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
csdn_franckjun
Paddle
提交
4b59646e
P
Paddle
项目概览
csdn_franckjun
/
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
4b59646e
编写于
2月 27, 2019
作者:
N
nhzlx
提交者:
ceci3
3月 08, 2019
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
fix comments and fix cpplint
test=develop
上级
5863c861
变更
13
隐藏空白更改
内联
并排
Showing
13 changed file
with
28 addition
and
6 deletion
+28
-6
paddle/fluid/framework/ir/fuse_pass_base.h
paddle/fluid/framework/ir/fuse_pass_base.h
+1
-1
paddle/fluid/inference/analysis/helper.h
paddle/fluid/inference/analysis/helper.h
+2
-0
paddle/fluid/inference/analysis/ir_pass_manager.h
paddle/fluid/inference/analysis/ir_pass_manager.h
+3
-0
paddle/fluid/inference/analysis/ir_passes/tensorrt_subgraph_pass.cc
...id/inference/analysis/ir_passes/tensorrt_subgraph_pass.cc
+1
-1
paddle/fluid/inference/analysis/ir_passes/tensorrt_subgraph_pass.h
...uid/inference/analysis/ir_passes/tensorrt_subgraph_pass.h
+4
-1
paddle/fluid/inference/api/analysis_predictor.h
paddle/fluid/inference/api/analysis_predictor.h
+1
-0
paddle/fluid/inference/tensorrt/convert/op_converter.h
paddle/fluid/inference/tensorrt/convert/op_converter.h
+1
-0
paddle/fluid/inference/tensorrt/convert/ut_helper.h
paddle/fluid/inference/tensorrt/convert/ut_helper.h
+2
-0
paddle/fluid/inference/tensorrt/plugin/split_op_plugin.h
paddle/fluid/inference/tensorrt/plugin/split_op_plugin.h
+1
-0
paddle/fluid/inference/tensorrt/plugin/trt_plugin.h
paddle/fluid/inference/tensorrt/plugin/trt_plugin.h
+1
-0
paddle/fluid/inference/tensorrt/plugin/trt_plugin_factory.h
paddle/fluid/inference/tensorrt/plugin/trt_plugin_factory.h
+1
-0
paddle/fluid/inference/tensorrt/test_engine.cc
paddle/fluid/inference/tensorrt/test_engine.cc
+6
-1
paddle/fluid/operators/tensorrt/tensorrt_engine_op.h
paddle/fluid/operators/tensorrt/tensorrt_engine_op.h
+4
-2
未找到文件。
paddle/fluid/framework/ir/fuse_pass_base.h
浏览文件 @
4b59646e
...
@@ -25,7 +25,7 @@ namespace ir {
...
@@ -25,7 +25,7 @@ namespace ir {
static
const
char
kParamScopeAttr
[]
=
"__param_scope__"
;
static
const
char
kParamScopeAttr
[]
=
"__param_scope__"
;
static
const
char
kFuseStatisAttr
[]
=
"__fuse_statis__"
;
static
const
char
kFuseStatisAttr
[]
=
"__fuse_statis__"
;
// When we use trt or other third_party lib, the parameters are manage
re
d by
// When we use trt or other third_party lib, the parameters are managed by
// the lib, but not the fluid. So we need to record them to avoid duplicate
// the lib, but not the fluid. So we need to record them to avoid duplicate
// allocation.
// allocation.
static
const
char
kRepetitiveParamAttr
[]
=
"__repetitive_param__"
;
static
const
char
kRepetitiveParamAttr
[]
=
"__repetitive_param__"
;
...
...
paddle/fluid/inference/analysis/helper.h
浏览文件 @
4b59646e
...
@@ -17,10 +17,12 @@ limitations under the License. */
...
@@ -17,10 +17,12 @@ limitations under the License. */
#include <sys/stat.h>
#include <sys/stat.h>
#include <cstdio>
#include <cstdio>
#include <fstream>
#include <fstream>
#include <memory>
#include <set>
#include <set>
#include <string>
#include <string>
#include <typeindex>
#include <typeindex>
#include <unordered_map>
#include <unordered_map>
#include <utility>
#include <vector>
#include <vector>
#include "paddle/fluid/framework/framework.pb.h"
#include "paddle/fluid/framework/framework.pb.h"
...
...
paddle/fluid/inference/analysis/ir_pass_manager.h
浏览文件 @
4b59646e
...
@@ -22,7 +22,10 @@
...
@@ -22,7 +22,10 @@
#pragma once
#pragma once
#include <memory>
#include <string>
#include <string>
#include <unordered_set>
#include <utility>
#include <vector>
#include <vector>
#include "paddle/fluid/framework/ir/graph.h"
#include "paddle/fluid/framework/ir/graph.h"
#include "paddle/fluid/framework/ir/pass.h"
#include "paddle/fluid/framework/ir/pass.h"
...
...
paddle/fluid/inference/analysis/ir_passes/tensorrt_subgraph_pass.cc
浏览文件 @
4b59646e
...
@@ -235,7 +235,7 @@ void TensorRtSubgraphPass::CreateTensorRTOp(
...
@@ -235,7 +235,7 @@ void TensorRtSubgraphPass::CreateTensorRTOp(
std
::
string
trt_engine_serialized_data
=
GetTrtEngineSerializedData
(
std
::
string
trt_engine_serialized_data
=
GetTrtEngineSerializedData
(
Get
<
std
::
string
>
(
"model_opt_cache_dir"
),
engine_key
);
Get
<
std
::
string
>
(
"model_opt_cache_dir"
),
engine_key
);
if
(
trt_engine_serialized_data
.
size
()
==
0
)
{
if
(
trt_engine_serialized_data
.
empty
()
)
{
LOG
(
INFO
)
<<
"Prepare TRT engine (Optimize model structure, Select OP "
LOG
(
INFO
)
<<
"Prepare TRT engine (Optimize model structure, Select OP "
"kernel etc). This process may cost a lot of time."
;
"kernel etc). This process may cost a lot of time."
;
std
::
unique_ptr
<
tensorrt
::
TensorRTEngine
>
trt_engine
(
std
::
unique_ptr
<
tensorrt
::
TensorRTEngine
>
trt_engine
(
...
...
paddle/fluid/inference/analysis/ir_passes/tensorrt_subgraph_pass.h
浏览文件 @
4b59646e
...
@@ -13,9 +13,12 @@
...
@@ -13,9 +13,12 @@
// limitations under the License.
// limitations under the License.
#pragma once
#pragma once
#include <
paddle/fluid/framework/ir/fuse_pass_base.h
>
#include <
memory
>
#include <string>
#include <string>
#include <unordered_map>
#include <unordered_set>
#include <vector>
#include <vector>
#include "paddle/fluid/framework/ir/fuse_pass_base.h"
#include "paddle/fluid/framework/ir/pass.h"
#include "paddle/fluid/framework/ir/pass.h"
namespace
paddle
{
namespace
paddle
{
...
...
paddle/fluid/inference/api/analysis_predictor.h
浏览文件 @
4b59646e
...
@@ -15,6 +15,7 @@
...
@@ -15,6 +15,7 @@
#pragma once
#pragma once
#include <algorithm>
#include <algorithm>
#include <map>
#include <map>
#include <memory>
#include <string>
#include <string>
#include <vector>
#include <vector>
#include "paddle/fluid/framework/naive_executor.h"
#include "paddle/fluid/framework/naive_executor.h"
...
...
paddle/fluid/inference/tensorrt/convert/op_converter.h
浏览文件 @
4b59646e
...
@@ -16,6 +16,7 @@ limitations under the License. */
...
@@ -16,6 +16,7 @@ limitations under the License. */
#include <string>
#include <string>
#include <unordered_map>
#include <unordered_map>
#include <unordered_set>
#include <vector>
#include <vector>
#include "paddle/fluid/framework/block_desc.h"
#include "paddle/fluid/framework/block_desc.h"
#include "paddle/fluid/framework/op_registry.h"
#include "paddle/fluid/framework/op_registry.h"
...
...
paddle/fluid/inference/tensorrt/convert/ut_helper.h
浏览文件 @
4b59646e
...
@@ -19,7 +19,9 @@ limitations under the License. */
...
@@ -19,7 +19,9 @@ limitations under the License. */
#pragma once
#pragma once
#include <memory>
#include <string>
#include <string>
#include <unordered_set>
#include <vector>
#include <vector>
#include "paddle/fluid/framework/lod_tensor.h"
#include "paddle/fluid/framework/lod_tensor.h"
...
...
paddle/fluid/inference/tensorrt/plugin/split_op_plugin.h
浏览文件 @
4b59646e
...
@@ -15,6 +15,7 @@
...
@@ -15,6 +15,7 @@
#pragma once
#pragma once
#include <thrust/device_vector.h>
#include <thrust/device_vector.h>
#include <utility>
#include <vector>
#include <vector>
#include "paddle/fluid/inference/tensorrt/plugin/trt_plugin.h"
#include "paddle/fluid/inference/tensorrt/plugin/trt_plugin.h"
...
...
paddle/fluid/inference/tensorrt/plugin/trt_plugin.h
浏览文件 @
4b59646e
...
@@ -17,6 +17,7 @@
...
@@ -17,6 +17,7 @@
#include <NvInfer.h>
#include <NvInfer.h>
#include <cstring>
#include <cstring>
#include <unordered_map>
#include <unordered_map>
#include <utility>
#include <vector>
#include <vector>
#include "paddle/fluid/inference/tensorrt/plugin/trt_plugin_utils.h"
#include "paddle/fluid/inference/tensorrt/plugin/trt_plugin_utils.h"
...
...
paddle/fluid/inference/tensorrt/plugin/trt_plugin_factory.h
浏览文件 @
4b59646e
...
@@ -17,6 +17,7 @@
...
@@ -17,6 +17,7 @@
#include <NvInfer.h>
#include <NvInfer.h>
#include <cstring>
#include <cstring>
#include <list>
#include <list>
#include <memory>
#include <string>
#include <string>
#include <unordered_map>
#include <unordered_map>
#include <vector>
#include <vector>
...
...
paddle/fluid/inference/tensorrt/test_engine.cc
浏览文件 @
4b59646e
...
@@ -35,7 +35,12 @@ class TensorRTEngineTest : public ::testing::Test {
...
@@ -35,7 +35,12 @@ class TensorRTEngineTest : public ::testing::Test {
engine_
->
InitNetwork
();
engine_
->
InitNetwork
();
}
}
void
TearDown
()
override
{
delete
engine_
;
}
void
TearDown
()
override
{
if
(
engine_
)
{
delete
engine_
;
engine_
=
nullptr
;
}
}
void
PrepareInputOutput
(
const
std
::
vector
<
float
>
&
input
,
void
PrepareInputOutput
(
const
std
::
vector
<
float
>
&
input
,
std
::
vector
<
int
>
output_shape
)
{
std
::
vector
<
int
>
output_shape
)
{
...
...
paddle/fluid/operators/tensorrt/tensorrt_engine_op.h
浏览文件 @
4b59646e
...
@@ -16,8 +16,10 @@
...
@@ -16,8 +16,10 @@
#ifdef PADDLE_WITH_CUDA
#ifdef PADDLE_WITH_CUDA
#include <memory>
#include <string>
#include <string>
#include <unordered_map>
#include <unordered_map>
#include <unordered_set>
#include <vector>
#include <vector>
#include "paddle/fluid/framework/executor.h"
#include "paddle/fluid/framework/executor.h"
...
@@ -220,11 +222,11 @@ class TensorRTEngineOp : public framework::OperatorBase {
...
@@ -220,11 +222,11 @@ class TensorRTEngineOp : public framework::OperatorBase {
TensorRTEngine
*
GetEngine
(
const
framework
::
Scope
&
scope
,
TensorRTEngine
*
GetEngine
(
const
framework
::
Scope
&
scope
,
const
platform
::
Place
&
dev_place
)
const
{
const
platform
::
Place
&
dev_place
)
const
{
if
(
trt_engine_
.
get
()
==
nullptr
)
{
if
(
!
trt_engine_
)
{
trt_engine_
.
reset
(
new
inference
::
tensorrt
::
TensorRTEngine
(
trt_engine_
.
reset
(
new
inference
::
tensorrt
::
TensorRTEngine
(
max_batch_size_
,
workspace_size_
,
enable_int8_
,
calibrator_
.
get
(),
max_batch_size_
,
workspace_size_
,
enable_int8_
,
calibrator_
.
get
(),
boost
::
get
<
platform
::
CUDAPlace
>
(
dev_place
).
device
));
boost
::
get
<
platform
::
CUDAPlace
>
(
dev_place
).
device
));
if
(
engine_serialized_data_
.
size
()
>
0
)
{
if
(
!
engine_serialized_data_
.
empty
()
)
{
trt_engine_
->
Deserialize
(
engine_serialized_data_
);
trt_engine_
->
Deserialize
(
engine_serialized_data_
);
}
else
{
}
else
{
PrepareTRTEngine
(
scope
,
trt_engine_
.
get
());
PrepareTRTEngine
(
scope
,
trt_engine_
.
get
());
...
...
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