Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
PaddlePaddle
Paddle
提交
d0521e6f
P
Paddle
项目概览
PaddlePaddle
/
Paddle
12 个月 前同步成功
通知
2292
Star
20931
Fork
5422
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1423
列表
看板
标记
里程碑
合并请求
543
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
Paddle
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
1,423
Issue
1,423
列表
看板
标记
里程碑
合并请求
543
合并请求
543
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
体验新版 GitCode,发现更多精彩内容 >>
提交
d0521e6f
编写于
9月 14, 2017
作者:
L
Luo Tao
浏览文件
操作
浏览文件
下载
差异文件
Merge branch 'develop' into use_op
上级
f657e21f
0e46f5eb
变更
6
隐藏空白更改
内联
并排
Showing
6 changed file
with
231 addition
and
95 deletion
+231
-95
paddle/operators/cos_sim_op.cc
paddle/operators/cos_sim_op.cc
+69
-26
paddle/operators/cos_sim_op.h
paddle/operators/cos_sim_op.h
+92
-55
paddle/operators/onehot_cross_entropy_op.cc
paddle/operators/onehot_cross_entropy_op.cc
+1
-1
python/paddle/v2/framework/tests/test_cos_sim_op.py
python/paddle/v2/framework/tests/test_cos_sim_op.py
+59
-5
python/paddle/v2/framework/tests/test_cross_entropy_op.py
python/paddle/v2/framework/tests/test_cross_entropy_op.py
+9
-7
python/paddle/v2/framework/tests/test_pad_op.py
python/paddle/v2/framework/tests/test_pad_op.py
+1
-1
未找到文件。
paddle/operators/cos_sim_op.cc
浏览文件 @
d0521e6f
...
...
@@ -25,16 +25,30 @@ class CosSimOp : public framework::OperatorWithKernel {
protected:
void
InferShape
(
const
framework
::
InferShapeContext
&
ctx
)
const
override
{
// notnull check
PADDLE_ENFORCE_NOT_NULL
(
ctx
.
InputVar
(
"X"
),
"Input(X) must not be null."
);
PADDLE_ENFORCE_NOT_NULL
(
ctx
.
InputVar
(
"Y"
),
"Input(Y) must not be null."
);
PADDLE_ENFORCE_EQ
(
ctx
.
Input
<
Tensor
>
(
"X"
)
->
dims
(),
ctx
.
Input
<
Tensor
>
(
"Y"
)
->
dims
(),
"Dimensions of Input(X) and Input(Y) must be the same."
);
auto
dims
=
ctx
.
Input
<
Tensor
>
(
"X"
)
->
dims
();
ctx
.
Output
<
Tensor
>
(
"Out"
)
->
Resize
({
dims
[
0
],
1
});
ctx
.
Output
<
Tensor
>
(
"XNorm"
)
->
Resize
({
dims
[
0
],
1
});
ctx
.
Output
<
Tensor
>
(
"YNorm"
)
->
Resize
({
dims
[
0
],
1
});
// shape check
auto
x_dims
=
ctx
.
Input
<
Tensor
>
(
"X"
)
->
dims
();
auto
y_dims
=
ctx
.
Input
<
Tensor
>
(
"Y"
)
->
dims
();
PADDLE_ENFORCE_EQ
(
x_dims
.
size
(),
y_dims
.
size
(),
"Ranks of Input(X) and Input(Y) must be equal."
);
PADDLE_ENFORCE_GE
(
x_dims
.
size
(),
2
,
"Rank of Input(X) must not be less than 2."
);
PADDLE_ENFORCE_EQ
(
framework
::
slice_ddim
(
x_dims
,
1
,
x_dims
.
size
()),
framework
::
slice_ddim
(
y_dims
,
1
,
y_dims
.
size
()),
"All dimensions except the 1st of Input(X) and Input(Y) "
"must be equal."
);
PADDLE_ENFORCE
(
x_dims
[
0
]
==
y_dims
[
0
]
||
y_dims
[
0
]
==
1
,
"The 1st dimension of Input(Y) must be equal to Input(X) or"
" just 1 (which will be broadcasted to match Input(X))."
);
// resize tensor
ctx
.
Output
<
Tensor
>
(
"Out"
)
->
Resize
({
x_dims
[
0
],
1
});
ctx
.
Output
<
Tensor
>
(
"XNorm"
)
->
Resize
({
x_dims
[
0
],
1
});
ctx
.
Output
<
Tensor
>
(
"YNorm"
)
->
Resize
({
y_dims
[
0
],
1
});
}
};
...
...
@@ -42,16 +56,27 @@ class CosSimOpMaker : public framework::OpProtoAndCheckerMaker {
public:
CosSimOpMaker
(
framework
::
OpProto
*
proto
,
framework
::
OpAttrChecker
*
op_checker
)
:
OpProtoAndCheckerMaker
(
proto
,
op_checker
)
{
AddInput
(
"X"
,
"The
fir
st input of cos_sim op."
);
AddInput
(
"Y"
,
"The
seco
nd input of cos_sim op."
);
AddInput
(
"X"
,
"The
1
st input of cos_sim op."
);
AddInput
(
"Y"
,
"The
2
nd input of cos_sim op."
);
AddOutput
(
"Out"
,
"The output of cos_sim op."
);
AddOutput
(
"XNorm"
,
"Row norm of the first input."
).
AsIntermediate
();
AddOutput
(
"YNorm"
,
"Row norm of the second input."
).
AsIntermediate
();
AddOutput
(
"XNorm"
,
"Norm of the first input, reduced along the 1st "
"dimension."
)
.
AsIntermediate
();
AddOutput
(
"YNorm"
,
"Norm of the second input, reduced along the 1st "
"dimension."
)
.
AsIntermediate
();
AddComment
(
R"DOC(
Cosine Similarity Operator.
The equation is: Out = X^T * Y / (sqrt(X^T * X) * sqrt(Y^T * Y))
The equation is: Out = X^T * Y / (sqrt(X^T * X) * sqrt(Y^T * Y)).
Input(X) and Input(Y) must have the same shape, except that the 1st dimension
of Input(Y) could be just 1 (different from Input(X)), which will be
broadcasted to match the shape of Input(X) before computing their cosine
similarity.
)DOC"
);
}
};
...
...
@@ -62,32 +87,50 @@ class CosSimOpGrad : public framework::OperatorWithKernel {
protected:
void
InferShape
(
const
framework
::
InferShapeContext
&
ctx
)
const
override
{
// notnull check
PADDLE_ENFORCE_NOT_NULL
(
ctx
.
InputVar
(
"X"
),
"Input(X) must not be null."
);
PADDLE_ENFORCE_NOT_NULL
(
ctx
.
InputVar
(
"Y"
),
"Input(Y) must not be null."
);
PADDLE_ENFORCE_NOT_NULL
(
ctx
.
InputVar
(
"XNorm"
),
"Input(XNorm) must not be null."
);
PADDLE_ENFORCE_NOT_NULL
(
ctx
.
InputVar
(
"YNorm"
),
"Input(YNorm) must not be null."
);
PADDLE_ENFORCE_NOT_NULL
(
ctx
.
InputVar
(
"Out"
),
"Input(Out) must not be null."
);
PADDLE_ENFORCE_NOT_NULL
(
ctx
.
InputVar
(
framework
::
GradVarName
(
"Out"
)),
"Input(Out@GRAD) must not be null."
);
// shape check
auto
x_dims
=
ctx
.
Input
<
Tensor
>
(
"X"
)
->
dims
();
auto
y_dims
=
ctx
.
Input
<
Tensor
>
(
"Y"
)
->
dims
();
auto
xnorm_dims
=
ctx
.
Input
<
Tensor
>
(
"XNorm"
)
->
dims
();
auto
ynorm_dims
=
ctx
.
Input
<
Tensor
>
(
"YNorm"
)
->
dims
();
auto
out_dims
=
ctx
.
Input
<
Tensor
>
(
framework
::
GradVarName
(
"Out"
))
->
dims
();
PADDLE_ENFORCE_EQ
(
x_dims
,
y_dims
,
"Dimensions of Input(X) and Input(Y) must be the same."
);
PADDLE_ENFORCE_EQ
(
xnorm_dims
[
0
],
x_dims
[
0
],
"1st dimension of XNorm must equal that of Input(X)."
);
PADDLE_ENFORCE_EQ
(
xnorm_dims
[
1
],
1
,
"2st dimension of XNorm must be one."
);
PADDLE_ENFORCE_EQ
(
ynorm_dims
[
0
],
y_dims
[
0
],
"1st dimension of YNorm must equal that of Input(Y)."
);
PADDLE_ENFORCE_EQ
(
ynorm_dims
[
1
],
1
,
"2st dimension of YNorm must be one."
);
PADDLE_ENFORCE_EQ
(
out_dims
[
0
],
x_dims
[
0
],
"1st dimension of Out@GRAD must equal that of Input(X)"
);
PADDLE_ENFORCE_EQ
(
out_dims
[
1
],
1
,
"1st dimension of Out@GRAD must be one."
);
auto
out_dims
=
ctx
.
Input
<
Tensor
>
(
"Out"
)
->
dims
();
auto
out_grad_dims
=
ctx
.
Input
<
Tensor
>
(
framework
::
GradVarName
(
"Out"
))
->
dims
();
PADDLE_ENFORCE_GE
(
x_dims
.
size
(),
y_dims
.
size
(),
"Ranks of Input(X) and Input(Y) must be equal."
);
PADDLE_ENFORCE_GE
(
x_dims
.
size
(),
2
,
"Rank of Input(X) must not be less than 2."
);
PADDLE_ENFORCE_EQ
(
framework
::
slice_ddim
(
x_dims
,
1
,
x_dims
.
size
()),
framework
::
slice_ddim
(
y_dims
,
1
,
y_dims
.
size
()),
"All dimensions except the 1st of Input(X) and Input(Y) "
"must be equal."
);
PADDLE_ENFORCE
(
x_dims
[
0
]
==
y_dims
[
0
]
||
y_dims
[
0
]
==
1
,
"The 1st dimension of Input(Y) must be equal to Input(X) or"
" just 1 (which will be broadcasted to match Input(X))."
);
auto
target_xnorm_dims
=
framework
::
make_ddim
({
x_dims
[
0
],
1
});
auto
target_ynorm_dims
=
framework
::
make_ddim
({
y_dims
[
0
],
1
});
PADDLE_ENFORCE_EQ
(
xnorm_dims
,
target_xnorm_dims
,
"Shape of Input(XNorm) must be [X.Dim(0), 1]."
);
PADDLE_ENFORCE_EQ
(
ynorm_dims
,
target_ynorm_dims
,
"Shape of Input(YNorm) must be [Y.Dim(0), 1]."
);
PADDLE_ENFORCE_EQ
(
out_dims
,
target_xnorm_dims
,
"Shape of Input(Out) must be [X.Dim(0), 1]."
);
PADDLE_ENFORCE_EQ
(
out_grad_dims
,
target_xnorm_dims
,
"Shape of Input(Out@Grad) must be [X.Dim(0), 1]."
);
// resize tensor
auto
*
x_grad
=
ctx
.
Output
<
Tensor
>
(
framework
::
GradVarName
(
"X"
));
auto
*
y_grad
=
ctx
.
Output
<
Tensor
>
(
framework
::
GradVarName
(
"Y"
));
if
(
x_grad
)
x_grad
->
Resize
(
x_dims
);
...
...
paddle/operators/cos_sim_op.h
浏览文件 @
d0521e6f
...
...
@@ -31,30 +31,38 @@ template <typename Place, typename T>
class
CosSimKernel
:
public
framework
::
OpKernel
{
public:
void
Compute
(
const
framework
::
ExecutionContext
&
context
)
const
override
{
auto
*
input_x
=
context
.
Input
<
Tensor
>
(
"X"
);
auto
*
input_y
=
context
.
Input
<
Tensor
>
(
"Y"
);
auto
*
output_z
=
context
.
Output
<
Tensor
>
(
"Out"
);
auto
*
output_x_norm
=
context
.
Output
<
Tensor
>
(
"XNorm"
);
auto
*
output_y_norm
=
context
.
Output
<
Tensor
>
(
"YNorm"
);
// get Tensor
auto
*
in_x
=
context
.
Input
<
Tensor
>
(
"X"
);
auto
*
in_y
=
context
.
Input
<
Tensor
>
(
"Y"
);
auto
*
out_z
=
context
.
Output
<
Tensor
>
(
"Out"
);
auto
*
out_x_norm
=
context
.
Output
<
Tensor
>
(
"XNorm"
);
auto
*
out_y_norm
=
context
.
Output
<
Tensor
>
(
"YNorm"
);
out_z
->
mutable_data
<
T
>
(
context
.
GetPlace
());
out_x_norm
->
mutable_data
<
T
>
(
context
.
GetPlace
());
out_y_norm
->
mutable_data
<
T
>
(
context
.
GetPlace
());
output_z
->
mutable_data
<
T
>
(
context
.
GetPlace
());
output_x_norm
->
mutable_data
<
T
>
(
context
.
GetPlace
());
output_y_norm
->
mutable_data
<
T
>
(
context
.
GetPlace
());
auto
dims
=
input_x
->
dims
();
int64_t
size
=
input_x
->
numel
();
auto
new_dims
=
framework
::
make_ddim
({
dims
[
0
],
size
/
dims
[
0
]});
auto
x
=
EigenMatrix
<
T
>::
From
(
*
input_x
,
new_dims
);
auto
y
=
EigenMatrix
<
T
>::
From
(
*
input_y
,
new_dims
);
auto
z
=
EigenVector
<
T
>::
Flatten
(
*
output_z
);
auto
x_norm
=
EigenVector
<
T
>::
Flatten
(
*
output_x_norm
);
auto
y_norm
=
EigenVector
<
T
>::
Flatten
(
*
output_y_norm
);
// convert Tensor to Eigen Tensor
int
rows_x
=
in_x
->
dims
()[
0
];
int
rows_y
=
in_y
->
dims
()[
0
];
auto
x
=
EigenMatrix
<
T
>::
Reshape
(
*
in_x
,
1
);
auto
y
=
EigenMatrix
<
T
>::
Reshape
(
*
in_y
,
1
);
auto
z
=
EigenVector
<
T
>::
Flatten
(
*
out_z
);
auto
x_norm
=
EigenVector
<
T
>::
Flatten
(
*
out_x_norm
);
auto
y_norm
=
EigenVector
<
T
>::
Flatten
(
*
out_y_norm
);
// compute
auto
place
=
context
.
GetEigenDevice
<
Place
>
();
auto
xy
=
(
x
*
y
).
sum
(
Eigen
::
array
<
int
,
1
>
({{
1
}}));
x_norm
.
device
(
place
)
=
x
.
square
().
sum
(
Eigen
::
array
<
int
,
1
>
({{
1
}})).
sqrt
();
y_norm
.
device
(
place
)
=
y
.
square
().
sum
(
Eigen
::
array
<
int
,
1
>
({{
1
}})).
sqrt
();
z
.
device
(
place
)
=
xy
/
x_norm
/
y_norm
;
auto
row_along
=
Eigen
::
array
<
int
,
1
>
({{
1
}});
x_norm
.
device
(
place
)
=
x
.
square
().
sum
(
row_along
).
sqrt
();
y_norm
.
device
(
place
)
=
y
.
square
().
sum
(
row_along
).
sqrt
();
if
(
rows_x
==
rows_y
)
{
auto
xy
=
(
x
*
y
).
sum
(
Eigen
::
array
<
int
,
1
>
({
1
}));
z
.
device
(
place
)
=
xy
/
x_norm
/
y_norm
;
}
else
{
Eigen
::
DSizes
<
int
,
2
>
bcast
(
rows_x
,
1
);
auto
xy
=
(
x
*
y
.
broadcast
(
bcast
)).
sum
(
row_along
);
z
.
device
(
place
)
=
xy
/
x_norm
/
y_norm
.
broadcast
(
bcast
);
}
}
};
...
...
@@ -62,43 +70,72 @@ template <typename Place, typename T>
class
CosSimGradKernel
:
public
framework
::
OpKernel
{
public:
void
Compute
(
const
framework
::
ExecutionContext
&
context
)
const
override
{
auto
*
input_x
=
context
.
Input
<
Tensor
>
(
"X"
);
auto
*
input_y
=
context
.
Input
<
Tensor
>
(
"Y"
);
auto
*
input_z
=
context
.
Input
<
Tensor
>
(
"Out"
);
auto
*
input_x_norm
=
context
.
Input
<
Tensor
>
(
"XNorm"
);
auto
*
input_y_norm
=
context
.
Input
<
Tensor
>
(
"YNorm"
);
auto
*
output_grad_x
=
context
.
Output
<
Tensor
>
(
framework
::
GradVarName
(
"X"
));
auto
*
output_grad_y
=
context
.
Output
<
Tensor
>
(
framework
::
GradVarName
(
"Y"
));
auto
*
input_grad_z
=
context
.
Input
<
Tensor
>
(
framework
::
GradVarName
(
"Out"
));
// get Tensor
auto
*
in_x
=
context
.
Input
<
Tensor
>
(
"X"
);
auto
*
in_y
=
context
.
Input
<
Tensor
>
(
"Y"
);
auto
*
in_z
=
context
.
Input
<
Tensor
>
(
"Out"
);
auto
*
in_x_norm
=
context
.
Input
<
Tensor
>
(
"XNorm"
);
auto
*
in_y_norm
=
context
.
Input
<
Tensor
>
(
"YNorm"
);
auto
*
out_grad_x
=
context
.
Output
<
Tensor
>
(
framework
::
GradVarName
(
"X"
));
auto
*
out_grad_y
=
context
.
Output
<
Tensor
>
(
framework
::
GradVarName
(
"Y"
));
auto
*
in_grad_z
=
context
.
Input
<
Tensor
>
(
framework
::
GradVarName
(
"Out"
));
auto
dims
=
input_x
->
dims
();
int64_t
size
=
input_x
->
numel
();
auto
new_dims
=
framework
::
make_ddim
({
dims
[
0
],
size
/
dims
[
0
]});
auto
x
=
EigenMatrix
<
T
>::
From
(
*
input_x
,
new_dims
);
auto
y
=
EigenMatrix
<
T
>::
From
(
*
input_y
,
new_dims
);
auto
z
=
EigenMatrix
<
T
>::
From
(
*
input_z
);
auto
x_norm
=
EigenMatrix
<
T
>::
From
(
*
input_x_norm
);
auto
y_norm
=
EigenMatrix
<
T
>::
From
(
*
input_y_norm
);
auto
dz
=
EigenMatrix
<
T
>::
From
(
*
input_grad_z
);
// convert Tensor to Eigen Tensor
auto
x
=
EigenMatrix
<
T
>::
Reshape
(
*
in_x
,
1
);
auto
y
=
EigenMatrix
<
T
>::
Reshape
(
*
in_y
,
1
);
auto
z
=
EigenMatrix
<
T
>::
Reshape
(
*
in_z
,
1
);
auto
x_norm
=
EigenMatrix
<
T
>::
Reshape
(
*
in_x_norm
,
1
);
auto
y_norm
=
EigenMatrix
<
T
>::
Reshape
(
*
in_y_norm
,
1
);
auto
dz
=
EigenMatrix
<
T
>::
Reshape
(
*
in_grad_z
,
1
);
Eigen
::
DSizes
<
int
,
2
>
bcast
(
1
,
new_dims
[
1
]);
auto
z_bcast
=
z
.
broadcast
(
bcast
);
auto
dz_bcast
=
dz
.
broadcast
(
bcast
);
// compute gradident
int
rows_x
=
in_x
->
dims
()[
0
];
int
rows_y
=
in_y
->
dims
()[
0
];
int
cols
=
framework
::
product
(
in_x
->
dims
())
/
rows_x
;
Eigen
::
DSizes
<
int
,
2
>
bcast_cols
(
1
,
cols
);
auto
z_bcast
=
z
.
broadcast
(
bcast_cols
);
auto
dz_bcast
=
dz
.
broadcast
(
bcast_cols
);
auto
x_snorm_bcast
=
x_norm
.
square
().
eval
().
broadcast
(
bcast_cols
);
auto
place
=
context
.
GetEigenDevice
<
Place
>
();
auto
x_snorm_bcast
=
x_norm
.
square
().
eval
().
broadcast
(
bcast
);
auto
y_snorm_bcast
=
y_norm
.
square
().
eval
().
broadcast
(
bcast
);
auto
norm_prod_bcast
=
(
x_norm
*
y_norm
).
eval
().
broadcast
(
bcast
);
if
(
output_grad_x
)
{
output_grad_x
->
mutable_data
<
T
>
(
context
.
GetPlace
());
auto
dx
=
EigenMatrix
<
T
>::
From
(
*
output_grad_x
,
new_dims
);
dx
.
device
(
place
)
=
dz_bcast
*
(
y
/
norm_prod_bcast
-
z_bcast
*
x
/
x_snorm_bcast
);
}
if
(
output_grad_y
)
{
output_grad_y
->
mutable_data
<
T
>
(
context
.
GetPlace
());
auto
dy
=
EigenMatrix
<
T
>::
From
(
*
output_grad_y
,
new_dims
);
dy
.
device
(
place
)
=
dz_bcast
*
(
x
/
norm_prod_bcast
-
z_bcast
*
y
/
y_snorm_bcast
);
if
(
rows_x
==
rows_y
)
{
auto
y_snorm_bcast
=
y_norm
.
square
().
eval
().
broadcast
(
bcast_cols
);
auto
norm_prod_bcast
=
(
x_norm
*
y_norm
).
eval
().
broadcast
(
bcast_cols
);
// compute dx
if
(
out_grad_x
)
{
out_grad_x
->
mutable_data
<
T
>
(
context
.
GetPlace
());
auto
dx
=
EigenMatrix
<
T
>::
Reshape
(
*
out_grad_x
,
1
);
auto
grad
=
y
/
norm_prod_bcast
-
z_bcast
*
x
/
x_snorm_bcast
;
dx
.
device
(
place
)
=
dz_bcast
*
grad
;
}
// compute dy
if
(
out_grad_y
)
{
out_grad_y
->
mutable_data
<
T
>
(
context
.
GetPlace
());
auto
dy
=
EigenMatrix
<
T
>::
Reshape
(
*
out_grad_y
,
1
);
auto
grad
=
x
/
norm_prod_bcast
-
z_bcast
*
y
/
y_snorm_bcast
;
dy
.
device
(
place
)
=
dz_bcast
*
grad
;
}
}
else
{
Eigen
::
DSizes
<
int
,
2
>
bcast_rows
(
rows_x
,
1
);
Eigen
::
DSizes
<
int
,
2
>
bcast_rows_cols
(
rows_x
,
cols
);
auto
y_bcast
=
y
.
broadcast
(
bcast_rows
);
auto
y_snorm_bcast
=
y_norm
.
square
().
eval
().
broadcast
(
bcast_rows_cols
);
auto
norm_prod_bcast
=
(
x_norm
*
y_norm
.
eval
().
broadcast
(
bcast_rows
))
.
eval
()
.
broadcast
(
bcast_cols
);
// compute dx
if
(
out_grad_x
)
{
out_grad_x
->
mutable_data
<
T
>
(
context
.
GetPlace
());
auto
dx
=
EigenMatrix
<
T
>::
Reshape
(
*
out_grad_x
,
1
);
auto
grad
=
y_bcast
/
norm_prod_bcast
-
z_bcast
*
x
/
x_snorm_bcast
;
dx
.
device
(
place
)
=
dz_bcast
*
grad
;
}
// compute dy
if
(
out_grad_y
)
{
out_grad_y
->
mutable_data
<
T
>
(
context
.
GetPlace
());
auto
dy
=
EigenMatrix
<
T
>::
Reshape
(
*
out_grad_y
,
1
);
auto
grad
=
x
/
norm_prod_bcast
-
z_bcast
*
y_bcast
/
y_snorm_bcast
;
dy
.
device
(
place
)
=
(
dz_bcast
*
grad
).
sum
(
Eigen
::
array
<
int
,
1
>
({
0
}));
}
}
}
};
...
...
paddle/operators/onehot_cross_entropy_op.cc
浏览文件 @
d0521e6f
...
...
@@ -29,7 +29,7 @@ class OnehotCrossEntropyOp : public framework::OperatorWithKernel {
PADDLE_ENFORCE_EQ
(
X
->
dims
().
size
(),
2
,
"X's dimension must be 2."
);
PADDLE_ENFORCE_EQ
(
label
->
dims
().
size
(),
1
,
"label's dimension must be 1."
);
PADDLE_ENFORCE_EQ
(
X
->
dims
()[
0
],
label
->
dims
()[
0
]);
ctx
.
Output
<
Tensor
>
(
"Y"
)
->
Resize
({
X
->
dims
()[
0
]});
ctx
.
Output
<
Tensor
>
(
"Y"
)
->
Resize
({
X
->
dims
()[
0
]
,
1
});
}
};
...
...
python/paddle/v2/framework/tests/test_cos_sim_op.py
浏览文件 @
d0521e6f
...
...
@@ -7,8 +7,8 @@ class TestCosSimOp(OpTest):
def
setUp
(
self
):
self
.
op_type
=
"cos_sim"
self
.
inputs
=
{
'X'
:
np
.
random
.
random
((
10
,
5
)).
astype
(
"float32"
),
'Y'
:
np
.
random
.
random
((
10
,
5
)).
astype
(
"float32"
)
'X'
:
np
.
random
.
random
((
6
,
5
)).
astype
(
"float32"
),
'Y'
:
np
.
random
.
random
((
6
,
5
)).
astype
(
"float32"
)
}
expect_x_norm
=
np
.
linalg
.
norm
(
self
.
inputs
[
'X'
],
axis
=
1
)
expect_y_norm
=
np
.
linalg
.
norm
(
self
.
inputs
[
'Y'
],
axis
=
1
)
...
...
@@ -28,12 +28,66 @@ class TestCosSimOp(OpTest):
def
test_check_grad_ingore_x
(
self
):
self
.
check_grad
(
[
'Y'
],
'Out'
,
max_relative_error
=
0.05
,
no_grad_set
=
set
(
'X'
))
[
'Y'
],
'Out'
,
max_relative_error
=
0.05
,
no_grad_set
=
set
(
"X"
))
def
test_check_grad_i
gn
ore_y
(
self
):
def
test_check_grad_i
ng
ore_y
(
self
):
self
.
check_grad
(
[
'X'
],
'Out'
,
max_relative_error
=
0.05
,
no_grad_set
=
set
(
'Y'
))
if
__name__
==
"__main__"
:
class
TestCosSimOp2
(
TestCosSimOp
):
def
setUp
(
self
):
self
.
op_type
=
"cos_sim"
self
.
inputs
=
{
'X'
:
np
.
random
.
random
((
6
,
5
)).
astype
(
"float32"
),
'Y'
:
np
.
random
.
random
((
1
,
5
)).
astype
(
"float32"
)
}
expect_x_norm
=
np
.
linalg
.
norm
(
self
.
inputs
[
'X'
],
axis
=
1
)
expect_y_norm
=
np
.
linalg
.
norm
(
self
.
inputs
[
'Y'
],
axis
=
1
)
expect_out
=
(
self
.
inputs
[
'X'
]
*
self
.
inputs
[
'Y'
]).
sum
(
axis
=
1
)
/
\
expect_x_norm
/
expect_y_norm
self
.
outputs
=
{
'XNorm'
:
np
.
expand_dims
(
expect_x_norm
,
1
),
'YNorm'
:
np
.
expand_dims
(
expect_y_norm
,
1
),
'Out'
:
np
.
expand_dims
(
expect_out
,
1
)
}
class
TestCosSimOp3
(
TestCosSimOp
):
def
setUp
(
self
):
self
.
op_type
=
"cos_sim"
self
.
inputs
=
{
'X'
:
np
.
random
.
random
((
6
,
5
,
2
)).
astype
(
"float32"
),
'Y'
:
np
.
random
.
random
((
6
,
5
,
2
)).
astype
(
"float32"
)
}
expect_x_norm
=
np
.
linalg
.
norm
(
self
.
inputs
[
'X'
],
axis
=
(
1
,
2
))
expect_y_norm
=
np
.
linalg
.
norm
(
self
.
inputs
[
'Y'
],
axis
=
(
1
,
2
))
expect_out
=
(
self
.
inputs
[
'X'
]
*
self
.
inputs
[
'Y'
]).
sum
(
axis
=
(
1
,
2
))
/
\
expect_x_norm
/
expect_y_norm
self
.
outputs
=
{
'XNorm'
:
np
.
expand_dims
(
expect_x_norm
,
1
),
'YNorm'
:
np
.
expand_dims
(
expect_y_norm
,
1
),
'Out'
:
np
.
expand_dims
(
expect_out
,
1
)
}
class
TestCosSimOp4
(
TestCosSimOp
):
def
setUp
(
self
):
self
.
op_type
=
"cos_sim"
self
.
inputs
=
{
'X'
:
np
.
random
.
random
((
6
,
5
,
2
)).
astype
(
"float32"
),
'Y'
:
np
.
random
.
random
((
1
,
5
,
2
)).
astype
(
"float32"
)
}
expect_x_norm
=
np
.
linalg
.
norm
(
self
.
inputs
[
'X'
],
axis
=
(
1
,
2
))
expect_y_norm
=
np
.
linalg
.
norm
(
self
.
inputs
[
'Y'
],
axis
=
(
1
,
2
))
expect_out
=
(
self
.
inputs
[
'X'
]
*
self
.
inputs
[
'Y'
]).
sum
(
axis
=
(
1
,
2
))
/
\
expect_x_norm
/
expect_y_norm
self
.
outputs
=
{
'XNorm'
:
np
.
expand_dims
(
expect_x_norm
,
1
),
'YNorm'
:
np
.
expand_dims
(
expect_y_norm
,
1
),
'Out'
:
np
.
expand_dims
(
expect_out
,
1
)
}
if
__name__
==
'__main__'
:
unittest
.
main
()
python/paddle/v2/framework/tests/test_cross_entropy_op.py
浏览文件 @
d0521e6f
...
...
@@ -8,20 +8,22 @@ class TestCrossEntropy(OpTest):
self
.
op_type
=
"onehot_cross_entropy"
batch_size
=
30
class_num
=
10
X
=
numpy
.
random
.
uniform
(
0.1
,
1.0
,
[
batch_size
,
class_num
]).
astype
(
"float32"
)
label
=
(
class_num
/
2
)
*
numpy
.
ones
(
batch_size
).
astype
(
"int32"
)
self
.
inputs
=
{
'X'
:
X
,
'label'
:
label
}
Y
=
[]
for
i
in
range
(
0
,
batch_size
):
Y
.
append
(
-
numpy
.
log
(
X
[
i
][
label
[
i
]]))
self
.
outputs
=
{
'Y'
:
numpy
.
array
(
Y
).
astype
(
"float32"
)}
labels
=
numpy
.
random
.
randint
(
0
,
class_num
,
batch_size
,
dtype
=
"int32"
)
cross_entropy
=
numpy
.
asmatrix
(
[[
-
numpy
.
log
(
X
[
i
][
labels
[
i
]])]
for
i
in
range
(
X
.
shape
[
0
])],
dtype
=
"float32"
)
self
.
inputs
=
{
"X"
:
X
,
"label"
:
labels
}
self
.
outputs
=
{
"Y"
:
cross_entropy
}
def
test_check_output
(
self
):
self
.
check_output
()
def
test_check_grad
(
self
):
self
.
check_grad
([
'X'
],
'Y'
)
self
.
check_grad
([
"X"
],
"Y"
)
if
__name__
==
"__main__"
:
...
...
python/paddle/v2/framework/tests/test_pad_op.py
浏览文件 @
d0521e6f
...
...
@@ -22,7 +22,7 @@ class TestPadOp(OpTest):
self
.
check_output
()
def
test_check_grad_normal
(
self
):
self
.
check_grad
([
'X'
],
'Out'
)
self
.
check_grad
([
'X'
],
'Out'
,
max_relative_error
=
0.006
)
def
initTestCase
(
self
):
self
.
shape
=
(
16
,
16
)
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录