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
GoAI
Paddle
提交
782ddc5f
P
Paddle
项目概览
GoAI
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
1
Star
1
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1
列表
看板
标记
里程碑
合并请求
0
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
Issue
1
列表
看板
标记
里程碑
合并请求
0
合并请求
0
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
782ddc5f
编写于
1月 21, 2018
作者:
C
chengduoZH
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
follow comments
上级
cd38e2d1
变更
4
隐藏空白更改
内联
并排
Showing
4 changed file
with
9 addition
and
8 deletion
+9
-8
paddle/operators/math/matmul.h
paddle/operators/math/matmul.h
+1
-1
paddle/operators/matmul_op.cc
paddle/operators/matmul_op.cc
+2
-2
paddle/operators/matmul_op.h
paddle/operators/matmul_op.h
+1
-1
python/paddle/v2/fluid/tests/test_matmul_op.py
python/paddle/v2/fluid/tests/test_matmul_op.py
+5
-4
未找到文件。
paddle/operators/math/matmul.h
浏览文件 @
782ddc5f
...
...
@@ -49,7 +49,7 @@ class MatMulFunctor {
"The dimensions of X and Y must be the same, and both of "
"them should be %d-dimensional."
,
dim_b
.
size
());
// The f
ron
t rank-2 dimensions are accumulated on the batch_count, and the
// The f
irs
t rank-2 dimensions are accumulated on the batch_count, and the
// last two dimensions are used for matrix multiplication.
for
(
int
j
=
0
;
j
<
dim_a
.
size
()
-
2
;
++
j
)
{
PADDLE_ENFORCE_EQ
(
dim_b
[
j
],
dim_a
[
j
],
...
...
paddle/operators/matmul_op.cc
浏览文件 @
782ddc5f
...
...
@@ -51,7 +51,7 @@ class MatMulOp : public framework::OperatorWithKernel {
"them should be %d-dimensional."
,
dim_x
.
size
());
// The f
ron
t rank-2 dimensions are accumulated on the batch_count, and the
// The f
irs
t rank-2 dimensions are accumulated on the batch_count, and the
// last two dimensions are used for matrix multiplication.
for
(
int
j
=
0
;
j
<
dim_x
.
size
()
-
2
;
++
j
)
{
PADDLE_ENFORCE_EQ
(
dim_y
[
j
],
dim_x
[
j
],
...
...
@@ -196,7 +196,7 @@ The differences are:
- When the rank of the input data is less than or equal to 3, it
is similar to the `numpy.matmul` function.
- When the rank of the input is greater than 3, the rank of X and
Y must be equal, and the f
ron
t `rank - 2` dimensions must be equal.
Y must be equal, and the f
irs
t `rank - 2` dimensions must be equal.
- We add `transpose_X` and `transpose_Y` flags.
Both the input `X` and `Y` can carry the LoD (Level of Details) information,
...
...
paddle/operators/matmul_op.h
浏览文件 @
782ddc5f
...
...
@@ -138,7 +138,7 @@ class MatMulGradKernel : public framework::OpKernel<T> {
}
int
batch_count
=
0
;
// The f
ron
t rank-2 dimensions are accumulated on the batch_count, and the
// The f
irs
t rank-2 dimensions are accumulated on the batch_count, and the
// last two dimensions are used for matrix multiplication.
if
(
x_dims
.
size
()
>
3
)
{
batch_count
=
accumulate
(
x_dims
.
begin
(),
x_dims
.
end
()
-
2
,
1
,
...
...
python/paddle/v2/fluid/tests/test_matmul_op.py
浏览文件 @
782ddc5f
...
...
@@ -127,6 +127,7 @@ for dim_X in [1, 2, 3]:
})
# Test case n-dim
def
generate_compatible_shapes
(
dim
,
transpose_X
,
transpose_Y
):
M
=
2
N
=
4
...
...
@@ -135,14 +136,14 @@ def generate_compatible_shapes(dim, transpose_X, transpose_Y):
shape_Y
=
[
2
for
_
in
range
(
dim
-
2
)]
if
transpose_X
:
shape_X
=
shape_X
+
[
K
,
M
]
shape_X
+=
[
K
,
M
]
else
:
shape_X
=
shape_X
+
[
M
,
K
]
shape_X
+=
[
M
,
K
]
if
transpose_Y
:
shape_Y
=
shape_Y
+
[
N
,
K
]
shape_Y
+=
[
N
,
K
]
else
:
shape_Y
=
shape_Y
+
[
K
,
N
]
shape_Y
+=
[
K
,
N
]
return
shape_X
,
shape_Y
...
...
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