Skip to content

D
docs

MindSpore / docs

通知 4
Star 2
Fork 2
D
docs
提交 4e77628c 编写于 作者: L liuxiao78
浏览文件
操作

revise on_device_inference

上级 daf73f94
隐藏空白更改
内联 并排
Showing with 90 addition and 312 deletion
tutorials/source_en/advanced_use/on_device_inference.md
浏览文件 @ 4e77628c
......@@ -29,9 +29,10 @@ Mindspore Lite's framework consists of frontend, IR, backend, Lite, RT and micro
- Lite RT: In the inference runtime, session provides the external interface, kernel registry is operator registry, scheduler is operator heterogeneous scheduler and executor is operator executor. Lite RT and shares with Micro the underlying infrastructure layers such as operator library, memory allocation, runtime thread pool and parallel primitives.
- Micro: Code Gen generates .c files according to the model, and infrastructure such as the underlying operator library is shared with Lite RT.
## Compilation Method
You need to compile the MindSpore Predict by yourself. This section describes how to perform cross compilation in the Ubuntu environment.
You need to compile the MindSpore Lite by yourself. This section describes how to perform cross compilation in the Ubuntu environment.
The environment requirements are as follows:
......@@ -40,8 +41,8 @@ The environment requirements are as follows:
- Hard disk space: 10 GB or above
- System requirements
- System: Ubuntu = 18.04.02LTS (availability is checked)
- Kernel: 4.15.0-45-generic (availability is checked)
- System is limited on Linux
- Recommend system: Ubuntu = 18.04.02LTS
- Software dependencies
- [cmake](https://cmake.org/download/) >= 3.14.1
......@@ -53,13 +54,9 @@ The environment requirements are as follows:
- decorator
- scipy
> `numpy`, `decorator` and `scipy` can be installed through `pip`. The reference command is as following.
> `numpy decorator scipy` can be installed through `pip`. The reference command is as: `pip3 install numpy==1.16 decorator scipy`.
```bash
pip3 install numpy==1.16 decorator scipy
```
The compilation procedure is as follows:
1. Configure environment variables.
......@@ -75,18 +72,17 @@ The compilation procedure is as follows:
git clone https://gitee.com/mindspore/mindspore.git
```
3. Run the following command in the root directory of the source code to compile MindSpore Predict: -I indicates options for compiling MindSpore Predict and the parameter is the target platform architecture. Currently, only the Android arm64 platform is supported.
3. Run the following command in the root directory of the source code to compile MindSpore Lite.
```bash
sh build.sh -I arm64
cd mindspore/lite
sh build.sh
```
4. Obtain the compilation result.
Go to the `predict/output` directory of the source code to view the generated package. The package name is MSPredict-*Version number*-*Host platform*_*Device platform*.tar.gz, for example, MSPredict-0.1.0-linux_aarch64.tar.gz. The package contains the following directories:
- include: MindSpore Predict header file.
- lib: MindSpore Predict dynamic library.
Go to the `lite/build` directory of the source code to view the generated documents. Then you can use various tools after changing directory.
## Use of On-Device Inference
......@@ -171,162 +167,54 @@ Use the `.ms` model file and image data as input to create a session and impleme
![](./images/side_infer_process.png)
Figure 2 On-device inference sequence diagram
1. Load the `.ms` model file to the memory buffer. The ReadFile function needs to be implemented by users, according to the [C++ tutorial](http://www.cplusplus.com/doc/tutorial/files/).
```cpp
// read model file
std::string modelPath = "./models/lenet/lenet.ms";
size_t graphSize = 0;
/* ReadFile() here is a dummy function */
char *graphBuf = ReadFile(modelPath.c_str(), graphSize);
// Read Model File
std::string model_path = "./lenet.ms";
ReadFile(model_path.c_str(), &model_size, buf);
// Import Model
auto model = lite::Model::Import(content, size);
meta_graph.reset();
content = nullptr;
auto context = new lite::Context;
context->cpuBindMode = lite::NO_BIND;
context->deviceCtx.type = lite::DT_CPU;
context->threadNum = 4;
```
2. Call the CreateSession API to create a session. After the session is created, the model file in the memory buffer can be released.
2. Call the `CreateSession` API to get a session.
```cpp
// create session
Context ctx;
std::shared_ptr<Session> session = CreateSession(graphBuf, graphSize, ctx);
free(graphBuf);
// Create Session
auto session = session::LiteSession::CreateSession(context);
ASSERT_NE(nullptr, session);
```
3. Read the input data for inference from the memory buffer and call the `SetData` API to set the input data to `input tensor`.
3. Call the `CompileGraph` API in previous `Session` and transport model.
```cpp
// load input buffer
size_t inputSize = 0;
std::string imagePath = "./data/input/lenet.bin";
char *inputBuf = ReadFile(imagePath.c_str(), inputSize);
//get input tensors
std::vector<Tensor *> inputs = session->GetInput();
//set input buffer
inputs[0]->SetData(inputBuf);
// Compile Graph
auto ret = session->CompileGraph(model.get());
ASSERT_EQ(lite::RET_OK, ret);
```
4. Call the `Run` API in the `session` to perform inference.
4. Call the `GetInputs` API to get input `tensor`, then set graph information as `data`, `data` will be used in to perform model inference.
```cpp
// session run
int ret = session->Run(inputs);
auto inputs = session->GetInputs();
ASSERT_EQ(inputs.size(), 1);
auto inTensor = inputs.front();
ASSERT_NE(nullptr, inTensor);
(void)inTensor->MutableData();
```
5. Call the `GetAllOutput` API to obtain the output.
```cpp
// get output
std::map<std::string, std::vector<Tensor *>> outputs = session->GetAllOutput();
```
6. Call the `Getdata` API to get the output data.
5. Call the `RunGraph` API in the `Session` to perform inference.
```cpp
// get output data
float *data = nullptr;
for (auto output : outputs) {
auto tensors = output.second;
for (auto tensor : tensors) {
data = (float *)(tensor->GetData());
}
}
// Run Graph
ret = session->RunGraph();
ASSERT_EQ(lite::RET_OK, ret);
```
7. Release input and output tensors after the inference is complete.
6. Call the `GetOutputs` API to obtain the output.
```cpp
// free inputs and outputs
for (auto &input : inputs) {
delete input;
}
inputs.clear();
for (auto &output : outputs) {
for (auto &outputTensor : output.second) {
delete outputTensor;
}
}
outputs.clear();
// Get Outputs
auto outputs = session->GetOutputs();
```
Select the LeNet network and set the inference input to `lenet.bin`. The complete sample code `lenet.cpp` is as follows:
> MindSpore Predict uses `FlatBuffers` to define models. The `FlatBuffers` header file is required for parsing models. Therefore, you need to configure the `FlatBuffers` header file.
>
> Method: Copy the `flatbuffers` folder in MindSpore root directory`/third_party/flatbuffers/include` to the directory at the same level as `session.h`.
```cpp
#include <string>
#include <vector>
#include "context.h"
#include "session.h"
#include "tensor.h"
#include "errorcode.h"
using namespace mindspore::predict;
int main() {
std::string modelPath = "./models/lenet/lenet.ms";
std::string imagePath = "./data/input/lenet.bin";
// read model file
size_t graphSize = 0;
/* ReadFile() here is a dummy function */
char *graphBuf = ReadFile(modelPath.c_str(), graphSize);
if (graphBuf == nullptr) {
return -1;
}
// create session
Context ctx;
auto session = CreateSession(graphBuf, graphSize, ctx);
if (session == nullptr) {
free(graphBuf);
return -1;
}
free(graphBuf);
// load input buf
size_t inputSize = 0;
char *inputBuf = ReadFile(imagePath.c_str(), inputSize);
if (inputBuf == nullptr) {
return -1;
}
auto inputs = session->GetInput();
inputs[0]->SetData(inputBuf);
// session run
auto ret = session->Run(inputs);
if (ret != RET_OK) {
printf("run failed, error: %d\n", ret);
for (auto &input : inputs) {
delete input;
}
return -1;
}
// get output
auto outputs = session->GetAllOutput();
// get output data
float *data = nullptr;
for (auto output : outputs) {
auto tensors = output.second;
for (auto tensor : tensors) {
data = (float *)(tensor->GetData());
//print the contents of the data
for (size_t i = 0; i < tensor->GetElementSize(); ++i) {
printf(" %f ", data[i]);
}
printf("\n");
}
}
// free inputs and outputs
for (auto &input : inputs) {
delete input;
}
inputs.clear();
for (auto &output : outputs) {
for (auto &outputTensor : output.second) {
delete outputTensor;
}
}
outputs.clear();
return 0;
}
```
\ No newline at end of file
tutorials/source_zh_cn/advanced_use/on_device_inference.md
浏览文件 @ 4e77628c
......@@ -41,8 +41,8 @@ MindSpore Lite的框架主要由Frontend、IR、Backend、Lite RT、Micro构成
- 硬盘空间10GB以上
- 系统要求
- 系统：Ubuntu = 18.04.02LTS（验证可用）
- 内核：4.15.0-45-generic（验证可用）
- 系统环境仅支持Linux
- 推荐系统：Ubuntu = 18.04.02LTS
- 软件依赖
- [cmake](https://cmake.org/download/) >= 3.14.1
......@@ -54,18 +54,15 @@ MindSpore Lite的框架主要由Frontend、IR、Backend、Lite RT、Micro构成
- decorator
- scipy
> `numpy`、 `decorator`和`scipy`可以通过`pip`安装，参考命令如下。
> `numpy decorator scipy`可以通过`pip`安装，参考命令：`pip3 install numpy==1.16 decorator scipy`。
```bash
pip3 install numpy==1.16 decorator scipy
```
编译步骤如下：
1. 配置环境变量。
```bash
export LLVM_PATH={$LLVM_PATH}/clang+llvm-8.0.0-x86_64-linux-gnu-ubuntu-16.04/bin/llvm-config #设定llvm路径
export LLVM_PATH={$LLVM_PATH}/clang+llvm-8.0.0-x86_64-linux-gnu-ubuntu-18.04/bin/llvm-config #设定llvm路径
export ANDROID_NDK={$NDK_PATH}/android-ndk-r16b #设定ndk路径
```
......@@ -75,18 +72,16 @@ MindSpore Lite的框架主要由Frontend、IR、Backend、Lite RT、Micro构成
git clone https://gitee.com/mindspore/mindspore.git
```
3. 在源码根目录下，执行如下命令编译MindSpore Predict。-I为编译MindSpore Predict的编译参数，-I的参数为目标端侧平台，目前仅支持安卓arm64平台。
3. 在源码根目录下，执行如下命令编译MindSpore Lite。
```bash
sh build.sh -I arm64
cd mindspore/lite
sh build.sh
```
4. 获取编译结果。
进入源码的`predict/output`目录，即可查看生成的压缩包，包名为MSPredict-*版本号*-*HOST平台*_*DEVICE平台*.tar.gz，例如：MSPredict-0.1.0-linux_aarch64.tar.gz。 该压缩包包含以下目录：
- include：MindSpore Predict的头文件。
- lib：MindSpore Predict的动态库。
进入源码的`lite/build`目录，可查看编译后生成的文件。进入相对应的文件夹下执行命令，就可以使用MindSpore Lite的多种功能。
## 端侧推理使用
......@@ -172,161 +167,56 @@ if __name__ == '__main__':
![](./images/side_infer_process.png)
图2：端侧推理时序图
1. 加载`.ms`模型文件到内存缓冲区，ReadFile函数功能需要用户参考[C++教程](http://www.cplusplus.com/doc/tutorial/files/)自行实现。
```cpp
// read model file
std::string modelPath = "./models/lenet/lenet.ms";
size_t graphSize = 0;
/* ReadFile() here is a dummy function */
char *graphBuf = ReadFile(modelPath.c_str(), graphSize);
1. 读取MindSpore端侧模型文件信息。ReadFile函数功能需要用户参考[C++教程](http://www.cplusplus.com/doc/tutorial/files/)自行实现。
```cpp
// Read Model File
std::string model_path = "./lenet.ms";
ReadFile(model_path.c_str(), &model_size, buf);
// Import Model
auto model = lite::Model::Import(content, size);
meta_graph.reset();
content = nullptr;
auto context = new lite::Context;
context->cpuBindMode = lite::NO_BIND;
context->deviceCtx.type = lite::DT_CPU;
context->threadNum = 4;
```
2. 调用`CreateSession`接口创建`Session`，创建完成后可释放内存缓冲区中的模型文件。
2. 调用`CreateSession`接口创建`Session`。
```cpp
// create session
Context ctx;
std::shared_ptr<Session> session = CreateSession(graphBuf, graphSize, ctx);
free(graphBuf);
// Create Session
auto session = session::LiteSession::CreateSession(context);
ASSERT_NE(nullptr, session);
```
3. 从内存缓冲区中读取推理的输入数据，调用`SetData`接口将输入数据设置到`input tensor`中。
3. 调用`Session`中的`CompileGraph`方法，传入模型。
```cpp
// load input buffer
size_t inputSize = 0;
std::string imagePath = "./data/input/lenet.bin";
char *inputBuf = ReadFile(imagePath.c_str(), inputSize);
//get input tensors
std::vector<Tensor *> inputs = session->GetInput();
//set input buffer
inputs[0]->SetData(inputBuf);
// Compile Graph
auto ret = session->CompileGraph(model.get());
ASSERT_EQ(lite::RET_OK, ret);
```
4. 调用`Session`中的`Run`接口执行推理。
4. 调用`Session`中的`GetInputs`方法，获取输入`Tensor`，获取图片信息设置为`data`，`data`即为用于推理的输入数据。
```cpp
// session run
int ret = session->Run(inputs);
auto inputs = session->GetInputs();
ASSERT_EQ(inputs.size(), 1);
auto inTensor = inputs.front();
ASSERT_NE(nullptr, inTensor);
(void)inTensor->MutableData();
```
5. 调用`GetAllOutput`接口获取输出。
5. 调用`Session`中的`RunGraph`接口执行推理。
```cpp
// get output
std::map<std::string, std::vector<Tensor *>> outputs = session->GetAllOutput();
// Run Graph
ret = session->RunGraph();
ASSERT_EQ(lite::RET_OK, ret);
```
6. 调用`Tensor`的`GetData`接口获取输出数据。
6. 调用`GetOutputs`接口获取输出。
```cpp
// get output data
float *data = nullptr;
for (auto output : outputs) {
auto tensors = output.second;
for (auto tensor : tensors) {
data = (float *)(tensor->GetData());
}
}
// Get Outputs
auto outputs = session->GetOutputs();
```
7. 推理结束释放`input tensor`和`output tensor`。
```cpp
// free inputs and outputs
for (auto &input : inputs) {
delete input;
}
inputs.clear();
for (auto &output : outputs) {
for (auto &outputTensor : output.second) {
delete outputTensor;
}
}
outputs.clear();
```
选取LeNet网络，推理输入为“`lenet.bin`”，完整示例代码`lenet.cpp`如下。
> MindSpore Predict使用`FlatBuffers`定义模型，解析模型需要使用到`FlatBuffers`头文件，因此用户需要自行配置`FlatBuffers`头文件。
>
> 具体做法：将MindSpore根目录`/third_party/flatbuffers/include`下的`flatbuffers`文件夹拷贝到`session.h`的同级目录。
```cpp
#include <string>
#include <vector>
#include "context.h"
#include "session.h"
#include "tensor.h"
#include "errorcode.h"
using namespace mindspore::predict;
int main() {
std::string modelPath = "./models/lenet/lenet.ms";
std::string imagePath = "./data/input/lenet.bin";
// read model file
size_t graphSize = 0;
/* ReadFile() here is a dummy function */
char *graphBuf = ReadFile(modelPath.c_str(), graphSize);
if (graphBuf == nullptr) {
return -1;
}
// create session
Context ctx;
auto session = CreateSession(graphBuf, graphSize, ctx);
if (session == nullptr) {
free(graphBuf);
return -1;
}
free(graphBuf);
// load input buf
size_t inputSize = 0;
char *inputBuf = ReadFile(imagePath.c_str(), inputSize);
if (inputBuf == nullptr) {
return -1;
}
auto inputs = session->GetInput();
inputs[0]->SetData(inputBuf);
// session run
auto ret = session->Run(inputs);
if (ret != RET_OK) {
printf("run failed, error: %d\n", ret);
for (auto &input : inputs) {
delete input;
}
return -1;
}
// get output
auto outputs = session->GetAllOutput();
// get output data
float *data = nullptr;
for (auto output : outputs) {
auto tensors = output.second;
for (auto tensor : tensors) {
data = (float *)(tensor->GetData());
//print the contents of the data
for (size_t i = 0; i < tensor->GetElementSize(); ++i) {
printf(" %f ", data[i]);
}
printf("\n");
}
}
// free inputs and outputs
for (auto &input : inputs) {
delete input;
}
inputs.clear();
for (auto &output : outputs) {
for (auto &outputTensor : output.second) {
delete outputTensor;
}
}
outputs.clear();
return 0;
}
```
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或