Untitled.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "7a4812c2",
   "metadata": {},
   "source": [
    "# Lite预测功能测试\n",
    "\n",
    "Lite预测功能测试的主程序为`test_lite.sh`，可以测试基于Lite预测库的模型推理功能。\n",
    "\n",
    "## 1. 测试结论汇总\n",
    "\n",
    "目前Lite端的样本间支持以下四种方式的组合，共16个链条：\n",
    "\n",
    "| 模型类型 | 样本数量 | 功耗模式 | cpu多线程 |\n",
    "|  ----   |   ----   |  :----:  |  :----:  |\n",
    "| 正常模型 | 单样本/多样本 | LITE_POWER_HIGH/LITE_POWER_LOW | 支持 |\n",
    "| 正常模型 | 单样本/多样本 | LITE_POWER_HIGH/LITE_POWER_LOW | 支持 |\n",
    "| 量化模型 | 单样本/多样本 | LITE_POWER_HIGH/LITE_POWER_LOW | 支持 |\n",
    "| 量化模型 | 单样本/多样本 | LITE_POWER_HIGH/LITE_POWER_LOW | 支持 |\n",
    "\n",
    "## 2. 测试流程\n",
    "\n",
    "### 2.1 功能测试\n",
    "\n",
    "先运行`prepare.sh`准备数据和模型，模型和数据会打包到test_lite.tar中，将test_lite.tar上传到手机上，然后运行`test_lite.sh`进行测试，最终在`test_lite/output`目录下生成`lite_*.log`后缀的日志文件。\n",
    "\n",
    "```shell\n",
    "\n",
    "# 数据和模型准备\n",
    "bash PTDN/prepare.sh ./PTDN/configs/ppocr_det_mobile_params.txt \"lite_infer\"\n",
    "\n",
    "# 手机端测试:\n",
    "bash PTND/test_lite.sh ./PTDN/configs/ppocr_det_mobile_params.txt\n",
    "\n",
    "```  \n",
    "\n",
    "**注意**：数据上传到手机的方式有两种，一种是通过adb的方式通过数据线连接，可以参考[Lite端部署文档](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/lite/readme.md)。另外一种是通过在手机上开启虚拟终端在局域网内通过ssh的方式连接，无需数据线，第二种方式可以在手机上支持大部分linux命令，如vim、bash等，比较推荐，可以参考[安卓手机通过termux连接电脑](./termux_for_android.md)。\n",
    "\n",
    "#### 运行结果\n",
    "\n",
    "各测试的运行情况会打印在 `./output/` 中：\n",
    "运行成功时会输出：\n",
    "\n",
    "```\n",
    "Run successfully with command - ./ocr_db_crnn det ./models/ch_ppocr_mobile_v2.0_det_slim_opt.nb INT8 4 1 LITE_POWER_LOW ./test_data/icdar2015_lite/text_localization/ch4_test_images/img_233.jpg ./config.txt True > ./output/lite_ch_ppocr_mobile_v2.0_det_slim_opt.nb_precision_INT8_batchsize_1_threads_4_powermode_LITE_POWER_LOW_singleimg_True.log 2>&1!\n",
    "Run successfully with command xxx\n",
    "...\n",
    "```\n",
    "\n",
    "运行失败时会输出：\n",
    "\n",
    "```\n",
    "Run failed with command - ./ocr_db_crnn det ./models/ch_ppocr_mobile_v2.0_det_slim_opt.nb INT8 4 1 LITE_POWER_LOW ./test_data/icdar2015_lite/text_localization/ch4_test_images/img_233.jpg ./config.txt True > ./output/lite_ch_ppocr_mobile_v2.0_det_slim_opt.nb_precision_INT8_batchsize_1_threads_4_powermode_LITE_POWER_LOW_singleimg_True.log 2>&1!\n",
    "Run failed with command xxx\n",
    "...\n",
    "```\n",
    "\n",
    "在./output/文件夹下，会存在如下日志，每一个日志都是不同配置下的log结果：\n",
    "\n",
    "<img src=\"lite_log.png\" width=\"1000\">\n",
    "\n",
    "在每一个log中，都会调用autolog打印如下信息：\n",
    "\n",
    "<img src=\"lite_auto_log.png\" width=\"1000\">\n",
    "\n",
    "\n",
    "\n",
    "## 3. 更多教程\n",
    "\n",
    "本文档为功能测试用，更详细的Lite端预测使用教程请参考：[Lite端部署](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/lite/readme.md)  "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "005fdd25",
   "metadata": {},
   "source": [
    "# 安卓手机通过Termux连接电脑\n",
    "\n",
    "由于通过adb方式连接手机后，很多linux命令无法运行，自动化测试受阻，所以此处特此介绍另外一种通过Termux的连接方式，不仅可以运行大部分linux命令，方便开发者在手机上在线调试，甚至还可以多实现台机器同时连接手机。Termux不是真实的Linux环境，但是Termux可以安装真实的Linux，而且不会损失性能，与此同时，Termux不需要root。在配置Termux之前，请确保电脑已经安装adb工具，安装方式请参考[Lite端部署](https://github.com/PaddlePaddle/PaddleOCR/blob/develop/deploy/lite/readme.md) 。在运行以下命令后确保可以显示安卓设备信息。\n",
    "\n",
    "```\n",
    "adb devices\n",
    "```\n",
    "连接成功信息提示：\n",
    "```\n",
    "List of devices attached\n",
    "744be294    device\n",
    "```\n",
    "\n",
    "## 1.安卓手机安装termux app\n",
    "\n",
    "### 1.1 下载termux apk文件\n",
    "\n",
    "由于目前该app目前各大商城暂无，所以可以直接下载如下apk文件。\n",
    "\n",
    "打开电脑终端，执行以下命令：\n",
    "\n",
    "```\n",
    "wget http://10.12.121.133:8911/cuicheng01/fullchain/termux-v1.0.3.apk\n",
    "```\n",
    "\n",
    "### 1.2 安装termux到手机上\n",
    "\n",
    "在手机端的开发者模式下，允许USB调试，允许USB安装。在电脑终端，执行如下命令，将termux app安装到手机上：\n",
    "\n",
    "```\n",
    "adb install termux-v1.0.3.apk\n",
    "```\n",
    "\n",
    "此处需要手机端确认安装，点击确认。\n",
    "\n",
    "### 1.3 验证是否安装成功\n",
    "\n",
    "打开手机，检验termux是否安装成功，如果没有，重新执行1.2，如果有相应的app，点击进入，会有如下显示。\n",
    "\n",
    "<img src=\"termux.jpg\" width=\"300\" height = \"300\">\n",
    "\n",
    "接下来的配置环境需要在手机上此终端运行相关命令。\n",
    "\n",
    "## 2.手机端配置termux\n",
    "\n",
    "首先将手机联网，最好可以连接外网，部分的配置需要外网。打开Termux终端，执行以下命令安装基础件`proot`，并使用`termux-chroot`命令可以模拟 root 环境与标准的 Linux 目录结构。\n",
    "\n",
    "```\n",
    "pkg i -y proot\n",
    "termux-chroot\n",
    "```\n",
    "\n",
    "Termux 默认只能访问自身内部的数据，如果要访问手机中其它的数据，输入下面的命令后，手机弹出对请求权限的窗口，允许即可（方便对部分运行出的结果在手机端可视化）。\n",
    "\n",
    "```\n",
    "termux-setup-storage\n",
    "```\n",
    "\n",
    "### 2.1 配置SSH\n",
    "\n",
    "作为 Linux 终端或者服务器，必须有SSH。不管你是 SSH 连接到 Termux还是使用Termux去连其它主机，都需要先安装openssh。如果安装失败，请重复执行命令。\n",
    "\n",
    "```\n",
    "pkg i -y openssh\n",
    "```\n",
    "\n",
    "启动 SSH 服务端，默认端口号为8022\n",
    "\n",
    "```\n",
    "sshd\n",
    "```\n",
    "\n",
    "\n",
    "### 2.2 电脑通过SSH方式连接手机\n",
    "\n",
    "1.保证手机和电脑处于同一局域网下\n",
    "手机端分别输入以下命令获得ip地址和当前用户：\n",
    "\n",
    "```\n",
    "# 获取ip地址\n",
    "ifconfig\n",
    "\n",
    "# 获取当前用户\n",
    "whoami\n",
    "```\n",
    "\n",
    "如获取到的ip地址和当前用户分别是`172.24.162.117`和`u0_a374`。\n",
    "\n",
    "2.电脑端通过SSH连接手机\n",
    "\n",
    "```\n",
    "#默认端口号为8022\n",
    "ssh u0_a374@172.24.162.117 -p 8022\n",
    "```\n",
    "\n",
    "3.运行ls命令后，会有如下显示：\n",
    "\n",
    "```\n",
    "ls\n",
    "```\n",
    "\n",
    "<img src=\"ssh_termux_ls.png\" width=\"800\">\n",
    "\n",
    "\n",
    "### 2.3 通过scp传输数据\n",
    "\n",
    "1.在当前目录上新建test目录\n",
    "\n",
    "```\n",
    "mkdir test\n",
    "```\n",
    "\n",
    "2.测试scp功能\n",
    "\n",
    "将电脑中的某个文件拷贝到手机上：\n",
    "```\n",
    "scp -P 8022 test.txt u0_a374@172.24.162.117:/home/storage/test\n",
    "```\n",
    "\n",
    "3.手机端查看\n",
    "\n",
    "打开手机终端，在`/home/storage/test`下查看是否存在`test.txt`\n",
    "\n",
    "\n",
    "## 3. 更多教程\n",
    "\n",
    "本教程可以完成Termux基本配置，更多关于Termux的用法，请参考：[Termux高级终端安装使用配置教程](https://www.sqlsec.com/2018/05/termux.html)。\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9276a65a",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}