
return
Apply
(
self
,
[
x
,
y
],
[
x
.
type
()])
def
c_code_cache_version
(
self
):
return
(
1
,
0
)
return
(
1
,
1
)
def
c_code
(
self
,
node
,
nodename
,
inputs
,
outputs
,
sub
):
x
,
y
=
inputs
...
...
@@ -48,7 +48,7 @@ class GpuDot22(Op):
npy_intp dims[2];
dims[0] = CudaNdarray_HOST_DIMS(
%(x)
s)[0];
dims[1] = CudaNdarray_HOST_DIMS(
%(y)
s)[1];
%(z)
s = (CudaNdarray*)CudaNdarray_
new_null
();
%(z)
s = (CudaNdarray*)CudaNdarray_
New
();
if ((NULL ==
%(z)
s) || CudaNdarray_alloc_contiguous(
%(z)
s, 2, dims))
{
if (
%(z)
s)
...
...
@@ -90,7 +90,7 @@ class GpuDot22Scalar(Op):
return
Apply
(
self
,
[
x
,
y
,
a
],
[
x
.
type
()])
def
c_code_cache_version
(
self
):
return
(
1
,
0
)
return
(
1
,
1
)
def
c_code
(
self
,
node
,
name
,
inputs
,
outputs
,
sub
):
x
,
y
,
a
=
inputs
...
...
@@ -122,7 +122,7 @@ class GpuDot22Scalar(Op):
npy_intp dims[2];
dims[0] = CudaNdarray_HOST_DIMS(
%(x)
s)[0];
dims[1] = CudaNdarray_HOST_DIMS(
%(y)
s)[1];
%(z)
s = (CudaNdarray*)CudaNdarray_
new_null
();
%(z)
s = (CudaNdarray*)CudaNdarray_
New
();
if ((NULL ==
%(z)
s) || CudaNdarray_alloc_contiguous(
%(z)
s, 2, dims))
{
if (
%(z)
s)
...
...
@@ -436,7 +436,7 @@ class GpuDownsampleFactorMax(Op):
#def perform(self, node, input_storage, output_storage):
#raise NotImplementedError('only C is implemented')
def
c_code_cache_version
(
self
):
return
(
2
)
return
(
3
)
def
c_code
(
self
,
node
,
nodename
,
inp
,
out
,
sub
):
x
,
=
inp
z
,
=
out
...
...
@@ -473,7 +473,7 @@ class GpuDownsampleFactorMax(Op):
|| (CudaNdarray_HOST_DIMS(
%(z)
s)[3] != dims[3]))
{
Py_XDECREF(
%(z)
s);
%(z)
s = (CudaNdarray*)CudaNdarray_
new_null
();
%(z)
s = (CudaNdarray*)CudaNdarray_
New
();
if ((NULL ==
%(z)
s)
|| CudaNdarray_alloc_contiguous(
%(z)
s, 4, dims))
{
...
...
@@ -588,7 +588,7 @@ class GpuDownsampleFactorMaxGrad(Op):
return
Apply
(
self
,
[
x
,
z
,
gz
],
[
x
.
type
()])
def
c_code_cache_version
(
self
):
#return ()
return
(
4
,)
return
(
5
,)
def
c_code
(
self
,
node
,
nodename
,
inp
,
out
,
sub
):
x
,
z
,
gz
=
inp
...
...
@@ -611,7 +611,7 @@ class GpuDownsampleFactorMaxGrad(Op):
|| (CudaNdarray_HOST_DIMS(
%(gx)
s)[3] != CudaNdarray_HOST_DIMS(
%(x)
s)[3]))
{
Py_XDECREF(
%(gx)
s);
%(gx)
s = (CudaNdarray*)CudaNdarray_
new_null
();
%(gx)
s = (CudaNdarray*)CudaNdarray_
New
();
if ((NULL ==
%(gx)
s)
|| CudaNdarray_alloc_contiguous(
%(gx)
s, 4, CudaNdarray_HOST_DIMS(
%(x)
s)))
{
...
...
theano/sandbox/cuda/cuda_ndarray.cu
浏览文件 @
627bd58e
...
...
@@ -350,7 +350,7 @@ PyObject* CudaNdarray_ZEROS(int n, int * dims)
// total_elements now contains the size of the array, in reals
int total_size = total_elements * sizeof(real);
CudaNdarray* rval = (CudaNdarray*)CudaNdarray_
new_null
();
CudaNdarray* rval = (CudaNdarray*)CudaNdarray_
New
();
if (!rval)
{
PyErr_SetString(PyExc_RuntimeError, "CudaNdarray_ZEROS: call to new_null failed");
...
...
@@ -448,7 +448,7 @@ PyObject* CudaNdarray_Zeros(PyObject* dummy, PyObject* shape)
PyObject * CudaNdarray_Copy(CudaNdarray * self)
{
PyObject * rval = CudaNdarray_
new_null
();
PyObject * rval = CudaNdarray_
New
();
if ((!rval) || (-1 == self->nd))
{
return rval;
...
...
@@ -509,7 +509,7 @@ PyObject * CudaNdarray_ReduceSum(CudaNdarray * self, PyObject * py_reduce_mask)
PyErr_SetString(PyExc_TypeError, "length of reduce_mask must match self->nd");
return NULL;
}
CudaNdarray * self_sum = (CudaNdarray*)CudaNdarray_
new_null
();
CudaNdarray * self_sum = (CudaNdarray*)CudaNdarray_
New
();
if (!self_sum)
{
return NULL;
...
...
@@ -666,9 +666,8 @@ PyObject * CudaNdarray_Reshape(CudaNdarray * self, PyObject * shape)
}
// allocate new space (TODO: test to see if we can re-use old one)
CudaNdarray * rval = (CudaNdarray * )CudaNdarray_new_null();
if (!rval || CudaNdarray_alloc_contiguous(rval, rval_nd, rval_dims))
{
CudaNdarray * rval = (CudaNdarray * )CudaNdarray_New();
if (!rval || CudaNdarray_alloc_contiguous(rval, rval_nd, rval_dims)){
Py_XDECREF(rval);
free(rval_dims);
return NULL;
...
...
@@ -754,7 +753,7 @@ PyObject * CudaNdarray_SetShapeI(CudaNdarray * self, PyObject *args)
static PyObject *
CudaNdarray_exp(CudaNdarray* self)
{
CudaNdarray * rval = (CudaNdarray *)CudaNdarray_
new_null
();
CudaNdarray * rval = (CudaNdarray *)CudaNdarray_
New
();
if ((NULL == rval) || CudaNdarray_alloc_contiguous(rval, self->nd, CudaNdarray_HOST_DIMS(self)))
{
Py_XDECREF(rval);
...
...
@@ -872,7 +871,7 @@ CudaNdarray_add(PyObject* py_self, PyObject * py_other)
}
size *= (unsigned int) CudaNdarray_HOST_DIMS(self)[i];
}
CudaNdarray * rval = (CudaNdarray *)CudaNdarray_
new_null
();
CudaNdarray * rval = (CudaNdarray *)CudaNdarray_
New
();
if (!rval || CudaNdarray_alloc_contiguous(rval, self->nd, CudaNdarray_HOST_DIMS(self)))
{
Py_XDECREF(rval);
...
...
@@ -2061,7 +2060,7 @@ CudaNdarray_from_gpu_pointer(PyObject* _unused, PyObject* args)
return NULL;
}
rval = CudaNdarray_
new_null
();
rval = CudaNdarray_
New
();
if (CudaNdarray_set_nd((CudaNdarray *)rval, nd))
{
...
...
@@ -2136,7 +2135,7 @@ CudaNdarray_Dot(PyObject* _unused, PyObject* args)
PyErr_SetString(PyExc_TypeError, "need 2d CudaNdarray arg for now");
goto CudaNdarray_dot_fail;
}
rval = CudaNdarray_
new_null
();
rval = CudaNdarray_
New
();
if (!rval)
{
goto CudaNdarray_dot_fail;
...
...
@@ -2246,7 +2245,7 @@ filter(PyObject* __unsed_self, PyObject *args) // args = (data, broadcastable, s
}
else
{
rval = (CudaNdarray*) CudaNdarray_
new_null
();
rval = (CudaNdarray*) CudaNdarray_
New
();
}
if (rval)
{
...
...
@@ -2450,16 +2449,11 @@ CudaNdarray_is_c_contiguous(const CudaNdarray * self)
}
return c_contiguous;
}
PyObject *
CudaNdarray_new_null()
{
//TODO: this function is deprecated... do not use. Consider removing.
return CudaNdarray_New(-1);
}
PyObject *
CudaNdarray_new_nd(int nd)
{
CudaNdarray * rval = (CudaNdarray*) CudaNdarray_
new_null
();
CudaNdarray * rval = (CudaNdarray*) CudaNdarray_
New
();
if (!rval || CudaNdarray_set_nd(rval, nd))
{
Py_XDECREF(rval);
...
...
theano/sandbox/cuda/cuda_ndarray.cuh
浏览文件 @
627bd58e
...
...
@@ -81,7 +81,7 @@ struct CudaNdarray
* Return a CudaNdarray whose 'nd' dimensions are all 0.
*/
PyObject *
CudaNdarray_New(int nd);
CudaNdarray_New(int nd
=-1
);
/**
* Return 1 for a CudaNdarray otw 0
...
...
@@ -296,11 +296,6 @@ CudaNdarray_SIZE_Object(const CudaNdarray *self, void *closure)
}
/**
* Allocate a new CudaNdarray with nd==-1
*/
PyObject * CudaNdarray_new_null();
/**
* Allocate a new CudaNdarray with room for given number of dimensions
*
...
...
@@ -424,7 +419,7 @@ template<typename inttype>
PyObject *
CudaNdarray_NewDims(int nd, const inttype * dims)
{
CudaNdarray * rval = (CudaNdarray*)CudaNdarray_
new_null
();
CudaNdarray * rval = (CudaNdarray*)CudaNdarray_
New
();
if (rval)
{
if (CudaNdarray_alloc_contiguous(rval, nd, dims))
...
...
theano/sandbox/cuda/elemwise.py
浏览文件 @
627bd58e
...
...
@@ -37,7 +37,7 @@ def get_str_list_logical_scalar(node, value_str='ii_i%i_value', data_str='ii_i%i
class
NaiveAlgo
(
object
):
verbose
=
0
# 1, 2 or 3 for more verbose output.
cache_version
=
()
cache_version
=
(
'debug'
,
1
3
,
verbose
)
cache_version
=
(
'debug'
,
1
4
,
verbose
)
def
__init__
(
self
,
scalar_op
,
sync
=
True
,
inplace_pattern
=
{}):
"""
...
...
@@ -888,7 +888,7 @@ nd_collapse_[i]=0;
}
if (NULL ==
%(oname)
s)
{
%(oname)
s = (CudaNdarray*)CudaNdarray_
new_null
();
%(oname)
s = (CudaNdarray*)CudaNdarray_
New
();
if (!
%(oname)
s)
{
//error string already set
...
...
theano/sandbox/cuda/nnet.py
浏览文件 @
627bd58e
...
...
@@ -191,7 +191,7 @@ class GpuCrossentropySoftmax1HotWithBiasDx (Op):
def
make_node
(
self
,
dy
,
sm
,
y_idx
):
return
Apply
(
self
,
[
dy
,
sm
,
y_idx
],[
sm
.
type
()])
def
c_code_cache_version
(
self
):
return
(
3
,)
return
(
4
,)
#return ()
def
c_code
(
self
,
node
,
nodename
,
inp
,
out
,
sub
):
dnll
,
sm
,
y_idx
=
inp
...
...
@@ -221,7 +221,7 @@ class GpuCrossentropySoftmax1HotWithBiasDx (Op):
|| (CudaNdarray_HOST_DIMS(
%(dx)
s)[1] != CudaNdarray_HOST_DIMS(
%(sm)
s)[1]))
{
Py_XDECREF(
%(dx)
s);
%(dx)
s = (CudaNdarray*)CudaNdarray_
new_null
();
%(dx)
s = (CudaNdarray*)CudaNdarray_
New
();
if ((NULL ==
%(dx)
s)
|| CudaNdarray_alloc_contiguous(
%(dx)
s, 2, CudaNdarray_HOST_DIMS(
%(sm)
s)))
{
...
...
@@ -309,7 +309,7 @@ class GpuSoftmax (Op):
return
shape
def
c_code_cache_version
(
self
):
#return ()
return
(
2
,)
+
inline_softmax
.
code_version
return
(
3
,)
+
inline_softmax
.
code_version
def
c_code
(
self
,
node
,
nodename
,
inp
,
out
,
sub
):
x
,
=
inp
z
,
=
out
...
...
@@ -325,7 +325,7 @@ class GpuSoftmax (Op):
|| (CudaNdarray_HOST_DIMS(
%(z)
s)[1] != CudaNdarray_HOST_DIMS(
%(x)
s)[1]))
{
Py_XDECREF(
%(z)
s);
%(z)
s = (CudaNdarray*)CudaNdarray_
new_null
();
%(z)
s = (CudaNdarray*)CudaNdarray_
New
();
if ((NULL ==
%(z)
s)
|| CudaNdarray_alloc_contiguous(
%(z)
s, 2, CudaNdarray_HOST_DIMS(
%(x)
s)))
{
...
...
@@ -398,7 +398,7 @@ class GpuSoftmaxWithBias (Op):
return
[
shape
[
0
]]
def
c_code_cache_version
(
self
):
#return ()
return
(
2
,)
+
inline_softmax
.
code_version
return
(
3
,)
+
inline_softmax
.
code_version
def
c_code
(
self
,
node
,
nodename
,
inp
,
out
,
sub
):
x
,
b
=
inp
...
...
@@ -426,7 +426,7 @@ class GpuSoftmaxWithBias (Op):
|| (CudaNdarray_HOST_DIMS(
%(z)
s)[1] != CudaNdarray_HOST_DIMS(
%(x)
s)[1]))
{
Py_XDECREF(
%(z)
s);
%(z)
s = (CudaNdarray*)CudaNdarray_
new_null
();
%(z)
s = (CudaNdarray*)CudaNdarray_
New
();
if ((NULL ==
%(z)
s)
|| CudaNdarray_alloc_contiguous(
%(z)
s, 2, CudaNdarray_HOST_DIMS(
%(x)
s)))
{
...
...
编写
预览
Markdown
格式
0%
重试
或
添加新文件
添加附件
取消
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!
取消
请
注册
或者
登录
后发表评论