apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-05 08:37:06 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Doc] Change all PaddleLite or Paddle-Lite to Paddle Lite (#929)

Browse Source 
* [FlyCV] Bump up FlyCV -> official release 1.0.0

* change PaddleLite or Paddle-Lite to Paddle lite

* fix docs

* fix doc

Co-authored-by: DefTruth <qiustudent_r@163.com>
Co-authored-by: DefTruth <31974251+DefTruth@users.noreply.github.com>
This commit is contained in:
yeliang2258
2022-12-21 14:15:50 +08:00
committed by GitHub
parent 725fe52df3
commit b42ec302e6
21 changed files with 104 additions and 86 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
docs/cn/build_and_install/a311d.md
View File

@@ -1,6 +1,6 @@
# 晶晨 A311D 部署环境编译安装
FastDeploy 基于 Paddle-Lite 后端支持在晶晨 NPU 上进行部署推理。
FastDeploy 基于 Paddle Lite 后端支持在晶晨 NPU 上进行部署推理。
更多详细的信息请参考：[Paddle Lite部署示例](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/verisilicon_timvx.html)。
本文档介绍如何编译基于 Paddle Lite 的 C++ FastDeploy 交叉编译库。
@@ -67,7 +67,7 @@ cmake -DCMAKE_TOOLCHAIN_FILE=./../cmake/toolchain.cmake \
make -j8
make install
```
编译完成之后，会生成 fastdeploy-tmivx 目录，表示基于 PadddleLite TIM-VX 的 FastDeploy 库编译完成。
编译完成之后，会生成 fastdeploy-tmivx 目录，表示基于 Paddle Lite TIM-VX 的 FastDeploy 库编译完成。
## 准备设备运行环境
部署前要保证晶晨 Linux Kernel NPU 驱动 galcore.so 版本及所适用的芯片型号与依赖库保持一致，在部署前，请登录开发板，并通过命令行输入以下命令查询 NPU 驱动版本，晶晨建议的驱动版本为：6.4.4.3

2
docs/cn/build_and_install/android.md Normal file → Executable file
View File

@@ -1,6 +1,6 @@
# Android部署库编译
FastDeploy当前在Android仅支持Paddle-Lite后端推理，支持armeabi-v7a和arm64-v8a两种cpu架构，在armv8.2架构的arm设备支持fp16精度推理。相关编译选项说明如下：
FastDeploy当前在Android仅支持Paddle Lite后端推理，支持armeabi-v7a和arm64-v8a两种cpu架构，在armv8.2架构的arm设备支持fp16精度推理。相关编译选项说明如下：
|编译选项|默认值|说明|备注|
|:---|:---|:---|:---|

4
docs/cn/build_and_install/rv1126.md
View File

@@ -1,6 +1,6 @@
# 瑞芯微 RV1126 部署环境编译安装
FastDeploy基于 Paddle-Lite 后端支持在瑞芯微（Rockchip）Soc 上进行部署推理。
FastDeploy基于 Paddle Lite 后端支持在瑞芯微（Rockchip）Soc 上进行部署推理。
更多详细的信息请参考：[Paddle Lite部署示例](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/verisilicon_timvx.html)。
本文档介绍如何编译基于 Paddle Lite 的 C++ FastDeploy 交叉编译库。
@@ -67,7 +67,7 @@ cmake -DCMAKE_TOOLCHAIN_FILE=./../cmake/toolchain.cmake \
make -j8
make install
```
编译完成之后，会生成 fastdeploy-tmivx 目录，表示基于 PadddleLite TIM-VX 的 FastDeploy 库编译完成。
编译完成之后，会生成 fastdeploy-tmivx 目录，表示基于 Paddle Lite TIM-VX 的 FastDeploy 库编译完成。
## 准备设备运行环境
部署前要保证芯原 Linux Kernel NPU 驱动 galcore.so 版本及所适用的芯片型号与依赖库保持一致，在部署前，请登录开发板，并通过命令行输入以下命令查询 NPU 驱动版本，Rockchip建议的驱动版本为： 6.4.6.5

4
docs/cn/build_and_install/xpu.md
View File

@@ -1,6 +1,6 @@
# 昆仑芯 XPU 部署环境编译安装
FastDeploy 基于 Paddle-Lite 后端支持在昆仑芯 XPU 上进行部署推理。
FastDeploy 基于 Paddle Lite 后端支持在昆仑芯 XPU 上进行部署推理。
更多详细的信息请参考：[Paddle Lite部署示例](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/kunlunxin_xpu.html#xpu)。
本文档介绍如何编译基于 Paddle Lite 的 C++ FastDeploy 编译库。
@@ -52,7 +52,7 @@ cmake -DWITH_XPU=ON \
make -j8
make install
```
编译完成之后，会生成 fastdeploy-xpu 目录，表示基于 PadddleLite 的 FastDeploy 库编译完成。
编译完成之后，会生成 fastdeploy-xpu 目录，表示基于 Paddle Lite 的 FastDeploy 库编译完成。
## Python 编译
编译命令如下：

6
docs/en/build_and_install/a311d.md
View File

@@ -1,8 +1,8 @@
# How to Build A311D Deployment Environment
FastDeploy supports AI deployment on Rockchip Soc based on Paddle-Lite backend. For more detailed information, please refer to: [PaddleLite Deployment Example](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/verisilicon_timvx.html).
FastDeploy supports AI deployment on Rockchip Soc based on Paddle Lite backend. For more detailed information, please refer to: [Paddle Lite Deployment Example](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/verisilicon_timvx.html).
This document describes how to compile the PaddleLite-based C++ FastDeploy cross-compilation library.
This document describes how to compile the Paddle Lite based C++ FastDeploy cross-compilation library.
The relevant compilation options are described as follows:
|Compile Options|Default Values|Description|Remarks|
@@ -66,7 +66,7 @@ cmake -DCMAKE_TOOLCHAIN_FILE=./../cmake/toolchain.cmake \
make -j8
make install
```
After the compilation is complete, the fastdeploy-tmivx directory will be generated, indicating that the FastDeploy library based on PadddleLite TIM-VX has been compiled.
After the compilation is complete, the fastdeploy-tmivx directory will be generated, indicating that the FastDeploy library based on Paddle Lite TIM-VX has been compiled.
## Prepare the Soc environment
Before deployment, ensure that the version of the driver galcore.so of the Verisilicon Linux Kernel NPU meets the requirements. Before deployment, please log in to the development board, and enter the following command through the command line to query the NPU driver version. The recommended version of the Rockchip driver is: 6.4.4.3

4
docs/en/build_and_install/android.md Normal file → Executable file
View File

@@ -1,12 +1,12 @@
# How to Build FastDeploy Android C++ SDK
FastDeploy supports Paddle-Lite backend on Android. It supports both armeabi-v7a and arm64-v8a cpu architectures, and supports fp16 precision inference on the armv8.2 architecture. The relevant compilation options are described as follows:
FastDeploy supports Paddle Lite backend on Android. It supports both armeabi-v7a and arm64-v8a cpu architectures, and supports fp16 precision inference on the armv8.2 architecture. The relevant compilation options are described as follows:
|Option|Default|Description|Remark|
|:---|:---|:---|:---|
|ENABLE_LITE_BACKEND|OFF|It needs to be set to ON when compiling the Android library| - |
|WITH_OPENCV_STATIC|OFF|Whether to use the OpenCV static library| - |
|WITH_LITE_STATIC|OFF|Whether to use the Paddle-Lite static library| NOT Support now |
|WITH_LITE_STATIC|OFF|Whether to use the Paddle Lite static library| NOT Support now |
Please reference [FastDeploy Compile Options](./README.md) for more details.

6
docs/en/build_and_install/rv1126.md
View File

@@ -1,8 +1,8 @@
# How to Build RV1126 Deployment Environment
FastDeploy supports AI deployment on Rockchip Soc based on Paddle-Lite backend. For more detailed information, please refer to: [PaddleLite Deployment Example](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/verisilicon_timvx.html).
FastDeploy supports AI deployment on Rockchip Soc based on Paddle Lite backend. For more detailed information, please refer to: [Paddle Lite Deployment Example](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/verisilicon_timvx.html).
This document describes how to compile the PaddleLite-based C++ FastDeploy cross-compilation library.
This document describes how to compile the Paddle Lite based C++ FastDeploy cross-compilation library.
The relevant compilation options are described as follows:
|Compile Options|Default Values|Description|Remarks|
@@ -66,7 +66,7 @@ cmake -DCMAKE_TOOLCHAIN_FILE=./../cmake/toolchain.cmake \
make -j8
make install
```
After the compilation is complete, the fastdeploy-tmivx directory will be generated, indicating that the FastDeploy library based on PadddleLite TIM-VX has been compiled.
After the compilation is complete, the fastdeploy-tmivx directory will be generated, indicating that the FastDeploy library based on Paddle Lite TIM-VX has been compiled.
## Prepare the Soc environment
Before deployment, ensure that the version of the driver galcore.so of the Verisilicon Linux Kernel NPU meets the requirements. Before deployment, please log in to the development board, and enter the following command through the command line to query the NPU driver version. The recommended version of the Rockchip driver is: 6.4.6.5

4
docs/en/build_and_install/xpu.md
View File

@@ -1,6 +1,6 @@
# How to Build KunlunXin XPU Deployment Environment
FastDeploy supports deployment AI on KunlunXin XPU based on Paddle-Lite backend. For more detailed information, please refer to: [PaddleLite Deployment Example](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/kunlunxin_xpu.html#xpu)。
FastDeploy supports deployment AI on KunlunXin XPU based on Paddle Lite backend. For more detailed information, please refer to: [Paddle Lite Deployment Example](https://www.paddlepaddle.org.cn/lite/develop/demo_guides/kunlunxin_xpu.html#xpu)。
This document describes how to compile the C++ FastDeploy library based on Paddle Lite.
@@ -55,7 +55,7 @@ cmake -DWITH_XPU=ON \
make -j8
make install
```
After the compilation is complete, the fastdeploy-xpu directory will be generated, indicating that the PadddleLite-based FastDeploy library has been compiled.
After the compilation is complete, the fastdeploy-xpu directory will be generated, indicating that the Padddle Lite based FastDeploy library has been compiled.
## Python compile
The compilation command is as follows:

0
examples/application/js/converter/DEVELOPMENT.md Normal file → Executable file
View File

0
examples/application/js/converter/DEVELOPMENT_cn.md Normal file → Executable file
View File

50
examples/application/js/converter/convertToPaddleJSModel.py Normal file → Executable file
View File

@@ -9,19 +9,20 @@ import stat
import traceback
import copy
def cleanTempModel(optimizedModelTempDir):
""" 清理opt优化完的临时模型文件 """
if os.path.exists(optimizedModelTempDir):
print("Cleaning optimized temporary model...")
shutil.rmtree(optimizedModelTempDir, onerror=grantWritePermission)
def grantWritePermission(func, path, execinfo):
""" 文件授权 """
os.chmod(path, stat.S_IWRITE)
func(path)
def main():
"""
Example:
@@ -29,13 +30,34 @@ def main():
"""
try:
p = argparse.ArgumentParser(description='转化为PaddleJS模型参数解析')
p.add_argument('--inputDir', help='fluid模型所在目录。当且仅当使用分片参数文件时使用该参数。将过滤modelPath和paramsPath参数，且模型文件名必须为`__model__`', required=False)
p.add_argument('--modelPath', help='fluid模型文件所在路径，使用合并参数文件时使用该参数', required=False)
p.add_argument('--paramPath', help='fluid参数文件所在路径，使用合并参数文件时使用该参数', required=False)
p.add_argument("--outputDir", help='paddleJS模型输出路径，必要参数', required=True)
p.add_argument("--disableOptimize", type=int, default=0, help='是否关闭模型优化，非必要参数，1为关闭优化，0为开启优化，默认开启优化', required=False)
p.add_argument("--logModelInfo", type=int, default=0, help='是否输出模型结构信息，非必要参数，0为不输出，1为输出，默认不输出', required=False)
p.add_argument("--sliceDataSize", type=int, default=4096, help='分片输出参数文件时，每片文件的大小，单位：KB，非必要参数，默认4096KB', required=False)
p.add_argument(
'--inputDir',
help='fluid模型所在目录。当且仅当使用分片参数文件时使用该参数。将过滤modelPath和paramsPath参数，且模型文件名必须为`__model__`',
required=False)
p.add_argument(
'--modelPath', help='fluid模型文件所在路径，使用合并参数文件时使用该参数', required=False)
p.add_argument(
'--paramPath', help='fluid参数文件所在路径，使用合并参数文件时使用该参数', required=False)
p.add_argument(
"--outputDir", help='paddleJS模型输出路径，必要参数', required=True)
p.add_argument(
"--disableOptimize",
type=int,
default=0,
help='是否关闭模型优化，非必要参数，1为关闭优化，0为开启优化，默认开启优化',
required=False)
p.add_argument(
"--logModelInfo",
type=int,
default=0,
help='是否输出模型结构信息，非必要参数，0为不输出，1为输出，默认不输出',
required=False)
p.add_argument(
"--sliceDataSize",
type=int,
default=4096,
help='分片输出参数文件时，每片文件的大小，单位：KB，非必要参数，默认4096KB',
required=False)
p.add_argument('--useGPUOpt', help='转换模型是否执行GPU优化方法', required=False)
args = p.parse_args()
@@ -76,7 +98,6 @@ def main():
args.modelPath = os.path.join(optimizedModelTempDir, "model")
args.paramPath = os.path.join(optimizedModelTempDir, "params")
print("============Convert Model Args=============")
if inputDir:
print("inputDir: " + inputDir)
@@ -88,14 +109,14 @@ def main():
print("enableLogModelInfo: " + str(enableLogModelInfo))
print("sliceDataSize:" + str(sliceDataSize))
print("Starting...")
if enableOptimization:
print("Optimizing model...")
for param in ["inputDir", "modelPath", "paramPath", "outputDir"]:
if optArgs.__dict__[param]:
# 用""框起命令参数值，解决路径中的空格问题
optimizeCmd += " --" + param + "="+ '"' + str(optArgs.__dict__[param]) + '"'
optimizeCmd += " --" + param + "=" + '"' + str(
optArgs.__dict__[param]) + '"'
os.system(optimizeCmd)
try:
os.listdir(optimizedModelTempDir)
@@ -110,13 +131,16 @@ def main():
else:
print("\n\033[32mOptimizing model successfully.\033[0m")
else:
print("\033[33mYou choosed not to optimize model, consequently, optimizing model is skiped.\033[0m")
print(
"\033[33mYou choosed not to optimize model, consequently, optimizing model is skiped.\033[0m"
)
print("\nConverting model...")
for param in args.__dict__:
if args.__dict__[param]:
# 用""框起参数，解决路径中的空格问题
convertCmd += " --" + param + "=" + '"' + str(args.__dict__[param]) + '"'
convertCmd += " --" + param + "=" + '"' + str(args.__dict__[
param]) + '"'
os.system(convertCmd)
try:
file = os.listdir(outputDir)

38
examples/application/js/converter/fuseOps.py Normal file → Executable file
View File

@@ -1,20 +1,12 @@
#!/usr/bin/env python
# -*- coding: UTF-8 -*-
def opListFuse(ops):
""" 算子融合 """
fuseOpList = [
'relu',
'relu6',
'leaky_relu',
'scale',
'sigmoid',
'hard_sigmoid',
'pow',
'sqrt',
'tanh',
'hard_swish',
'dropout'
'relu', 'relu6', 'leaky_relu', 'scale', 'sigmoid', 'hard_sigmoid',
'pow', 'sqrt', 'tanh', 'hard_swish', 'dropout'
]
# 判断op是否为单节点
@@ -37,39 +29,41 @@ def opListFuse(ops):
else:
return False
for index in reversed(range(len(ops))):
if index > 0:
op = ops[index]
# 兼容paddlelite 算子融合字段
# 兼容 Paddle Lite 算子融合字段
if 'act_type' in op['attrs']:
name = op['attrs']['act_type']
op['attrs']['fuse_opt'] = {}
op['attrs']['fuse_opt'][name] = {}
if name == 'hard_swish':
op['attrs']['fuse_opt'][name]['offset'] = op['attrs']['hard_swish_offset']
op['attrs']['fuse_opt'][name]['scale'] = op['attrs']['hard_swish_scale']
op['attrs']['fuse_opt'][name]['threshold'] = op['attrs']['hard_swish_threshold']
op['attrs']['fuse_opt'][name]['offset'] = op['attrs'][
'hard_swish_offset']
op['attrs']['fuse_opt'][name]['scale'] = op['attrs'][
'hard_swish_scale']
op['attrs']['fuse_opt'][name]['threshold'] = op['attrs'][
'hard_swish_threshold']
if name == 'relu6':
op['attrs']['fuse_opt'][name]['threshold'] = op['attrs']['fuse_brelu_threshold']
op['attrs']['fuse_opt'][name]['threshold'] = op['attrs'][
'fuse_brelu_threshold']
for fuse in fuseOpList:
if op['type'] == fuse:
prevOp = ops[index - 1]
if opExistSingleNode(prevOp['outputs']['Out'][0]) and len(prevOp['outputs']['Out']) == 1 :
if opExistSingleNode(prevOp['outputs']['Out'][0]) and len(
prevOp['outputs']['Out']) == 1:
prevOp['attrs']['fuse_opt'] = {}
if 'fuse_opt' in op['attrs']:
prevOp['attrs']['fuse_opt'] = op['attrs']['fuse_opt']
prevOp['attrs']['fuse_opt'] = op['attrs'][
'fuse_opt']
del op['attrs']['fuse_opt']
prevOp['attrs']['fuse_opt'][fuse] = op['attrs']
prevOp['outputs']['Out'] = op['outputs']['Out']
del ops[index]

2
fastdeploy/runtime.h
View File

@@ -372,7 +372,7 @@ struct FASTDEPLOY_DECL RuntimeOption {
float ipu_available_memory_proportion = 1.0;
bool ipu_enable_half_partial = false;
// ======Only for Paddle-Lite Backend=====
// ======Only for Paddle Lite Backend=====
// 0: LITE_POWER_HIGH 1: LITE_POWER_LOW 2: LITE_POWER_FULL
// 3: LITE_POWER_NO_BIND 4: LITE_POWER_RAND_HIGH
// 5: LITE_POWER_RAND_LOW