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
Serving
提交
d986670e
S
Serving
项目概览
PaddlePaddle
/
Serving
大约 1 年 前同步成功
通知
185
Star
833
Fork
253
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
105
列表
看板
标记
里程碑
合并请求
10
Wiki
2
Wiki
分析
仓库
DevOps
项目成员
Pages
S
Serving
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
105
Issue
105
列表
看板
标记
里程碑
合并请求
10
合并请求
10
Pages
分析
分析
仓库分析
DevOps
Wiki
2
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
d986670e
编写于
2月 18, 2020
作者:
M
MRXLT
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
fix batch size
上级
4d435db4
变更
4
隐藏空白更改
内联
并排
Showing
4 changed file
with
31 addition
and
29 deletion
+31
-29
core/general-server/op/general_infer_helper.h
core/general-server/op/general_infer_helper.h
+8
-2
core/general-server/op/general_infer_op.cpp
core/general-server/op/general_infer_op.cpp
+10
-12
core/general-server/op/general_reader_op.cpp
core/general-server/op/general_reader_op.cpp
+3
-2
core/general-server/op/general_response_op.cpp
core/general-server/op/general_response_op.cpp
+10
-13
未找到文件。
core/general-server/op/general_infer_helper.h
浏览文件 @
d986670e
...
...
@@ -37,6 +37,8 @@ struct GeneralBlob {
double
infer_time
;
std
::
vector
<
std
::
string
>
fetch_name_vector
;
int
_batch_size
;
void
Clear
()
{
size_t
tensor_count
=
tensor_vector
.
size
();
for
(
size_t
ti
=
0
;
ti
<
tensor_count
;
++
ti
)
{
...
...
@@ -44,7 +46,11 @@ struct GeneralBlob {
}
tensor_vector
.
clear
();
}
int
SetBatchSize
(
int
batch_size
)
{
_batch_size
=
batch_size
;
}
int
GetBatchSize
()
const
{
return
_batch_size
;
}
/*
int GetBatchSize() const {
if (tensor_vector.size() > 0) {
if (tensor_vector[0].lod.size() == 1) {
...
...
@@ -56,7 +62,7 @@ struct GeneralBlob {
return -1;
}
}
*/
std
::
string
ShortDebugString
()
const
{
return
"Not implemented!"
;
}
};
...
...
core/general-server/op/general_infer_op.cpp
浏览文件 @
d986670e
...
...
@@ -12,11 +12,11 @@
// See the License for the specific language governing permissions and
// limitations under the License.
#include "core/general-server/op/general_infer_op.h"
#include <algorithm>
#include <iostream>
#include <memory>
#include <sstream>
#include "core/general-server/op/general_infer_op.h"
#include "core/predictor/framework/infer.h"
#include "core/predictor/framework/memory.h"
#include "core/predictor/framework/resource.h"
...
...
@@ -36,33 +36,31 @@ using baidu::paddle_serving::predictor::InferManager;
using
baidu
::
paddle_serving
::
predictor
::
PaddleGeneralModelConfig
;
int
GeneralInferOp
::
inference
()
{
const
GeneralBlob
*
input_blob
=
get_depend_argument
<
GeneralBlob
>
(
pre_name
());
const
GeneralBlob
*
input_blob
=
get_depend_argument
<
GeneralBlob
>
(
pre_name
());
GeneralBlob
*
output_blob
=
mutable_data
<
GeneralBlob
>
();
GeneralBlob
*
output_blob
=
mutable_data
<
GeneralBlob
>
();
if
(
!
input_blob
)
{
LOG
(
ERROR
)
<<
"Failed mutable depended argument, op:"
<<
pre_name
();
LOG
(
ERROR
)
<<
"Failed mutable depended argument, op:"
<<
pre_name
();
return
-
1
;
}
const
TensorVector
*
in
=
&
input_blob
->
tensor_vector
;
TensorVector
*
out
=
&
output_blob
->
tensor_vector
;
int
batch_size
=
input_blob
->
GetBatchSize
();
output_blob
->
SetBatchSize
(
batch_size
);
VLOG
(
2
)
<<
"infer batch size: "
<<
batch_size
;
// infer
Timer
timeline
;
double
infer_time
=
0.0
;
timeline
.
Start
();
//
Timer timeline;
//
double infer_time = 0.0;
//
timeline.Start();
if
(
InferManager
::
instance
().
infer
(
GENERAL_MODEL_NAME
,
in
,
out
,
batch_size
))
{
LOG
(
ERROR
)
<<
"Failed do infer in fluid model: "
<<
GENERAL_MODEL_NAME
;
return
-
1
;
}
timeline
.
Pause
();
infer_time
=
timeline
.
ElapsedUS
();
// timeline.Pause();
// infer_time = timeline.ElapsedUS();
return
0
;
}
DEFINE_OP
(
GeneralInferOp
);
...
...
core/general-server/op/general_reader_op.cpp
浏览文件 @
d986670e
...
...
@@ -12,12 +12,12 @@
// See the License for the specific language governing permissions and
// limitations under the License.
#include "core/general-server/op/general_reader_op.h"
#include <algorithm>
#include <iostream>
#include <memory>
#include <sstream>
#include "core/general-server/op/general_infer_helper.h"
#include "core/general-server/op/general_reader_op.h"
#include "core/predictor/framework/infer.h"
#include "core/predictor/framework/memory.h"
...
...
@@ -73,7 +73,6 @@ int GeneralReaderOp::inference() {
int
batch_size
=
req
->
insts_size
();
int
input_var_num
=
0
;
std
::
vector
<
int64_t
>
elem_type
;
std
::
vector
<
int64_t
>
elem_size
;
std
::
vector
<
int64_t
>
capacity
;
...
...
@@ -81,6 +80,8 @@ int GeneralReaderOp::inference() {
GeneralBlob
*
res
=
mutable_data
<
GeneralBlob
>
();
TensorVector
*
out
=
&
res
->
tensor_vector
;
res
->
SetBatchSize
(
batch_size
);
if
(
!
res
)
{
LOG
(
ERROR
)
<<
"Failed get op tls reader object output"
;
}
...
...
core/general-server/op/general_response_op.cpp
浏览文件 @
d986670e
...
...
@@ -12,12 +12,12 @@
// See the License for the specific language governing permissions and
// limitations under the License.
#include "core/general-server/op/general_response_op.h"
#include <algorithm>
#include <iostream>
#include <memory>
#include <sstream>
#include "core/general-server/op/general_infer_helper.h"
#include "core/general-server/op/general_response_op.h"
#include "core/predictor/framework/infer.h"
#include "core/predictor/framework/memory.h"
#include "core/predictor/framework/resource.h"
...
...
@@ -37,12 +37,10 @@ using baidu::paddle_serving::predictor::InferManager;
using
baidu
::
paddle_serving
::
predictor
::
PaddleGeneralModelConfig
;
int
GeneralResponseOp
::
inference
()
{
const
GeneralBlob
*
input_blob
=
get_depend_argument
<
GeneralBlob
>
(
pre_name
());
const
GeneralBlob
*
input_blob
=
get_depend_argument
<
GeneralBlob
>
(
pre_name
());
if
(
!
input_blob
)
{
LOG
(
ERROR
)
<<
"Failed mutable depended argument, op: "
<<
pre_name
();
LOG
(
ERROR
)
<<
"Failed mutable depended argument, op: "
<<
pre_name
();
return
-
1
;
}
...
...
@@ -56,7 +54,7 @@ int GeneralResponseOp::inference() {
VLOG
(
2
)
<<
"start to call load general model_conf op"
;
baidu
::
paddle_serving
::
predictor
::
Resource
&
resource
=
baidu
::
paddle_serving
::
predictor
::
Resource
::
instance
();
VLOG
(
2
)
<<
"get resource pointer done."
;
std
::
shared_ptr
<
PaddleGeneralModelConfig
>
model_config
=
resource
.
get_general_model_config
();
...
...
@@ -67,7 +65,7 @@ int GeneralResponseOp::inference() {
fetch_index
[
i
]
=
model_config
->
_fetch_alias_name_to_index
[
req
->
fetch_var_names
(
i
)];
}
// response inst with only fetch_var_names
Response
*
res
=
mutable_data
<
Response
>
();
...
...
@@ -75,7 +73,7 @@ int GeneralResponseOp::inference() {
for
(
int
i
=
0
;
i
<
batch_size
;
++
i
)
{
FetchInst
*
fetch_inst
=
res
->
add_insts
();
for
(
auto
&
idx
:
fetch_index
)
{
for
(
auto
&
idx
:
fetch_index
)
{
Tensor
*
tensor
=
fetch_inst
->
add_tensor_array
();
// currently only response float tensor or lod_tensor
tensor
->
set_elem_type
(
1
);
...
...
@@ -85,8 +83,7 @@ int GeneralResponseOp::inference() {
}
else
{
VLOG
(
2
)
<<
"out["
<<
idx
<<
"] is tensor"
;
for
(
int
k
=
1
;
k
<
in
->
at
(
idx
).
shape
.
size
();
++
k
)
{
VLOG
(
2
)
<<
"shape["
<<
k
-
1
<<
"]: "
<<
in
->
at
(
idx
).
shape
[
k
];
VLOG
(
2
)
<<
"shape["
<<
k
-
1
<<
"]: "
<<
in
->
at
(
idx
).
shape
[
k
];
tensor
->
add_shape
(
in
->
at
(
idx
).
shape
[
k
]);
}
}
...
...
@@ -94,7 +91,7 @@ int GeneralResponseOp::inference() {
}
int
var_idx
=
0
;
for
(
auto
&
idx
:
fetch_index
)
{
for
(
auto
&
idx
:
fetch_index
)
{
float
*
data_ptr
=
static_cast
<
float
*>
(
in
->
at
(
idx
).
data
.
data
());
int
cap
=
1
;
for
(
int
j
=
1
;
j
<
in
->
at
(
idx
).
shape
.
size
();
++
j
)
{
...
...
@@ -102,8 +99,8 @@ int GeneralResponseOp::inference() {
}
if
(
model_config
->
_is_lod_fetch
[
idx
])
{
for
(
int
j
=
0
;
j
<
batch_size
;
++
j
)
{
for
(
int
k
=
in
->
at
(
idx
).
lod
[
0
][
j
];
k
<
in
->
at
(
idx
).
lod
[
0
][
j
+
1
];
k
++
)
{
for
(
int
k
=
in
->
at
(
idx
).
lod
[
0
][
j
];
k
<
in
->
at
(
idx
).
lod
[
0
][
j
+
1
];
k
++
)
{
res
->
mutable_insts
(
j
)
->
mutable_tensor_array
(
var_idx
)
->
add_data
(
reinterpret_cast
<
char
*>
(
&
(
data_ptr
[
k
])),
sizeof
(
float
));
}
...
...
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