From 3cf2584cfb70a97f30ffe91dade629849743b2a4 Mon Sep 17 00:00:00 2001 From: wangjincheng123456 <35827074+wangjincheng123456@users.noreply.github.com> Date: Fri, 13 May 2022 17:18:10 +0800 Subject: [PATCH] Add files via upload --- docs/practices/jit/Face recognition-jit.ipynb | 2821 +++++++++++++++++ docs/practices/jit/Image denoising-jit.ipynb | 1297 ++++++++ .../jit/Semantic segmentation-jit.ipynb | 1040 ++++++ 3 files changed, 5158 insertions(+) create mode 100644 docs/practices/jit/Face recognition-jit.ipynb create mode 100644 docs/practices/jit/Image denoising-jit.ipynb create mode 100644 docs/practices/jit/Semantic segmentation-jit.ipynb diff --git a/docs/practices/jit/Face recognition-jit.ipynb b/docs/practices/jit/Face recognition-jit.ipynb new file mode 100644 index 00000000000..da1d18b4ed2 --- /dev/null +++ b/docs/practices/jit/Face recognition-jit.ipynb @@ -0,0 +1,2821 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-03T09:37:09.082319Z", + "iopub.status.busy": "2022-05-03T09:37:09.081349Z", + "iopub.status.idle": "2022-05-03T09:37:09.312566Z", + "shell.execute_reply": "2022-05-03T09:37:09.311225Z", + "shell.execute_reply.started": "2022-05-03T09:37:09.082262Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "data39602\n" + ] + } + ], + "source": [ + "# 查看当前挂载的数据集目录, 该目录下的变更重启环境后会自动还原\n", + "# View dataset directory. \n", + "# This directory will be recovered automatically after resetting environment. \n", + "!ls /home/aistudio/data" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-03T09:37:15.704486Z", + "iopub.status.busy": "2022-05-03T09:37:15.703512Z", + "iopub.status.idle": "2022-05-03T09:37:15.929880Z", + "shell.execute_reply": "2022-05-03T09:37:15.928738Z", + "shell.execute_reply.started": "2022-05-03T09:37:15.704438Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "# 查看工作区文件, 该目录下的变更将会持久保存. 请及时清理不必要的文件, 避免加载过慢.\n", + "# View personal work directory. \n", + "# All changes under this directory will be kept even after reset. \n", + "# Please clean unnecessary files in time to speed up environment loading. \n", + "!ls /home/aistudio/work" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-03T09:37:22.035866Z", + "iopub.status.busy": "2022-05-03T09:37:22.035240Z", + "iopub.status.idle": "2022-05-03T09:37:24.853978Z", + "shell.execute_reply": "2022-05-03T09:37:24.852354Z", + "shell.execute_reply.started": "2022-05-03T09:37:22.035825Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple\n", + "Collecting beautifulsoup4\n", + " Downloading https://pypi.tuna.tsinghua.edu.cn/packages/9c/d8/909c4089dbe4ade9f9705f143c9f13f065049a9d5e7d34c828aefdd0a97c/beautifulsoup4-4.11.1-py3-none-any.whl (128 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m128.2/128.2 KB\u001b[0m \u001b[31m13.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25hCollecting soupsieve>1.2\n", + " Downloading https://pypi.tuna.tsinghua.edu.cn/packages/16/e3/4ad79882b92617e3a4a0df1960d6bce08edfb637737ac5c3f3ba29022e25/soupsieve-2.3.2.post1-py3-none-any.whl (37 kB)\n", + "Installing collected packages: soupsieve, beautifulsoup4\n", + "Successfully installed beautifulsoup4-4.11.1 soupsieve-2.3.2.post1\n" + ] + } + ], + "source": [ + "# 如果需要进行持久化安装, 需要使用持久化路径, 如下方代码示例:\n", + "# If a persistence installation is required, \n", + "# you need to use the persistence path as the following: \n", + "!mkdir /home/aistudio/external-libraries\n", + "!pip install beautifulsoup4 -t /home/aistudio/external-libraries" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-03T09:37:31.215094Z", + "iopub.status.busy": "2022-05-03T09:37:31.214588Z", + "iopub.status.idle": "2022-05-03T09:37:31.220602Z", + "shell.execute_reply": "2022-05-03T09:37:31.218962Z", + "shell.execute_reply.started": "2022-05-03T09:37:31.215055Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "# 同时添加如下代码, 这样每次环境(kernel)启动的时候只要运行下方代码即可: \n", + "# Also add the following code, \n", + "# so that every time the environment (kernel) starts, \n", + "# just run the following code: \n", + "import sys \n", + "sys.path.append('/home/aistudio/external-libraries')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 数据处理" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T09:38:19.500231Z", + "iopub.status.busy": "2022-05-03T09:38:19.499662Z", + "iopub.status.idle": "2022-05-03T09:38:20.022909Z", + "shell.execute_reply": "2022-05-03T09:38:20.021817Z", + "shell.execute_reply.started": "2022-05-03T09:38:19.500189Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Archive: /home/aistudio/data/data39602/images.zip\n", + " creating: images/\n", + " creating: images/face/\n", + " creating: images/face/dilireba/\n", + " inflating: images/face/dilireba/15911604352.jpg \n", + " inflating: images/face/dilireba/15911604353.jpg \n", + " inflating: images/face/dilireba/15911604354.jpg \n", + " inflating: images/face/dilireba/15911604355.jpg \n", + " inflating: images/face/dilireba/159116043610.jpg \n", + " inflating: images/face/dilireba/159116043611.jpg \n", + " inflating: images/face/dilireba/159116043612.jpg \n", + " inflating: images/face/dilireba/159116043613.jpg \n", + " inflating: images/face/dilireba/159116043614.jpg \n", + " inflating: images/face/dilireba/159116043615.jpg \n", + " inflating: images/face/dilireba/159116043616.jpg \n", + " inflating: images/face/dilireba/159116043618.jpg \n", + " inflating: images/face/dilireba/159116043620.jpg \n", + " inflating: images/face/dilireba/159116043622.jpg \n", + " inflating: images/face/dilireba/159116043623.jpg \n", + " inflating: images/face/dilireba/159116043624.jpg \n", + " inflating: images/face/dilireba/159116043625.jpg \n", + " inflating: images/face/dilireba/159116043626.jpg \n", + " inflating: images/face/dilireba/159116043627.jpg \n", + " inflating: images/face/dilireba/159116043628.jpg \n", + " inflating: images/face/dilireba/159116043629.jpg \n", + " inflating: images/face/dilireba/159116043631.jpg \n", + " inflating: images/face/dilireba/159116043632.jpg \n", + " inflating: images/face/dilireba/159116043633.jpg \n", + " inflating: images/face/dilireba/159116043634.jpg \n", + " inflating: images/face/dilireba/159116043635.jpg \n", + " inflating: images/face/dilireba/159116043636.jpg \n", + " inflating: images/face/dilireba/159116043637.jpg \n", + " inflating: images/face/dilireba/159116043638.jpg \n", + " inflating: images/face/dilireba/159116043639.jpg \n", + " inflating: images/face/dilireba/159116043640.jpg \n", + " inflating: images/face/dilireba/159116043641.jpg \n", + " inflating: images/face/dilireba/159116043642.jpg \n", + " inflating: images/face/dilireba/159116043643.jpg \n", + " inflating: images/face/dilireba/159116043644.jpg \n", + " inflating: images/face/dilireba/159116043645.jpg \n", + " inflating: images/face/dilireba/159116043646.jpg \n", + " inflating: images/face/dilireba/159116043647.jpg \n", + " inflating: images/face/dilireba/159116043648.jpg \n", + " inflating: images/face/dilireba/159116043649.jpg \n", + " inflating: images/face/dilireba/159116043650.jpg \n", + " inflating: images/face/dilireba/159116043651.jpg \n", + " inflating: images/face/dilireba/159116043652.jpg \n", + " inflating: images/face/dilireba/159116043653.jpg \n", + " inflating: images/face/dilireba/159116043654.jpg \n", + " inflating: images/face/dilireba/159116043655.jpg \n", + " inflating: images/face/dilireba/159116043656.jpg \n", + " inflating: images/face/dilireba/159116043657.jpg \n", + " inflating: images/face/dilireba/159116043658.jpg \n", + " inflating: images/face/dilireba/159116043659.jpg \n", + " inflating: images/face/dilireba/15911604366.jpg \n", + " inflating: images/face/dilireba/159116043660.jpg \n", + " inflating: images/face/dilireba/15911604367.jpg \n", + " inflating: images/face/dilireba/15911604368.jpg \n", + " inflating: images/face/dilireba/15911604369.jpg \n", + " inflating: images/face/dilireba/1591160437100.jpg \n", + " inflating: images/face/dilireba/1591160437101.jpg \n", + " inflating: images/face/dilireba/1591160437102.jpg \n", + " inflating: images/face/dilireba/1591160437103.jpg \n", + " inflating: images/face/dilireba/1591160437104.jpg \n", + " inflating: images/face/dilireba/1591160437105.jpg \n", + " inflating: images/face/dilireba/1591160437107.jpg \n", + " inflating: images/face/dilireba/1591160437108.jpg \n", + " inflating: images/face/dilireba/1591160437109.jpg \n", + " inflating: images/face/dilireba/1591160437110.jpg \n", + " inflating: images/face/dilireba/1591160437111.jpg \n", + " inflating: images/face/dilireba/1591160437112.jpg \n", + " inflating: images/face/dilireba/1591160437114.jpg \n", + " inflating: images/face/dilireba/159116043761.jpg \n", + " inflating: images/face/dilireba/159116043762.jpg \n", + " inflating: images/face/dilireba/159116043763.jpg \n", + " inflating: images/face/dilireba/159116043764.jpg \n", + " inflating: images/face/dilireba/159116043765.jpg \n", + " inflating: images/face/dilireba/159116043766.jpg \n", + " inflating: images/face/dilireba/159116043767.jpg \n", + " inflating: images/face/dilireba/159116043768.jpg \n", + " inflating: images/face/dilireba/159116043769.jpg \n", + " inflating: images/face/dilireba/159116043770.jpg \n", + " inflating: images/face/dilireba/159116043771.jpg \n", + " inflating: images/face/dilireba/159116043772.jpg \n", + " inflating: images/face/dilireba/159116043773.jpg \n", + " inflating: images/face/dilireba/159116043775.jpg \n", + " inflating: images/face/dilireba/159116043776.jpg \n", + " inflating: images/face/dilireba/159116043777.jpg \n", + " inflating: images/face/dilireba/159116043779.jpg \n", + " inflating: images/face/dilireba/159116043781.jpg \n", + " inflating: images/face/dilireba/159116043783.jpg \n", + " inflating: images/face/dilireba/159116043784.jpg \n", + " inflating: images/face/dilireba/159116043785.jpg \n", + " inflating: images/face/dilireba/159116043786.jpg \n", + " inflating: images/face/dilireba/159116043787.jpg \n", + " inflating: images/face/dilireba/159116043788.jpg \n", + " inflating: images/face/dilireba/159116043789.jpg \n", + " inflating: images/face/dilireba/159116043791.jpg \n", + " inflating: images/face/dilireba/159116043792.jpg \n", + " inflating: images/face/dilireba/159116043793.jpg \n", + " inflating: images/face/dilireba/159116043794.jpg \n", + " inflating: images/face/dilireba/159116043796.jpg \n", + " inflating: images/face/dilireba/159116043797.jpg \n", + " inflating: images/face/dilireba/159116043798.jpg \n", + " inflating: images/face/dilireba/159116043799.jpg \n", + " inflating: images/face/dilireba/1591160438115.jpg \n", + " inflating: images/face/dilireba/1591160438116.jpg \n", + " inflating: images/face/dilireba/1591160438117.jpg \n", + " inflating: images/face/dilireba/1591160438118.jpg \n", + " inflating: images/face/dilireba/1591160438119.jpg \n", + " inflating: images/face/dilireba/1591160438120.jpg \n", + " inflating: images/face/dilireba/1591160438121.jpg \n", + " inflating: images/face/dilireba/1591160438122.jpg \n", + " inflating: images/face/dilireba/1591160438123.jpg \n", + " inflating: images/face/dilireba/1591160438124.jpg \n", + " inflating: images/face/dilireba/1591160438125.jpg \n", + " creating: images/face/jiangwen/\n", + " inflating: images/face/jiangwen/0acb8d12-f929-11e8-ac67-005056c00008.jpg \n", + " inflating: images/face/jiangwen/0b1937e2-f929-11e8-8a8a-005056c00008.jpg \n", + " inflating: images/face/jiangwen/10acb8d12-f929-11e8-ac67-005056c00008.jpg \n", + " inflating: images/face/jiangwen/10b1937e2-f929-11e8-8a8a-005056c00008.jpg \n", + " inflating: images/face/jiangwen/118207c00-f929-11e8-bee9-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1185e2140-f929-11e8-9b08-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/118b60540-f929-11e8-aefd-005056c00008.jpg \n", + " inflating: images/face/jiangwen/11ad8ed12-f929-11e8-ab1f-005056c00008.jpg \n", + " inflating: images/face/jiangwen/11b9203e2-f929-11e8-970c-005056c00008.jpg \n", + " inflating: images/face/jiangwen/11bc5be12-f929-11e8-860a-005056c00008.jpg \n", + " inflating: images/face/jiangwen/11c749bb0-f929-11e8-bfb9-005056c00008.jpg \n", + " inflating: images/face/jiangwen/11cc97270-f929-11e8-8f86-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/11de1c040-f929-11e8-b9f1-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/11e2c7540-f929-11e8-b467-005056c00008.jpg \n", + " inflating: images/face/jiangwen/11e78d7f0-f929-11e8-b323-005056c00008.jpg \n", + " inflating: images/face/jiangwen/11eb9ffa2-f929-11e8-a78c-005056c00008.jpg \n", + " inflating: images/face/jiangwen/11fe84670-f929-11e8-8e1b-005056c00008.jpg \n", + " inflating: images/face/jiangwen/13a0d63f0-f929-11e8-9dce-005056c00008.png \n", + " inflating: images/face/jiangwen/13ac6eff0-f929-11e8-bd75-005056c00008.jpg \n", + " inflating: images/face/jiangwen/13b4791a2-f929-11e8-8555-005056c00008.jpg \n", + " inflating: images/face/jiangwen/13b938ec0-f929-11e8-85af-005056c00008.jpg \n", + " inflating: images/face/jiangwen/13bc1a3a2-f929-11e8-bdf1-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/13d309cf0-f929-11e8-855f-005056c00008.jpg \n", + " inflating: images/face/jiangwen/13f521d62-f929-11e8-a72b-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/140040840-f929-11e8-b9ec-005056c00008.jpg \n", + " inflating: images/face/jiangwen/14047c800-f929-11e8-9bbd-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1413cfd70-f929-11e8-a030-005056c00008.jpg \n", + " inflating: images/face/jiangwen/141625fc0-f929-11e8-bcf4-005056c00008.jpg \n", + " inflating: images/face/jiangwen/14eb97192-f929-11e8-a1ff-005056c00008.jpg \n", + " inflating: images/face/jiangwen/14f2308d2-f929-11e8-b45b-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/16c874bc0-f929-11e8-bad9-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/170bedaf0-f929-11e8-b1ba-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/1741224a2-f929-11e8-9e25-005056c00008.jpg \n", + " inflating: images/face/jiangwen/18207c00-f929-11e8-bee9-005056c00008.jpg \n", + " inflating: images/face/jiangwen/18358d170-f929-11e8-b2a3-005056c00008.jpg \n", + " inflating: images/face/jiangwen/185e2140-f929-11e8-9b08-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/18900b980-f929-11e8-b547-005056c00008.jpg \n", + " inflating: images/face/jiangwen/18b60540-f929-11e8-aefd-005056c00008.jpg \n", + " inflating: images/face/jiangwen/18e42b3d2-f929-11e8-ad0c-005056c00008.jpg \n", + " inflating: images/face/jiangwen/18f6fc222-f929-11e8-86ae-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1905985e2-f929-11e8-a7e1-005056c00008.png \n", + " inflating: images/face/jiangwen/191270f12-f929-11e8-8bc6-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/192c57cd2-f929-11e8-855d-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1930285d2-f929-11e8-a039-005056c00008.jpg \n", + " inflating: images/face/jiangwen/194b19562-f929-11e8-88e0-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1a2935b00-f929-11e8-8ef8-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1ad8ed12-f929-11e8-ab1f-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1b055d8d2-f929-11e8-a6e5-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1b18ece00-f929-11e8-8e00-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1b20584a2-f929-11e8-8681-005056c00008.png \n", + " inflating: images/face/jiangwen/1b56b44e2-f929-11e8-995c-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1b74c60f0-f929-11e8-8eea-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1b91699a2-f929-11e8-a82b-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1b9203e2-f929-11e8-970c-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1b9969f12-f929-11e8-8e46-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1ba5952d2-f929-11e8-bcf4-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1bc5be12-f929-11e8-860a-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1bd622600-f929-11e8-8623-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1bebb2652-f929-11e8-a1e8-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1c2128eb0-f929-11e8-8a05-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1c749bb0-f929-11e8-bfb9-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1cc97270-f929-11e8-8f86-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/1d3134792-f929-11e8-8194-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1d3725780-f929-11e8-b066-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1de1c040-f929-11e8-b9f1-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/1e2c7540-f929-11e8-b467-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1e78d7f0-f929-11e8-b323-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1e9fde062-f928-11e8-acbd-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1ea6c3292-f928-11e8-b3ee-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1ea95b392-f928-11e8-a520-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1eade4870-f929-11e8-befb-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1eb0ae392-f928-11e8-81ce-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1eb9ffa2-f929-11e8-a78c-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1f896b422-f929-11e8-b4fb-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1f9c8b152-f928-11e8-b6ca-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1fa115312-f928-11e8-91fb-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1faba8b62-f928-11e8-a0bc-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/1fb42ce30-f928-11e8-b69d-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1fb8be522-f928-11e8-8016-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1fc638070-f928-11e8-9d12-005056c00008.jpg \n", + " inflating: images/face/jiangwen/1fe84670-f929-11e8-8e1b-005056c00008.jpg \n", + " inflating: images/face/jiangwen/3a0d63f0-f929-11e8-9dce-005056c00008.png \n", + " inflating: images/face/jiangwen/3ac6eff0-f929-11e8-bd75-005056c00008.jpg \n", + " inflating: images/face/jiangwen/3b4791a2-f929-11e8-8555-005056c00008.jpg \n", + " inflating: images/face/jiangwen/3b938ec0-f929-11e8-85af-005056c00008.jpg \n", + " inflating: images/face/jiangwen/3bc1a3a2-f929-11e8-bdf1-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/3d309cf0-f929-11e8-855f-005056c00008.jpg \n", + " inflating: images/face/jiangwen/3f521d62-f929-11e8-a72b-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/40040840-f929-11e8-b9ec-005056c00008.jpg \n", + " inflating: images/face/jiangwen/4047c800-f929-11e8-9bbd-005056c00008.jpg \n", + " inflating: images/face/jiangwen/413cfd70-f929-11e8-a030-005056c00008.jpg \n", + " inflating: images/face/jiangwen/41625fc0-f929-11e8-bcf4-005056c00008.jpg \n", + " inflating: images/face/jiangwen/4eb97192-f929-11e8-a1ff-005056c00008.jpg \n", + " inflating: images/face/jiangwen/4f2308d2-f929-11e8-b45b-005056c00008.jpg \n", + " inflating: images/face/jiangwen/6c874bc0-f929-11e8-bad9-005056c00008.jpg \n", + " inflating: images/face/jiangwen/70bedaf0-f929-11e8-b1ba-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/741224a2-f929-11e8-9e25-005056c00008.jpg \n", + " inflating: images/face/jiangwen/8358d170-f929-11e8-b2a3-005056c00008.jpg \n", + " inflating: images/face/jiangwen/8900b980-f929-11e8-b547-005056c00008.jpg \n", + " inflating: images/face/jiangwen/8e42b3d2-f929-11e8-ad0c-005056c00008.jpg \n", + " inflating: images/face/jiangwen/8f6fc222-f929-11e8-86ae-005056c00008.jpg \n", + " inflating: images/face/jiangwen/905985e2-f929-11e8-a7e1-005056c00008.png \n", + " inflating: images/face/jiangwen/91270f12-f929-11e8-8bc6-005056c00008.jpeg \n", + " inflating: images/face/jiangwen/92c57cd2-f929-11e8-855d-005056c00008.jpg \n", + " inflating: images/face/jiangwen/930285d2-f929-11e8-a039-005056c00008.jpg \n", + " inflating: images/face/jiangwen/94b19562-f929-11e8-88e0-005056c00008.jpg \n", + " inflating: images/face/jiangwen/a2935b00-f929-11e8-8ef8-005056c00008.jpg \n", + " inflating: images/face/jiangwen/aff912d2-f929-11e8-8807-005056c00008.jpg \n", + " inflating: images/face/jiangwen/b055d8d2-f929-11e8-a6e5-005056c00008.jpg \n", + " inflating: images/face/jiangwen/b18ece00-f929-11e8-8e00-005056c00008.jpg \n", + " inflating: images/face/jiangwen/b20584a2-f929-11e8-8681-005056c00008.png \n", + " inflating: images/face/jiangwen/b56b44e2-f929-11e8-995c-005056c00008.jpg \n", + " inflating: images/face/jiangwen/b74c60f0-f929-11e8-8eea-005056c00008.jpg \n", + " inflating: images/face/jiangwen/b91699a2-f929-11e8-a82b-005056c00008.jpg \n", + " inflating: images/face/jiangwen/b9969f12-f929-11e8-8e46-005056c00008.jpg \n", + " inflating: images/face/jiangwen/bd622600-f929-11e8-8623-005056c00008.jpg \n", + " inflating: images/face/jiangwen/bebb2652-f929-11e8-a1e8-005056c00008.jpg \n", + " inflating: images/face/jiangwen/c2128eb0-f929-11e8-8a05-005056c00008.jpg \n", + " inflating: images/face/jiangwen/d3134792-f929-11e8-8194-005056c00008.jpg \n", + " inflating: images/face/jiangwen/d3725780-f929-11e8-b066-005056c00008.jpg \n", + " inflating: images/face/jiangwen/e9fde062-f928-11e8-acbd-005056c00008.jpg \n", + " inflating: images/face/jiangwen/ea6c3292-f928-11e8-b3ee-005056c00008.jpg \n", + " inflating: images/face/jiangwen/ea95b392-f928-11e8-a520-005056c00008.jpg \n", + " inflating: images/face/jiangwen/eade4870-f929-11e8-befb-005056c00008.jpg \n", + " inflating: images/face/jiangwen/eb0ae392-f928-11e8-81ce-005056c00008.jpg \n", + " inflating: images/face/jiangwen/f896b422-f929-11e8-b4fb-005056c00008.jpg \n", + " inflating: images/face/jiangwen/f97205d2-f928-11e8-9d91-005056c00008.jpg \n", + " inflating: images/face/jiangwen/f9c8b152-f928-11e8-b6ca-005056c00008.jpg \n", + " inflating: images/face/jiangwen/fa115312-f928-11e8-91fb-005056c00008.jpg \n", + " inflating: images/face/jiangwen/faba8b62-f928-11e8-a0bc-005056c00008.jpg \n", + " inflating: images/face/jiangwen/fb42ce30-f928-11e8-b69d-005056c00008.jpg \n", + " inflating: images/face/jiangwen/fb8be522-f928-11e8-8016-005056c00008.jpg \n", + " inflating: images/face/jiangwen/fc638070-f928-11e8-9d12-005056c00008.jpg \n", + " creating: images/face/pengyuyan/\n", + " inflating: images/face/pengyuyan/20181206160937.png \n", + " inflating: images/face/pengyuyan/20181206161041.png \n", + " inflating: images/face/pengyuyan/20181206161053.png \n", + " inflating: images/face/pengyuyan/20181206161115.png \n", + " inflating: images/face/pengyuyan/20181206161127.png \n", + " inflating: images/face/pengyuyan/20181206161153.png \n", + " inflating: images/face/pengyuyan/20181206161233.png \n", + " inflating: images/face/pengyuyan/20181206161255.png \n", + " inflating: images/face/pengyuyan/20181206161312.png \n", + " inflating: images/face/pengyuyan/20181206161331.png \n", + " inflating: images/face/pengyuyan/20181206161448.png \n", + " inflating: images/face/pengyuyan/20181206161502.png \n", + " inflating: images/face/pengyuyan/20181206161516.png \n", + " inflating: images/face/pengyuyan/20181206161530.png \n", + " inflating: images/face/pengyuyan/20181206161541.png \n", + " inflating: images/face/pengyuyan/20181206161554.png \n", + " inflating: images/face/pengyuyan/20181206161607.png \n", + " inflating: images/face/pengyuyan/20181206161634.png \n", + " inflating: images/face/pengyuyan/20181206161648.png \n", + " inflating: images/face/pengyuyan/20181206161702.png \n", + " inflating: images/face/pengyuyan/20181206161734.png \n", + " inflating: images/face/pengyuyan/20181206161755.png \n", + " inflating: images/face/pengyuyan/20181206161806.png \n", + " inflating: images/face/pengyuyan/20181206161828.png \n", + " inflating: images/face/pengyuyan/20181206161848.png \n", + " inflating: images/face/pengyuyan/20181206161903.png \n", + " inflating: images/face/pengyuyan/20181206161917.png \n", + " inflating: images/face/pengyuyan/20181206161943.png \n", + " inflating: images/face/pengyuyan/20181206161954.png \n", + " inflating: images/face/pengyuyan/20181206162010.png \n", + " inflating: images/face/pengyuyan/20181206162027.png \n", + " inflating: images/face/pengyuyan/20181206162042.png \n", + " inflating: images/face/pengyuyan/20181206162057.png \n", + " inflating: images/face/pengyuyan/20181206162112.png \n", + " inflating: images/face/pengyuyan/20181206162157.png \n", + " inflating: images/face/pengyuyan/20181206162211.png \n", + " inflating: images/face/pengyuyan/20181206162223.png \n", + " inflating: images/face/pengyuyan/20181206162234.png \n", + " inflating: images/face/pengyuyan/20181206162254.png \n", + " inflating: images/face/pengyuyan/20181206162311.png \n", + " inflating: images/face/pengyuyan/20181206162340.png \n", + " inflating: images/face/pengyuyan/20181206162351.png \n", + " inflating: images/face/pengyuyan/20181206162411.png \n", + " inflating: images/face/pengyuyan/20181206162423.png \n", + " inflating: images/face/pengyuyan/20181206162442.png \n", + " inflating: images/face/pengyuyan/20181206162454.png \n", + " inflating: images/face/pengyuyan/20181206162507.png \n", + " inflating: images/face/pengyuyan/20181206162522.png \n", + " inflating: images/face/pengyuyan/20181206162540.png \n", + " inflating: images/face/pengyuyan/20181206162552.png \n", + " inflating: images/face/pengyuyan/20181206162602.png \n", + " inflating: images/face/pengyuyan/20181206162613.png \n", + " inflating: images/face/pengyuyan/20181206162625.png \n", + " inflating: images/face/pengyuyan/20181206162641.png \n", + " inflating: images/face/pengyuyan/20181206162654.png \n", + " inflating: images/face/pengyuyan/20181206162707.png \n", + " inflating: images/face/pengyuyan/20181206162717.png \n", + " inflating: images/face/pengyuyan/20181206162732.png \n", + " inflating: images/face/pengyuyan/20181206162749.png \n", + " inflating: images/face/pengyuyan/20181206162804.png \n", + " inflating: images/face/pengyuyan/20181206162816.png \n", + " inflating: images/face/pengyuyan/20181206162836.png \n", + " inflating: images/face/pengyuyan/20181206162850.png \n", + " inflating: images/face/pengyuyan/20181206162904.png \n", + " inflating: images/face/pengyuyan/20181206162915.png \n", + " inflating: images/face/pengyuyan/20181206162932.png \n", + " inflating: images/face/pengyuyan/20181206162941.png \n", + " inflating: images/face/pengyuyan/20181206162956.png \n", + " inflating: images/face/pengyuyan/20181206163011.png \n", + " inflating: images/face/pengyuyan/20181206163213.png \n", + " inflating: images/face/pengyuyan/20181206163224.png \n", + " inflating: images/face/pengyuyan/20181206163238.png \n", + " inflating: images/face/pengyuyan/20181206163250.png \n", + " inflating: images/face/pengyuyan/20181206163306.png \n", + " inflating: images/face/pengyuyan/20181206163318.png \n", + " inflating: images/face/pengyuyan/20181206163331.png \n", + " inflating: images/face/pengyuyan/20181206163344.png \n", + " inflating: images/face/pengyuyan/20181206163359.png \n", + " inflating: images/face/pengyuyan/20181206163412.png \n", + " inflating: images/face/pengyuyan/20181206163439.png \n", + " inflating: images/face/pengyuyan/20181206163452.png \n", + " inflating: images/face/pengyuyan/20181206164423.png \n", + " inflating: images/face/pengyuyan/20181206164436.png \n", + " inflating: images/face/pengyuyan/20181206164450.png \n", + " inflating: images/face/pengyuyan/20181206164502.png \n", + " inflating: images/face/pengyuyan/20181206164513.png \n", + " inflating: images/face/pengyuyan/20181206164526.png \n", + " inflating: images/face/pengyuyan/20181206164539.png \n", + " inflating: images/face/pengyuyan/20181206164550.png \n", + " inflating: images/face/pengyuyan/20181206164628.png \n", + " inflating: images/face/pengyuyan/20181206164640.png \n", + " inflating: images/face/pengyuyan/20181206164655.png \n", + " inflating: images/face/pengyuyan/20181206164707.png \n", + " inflating: images/face/pengyuyan/20181206164718.png \n", + " inflating: images/face/pengyuyan/20181206164727.png \n", + " inflating: images/face/pengyuyan/20181206164742.png \n", + " inflating: images/face/pengyuyan/20181206164753.png \n", + " inflating: images/face/pengyuyan/20181206164804.png \n", + " inflating: images/face/pengyuyan/20181206164816.png \n", + " inflating: images/face/pengyuyan/20181206164827.png \n", + " inflating: images/face/pengyuyan/20181206164840.png \n", + " inflating: images/face/pengyuyan/20181206164853.png \n", + " inflating: images/face/pengyuyan/20181206164912.png \n", + " inflating: images/face/pengyuyan/20181206165014.png \n", + " inflating: images/face/pengyuyan/20181206165044.png \n", + " inflating: images/face/pengyuyan/20181206165057.png \n", + " inflating: images/face/pengyuyan/20181206165108.png \n", + " inflating: images/face/pengyuyan/20181206165118.png \n", + " inflating: images/face/pengyuyan/20181206165131.png \n", + " inflating: images/face/pengyuyan/20181206165143.png \n", + " inflating: images/face/pengyuyan/20181206165157.png \n", + " inflating: images/face/pengyuyan/20181206165211.png \n", + " inflating: images/face/pengyuyan/20181206165238.png \n", + " inflating: images/face/pengyuyan/20181206165249.png \n", + " creating: images/face/zhangyan/\n", + "images/face/zhangyan/微信图片_20200610094637.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094637.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094637.jpg \n", + "images/face/zhangyan/微信图片_20200610094707.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094707.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094707.jpg \n", + "images/face/zhangyan/微信图片_20200610094710.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094710.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094710.jpg \n", + "images/face/zhangyan/微信图片_20200610094714.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094714.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094714.jpg \n", + "images/face/zhangyan/微信图片_20200610094717.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094717.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094717.jpg \n", + "images/face/zhangyan/微信图片_20200610094720.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094720.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094720.jpg \n", + "images/face/zhangyan/微信图片_20200610094723.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094723.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094723.jpg \n", + "images/face/zhangyan/微信图片_20200610094727.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094727.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094727.jpg \n", + "images/face/zhangyan/微信图片_20200610094730.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094730.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094730.jpg \n", + "images/face/zhangyan/微信图片_20200610094733.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094733.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094733.jpg \n", + "images/face/zhangyan/微信图片_20200610094736.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094736.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094736.jpg \n", + "images/face/zhangyan/微信图片_20200610094739.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094739.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094739.jpg \n", + "images/face/zhangyan/微信图片_20200610094742.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094742.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094742.jpg \n", + "images/face/zhangyan/微信图片_20200610094745.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094745.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094745.jpg \n", + "images/face/zhangyan/微信图片_20200610094749.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094749.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094749.jpg \n", + "images/face/zhangyan/微信图片_20200610094752.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094752.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094752.jpg \n", + "images/face/zhangyan/微信图片_20200610094755.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094755.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094755.jpg \n", + "images/face/zhangyan/微信图片_20200610094759.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094759.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094759.jpg \n", + "images/face/zhangyan/微信图片_20200610094801.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094801.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094801.jpg \n", + "images/face/zhangyan/微信图片_20200610094804.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094804.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094804.jpg \n", + "images/face/zhangyan/微信图片_20200610094807.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094807.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094807.jpg \n", + "images/face/zhangyan/微信图片_20200610094810.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094810.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094810.jpg \n", + "images/face/zhangyan/微信图片_20200610094813.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094813.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094813.jpg \n", + "images/face/zhangyan/微信图片_20200610094816.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094816.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094816.jpg \n", + "images/face/zhangyan/微信图片_20200610094818.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094818.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094818.jpg \n", + "images/face/zhangyan/微信图片_20200610094821.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094821.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094821.jpg \n", + "images/face/zhangyan/微信图片_20200610094824.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094824.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094824.jpg \n", + "images/face/zhangyan/微信图片_20200610094827.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094827.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094827.jpg \n", + "images/face/zhangyan/微信图片_20200610094829.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094829.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094829.jpg \n", + "images/face/zhangyan/微信图片_20200610094832.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094832.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094832.jpg \n", + "images/face/zhangyan/微信图片_20200610094835.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094835.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094835.jpg \n", + "images/face/zhangyan/微信图片_20200610094838.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094838.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094838.jpg \n", + "images/face/zhangyan/微信图片_20200610094841.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094841.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094841.jpg \n", + "images/face/zhangyan/微信图片_20200610094843.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094843.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094843.jpg \n", + "images/face/zhangyan/微信图片_20200610094846.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094846.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094846.jpg \n", + "images/face/zhangyan/微信图片_20200610094848.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094848.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094848.jpg \n", + "images/face/zhangyan/微信图片_20200610094852.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094852.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094852.jpg \n", + "images/face/zhangyan/微信图片_20200610094854.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094854.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094854.jpg \n", + "images/face/zhangyan/微信图片_20200610094857.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094857.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094857.jpg \n", + "images/face/zhangyan/微信图片_20200610094900.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094900.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094900.jpg \n", + "images/face/zhangyan/微信图片_20200610094902.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094902.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094902.jpg \n", + "images/face/zhangyan/微信图片_20200610094905.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094905.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094905.jpg \n", + "images/face/zhangyan/微信图片_20200610094907.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094907.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094907.jpg \n", + "images/face/zhangyan/微信图片_20200610094910.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094910.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094910.jpg \n", + "images/face/zhangyan/微信图片_20200610094913.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094913.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094913.jpg \n", + "images/face/zhangyan/微信图片_20200610094917.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094917.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094917.jpg \n", + "images/face/zhangyan/微信图片_20200610094920.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094920.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094920.jpg \n", + "images/face/zhangyan/微信图片_20200610094922.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094922.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094922.jpg \n", + "images/face/zhangyan/微信图片_20200610094925.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094925.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094925.jpg \n", + "images/face/zhangyan/微信图片_20200610094927.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094927.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094927.jpg \n", + "images/face/zhangyan/微信图片_20200610094931.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094931.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094931.jpg \n", + "images/face/zhangyan/微信图片_20200610094933.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094933.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094933.jpg \n", + "images/face/zhangyan/微信图片_20200610094936.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094936.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094936.jpg \n", + "images/face/zhangyan/微信图片_20200610094940.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094940.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094940.jpg \n", + "images/face/zhangyan/微信图片_20200610094942.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094942.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094942.jpg \n", + "images/face/zhangyan/微信图片_20200610094945.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094945.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094945.jpg \n", + "images/face/zhangyan/微信图片_20200610094947.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094947.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094947.jpg \n", + "images/face/zhangyan/微信图片_20200610094949.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094949.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094949.jpg \n", + "images/face/zhangyan/微信图片_20200610094952.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094952.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094952.jpg \n", + "images/face/zhangyan/微信图片_20200610094954.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094954.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094954.jpg \n", + "images/face/zhangyan/微信图片_20200610094957.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610094957.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610094957.jpg \n", + "images/face/zhangyan/微信图片_20200610095000.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095000.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095000.jpg \n", + "images/face/zhangyan/微信图片_20200610095002.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095002.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095002.jpg \n", + "images/face/zhangyan/微信图片_20200610095004.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095004.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095004.jpg \n", + "images/face/zhangyan/微信图片_20200610095007.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095007.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095007.jpg \n", + "images/face/zhangyan/微信图片_20200610095010.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095010.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095010.jpg \n", + "images/face/zhangyan/微信图片_20200610095013.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095013.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095013.jpg \n", + "images/face/zhangyan/微信图片_20200610095016.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095016.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095016.jpg \n", + "images/face/zhangyan/微信图片_20200610095018.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095018.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095018.jpg \n", + "images/face/zhangyan/微信图片_20200610095020.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095020.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095020.jpg \n", + "images/face/zhangyan/微信图片_20200610095023.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095023.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095023.jpg \n", + "images/face/zhangyan/微信图片_20200610095025.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095025.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095025.jpg \n", + "images/face/zhangyan/微信图片_20200610095028.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095028.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095028.jpg \n", + "images/face/zhangyan/微信图片_20200610095030.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095030.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095030.jpg \n", + "images/face/zhangyan/微信图片_20200610095033.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095033.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095033.jpg \n", + "images/face/zhangyan/微信图片_20200610095036.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095036.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095036.jpg \n", + "images/face/zhangyan/微信图片_20200610095039.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095039.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095039.jpg \n", + "images/face/zhangyan/微信图片_20200610095041.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095041.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095041.jpg \n", + "images/face/zhangyan/微信图片_20200610095044.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095044.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095044.jpg \n", + "images/face/zhangyan/微信图片_20200610095046.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095046.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095046.jpg \n", + "images/face/zhangyan/微信图片_20200610095049.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095049.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095049.jpg \n", + "images/face/zhangyan/微信图片_20200610095052.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095052.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095052.jpg \n", + "images/face/zhangyan/微信图片_20200610095054.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095054.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095054.jpg \n", + "images/face/zhangyan/微信图片_20200610095057.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095057.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095057.jpg \n", + "images/face/zhangyan/微信图片_20200610095059.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095059.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095059.jpg \n", + "images/face/zhangyan/微信图片_20200610095102.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095102.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095102.jpg \n", + "images/face/zhangyan/微信图片_20200610095104.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095104.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095104.jpg \n", + "images/face/zhangyan/微信图片_20200610095107.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095107.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095107.jpg \n", + "images/face/zhangyan/微信图片_20200610095110.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095110.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095110.jpg \n", + "images/face/zhangyan/微信图片_20200610095113.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095113.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095113.jpg \n", + "images/face/zhangyan/微信图片_20200610095116.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095116.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095116.jpg \n", + "images/face/zhangyan/微信图片_20200610095118.jpg: mismatching \"local\" filename (images/face/zhangyan/х╛оф┐бхЫ╛чЙЗ_20200610095118.jpg),\n", + " continuing with \"central\" filename version\n", + " inflating: images/face/zhangyan/微信图片_20200610095118.jpg \n", + " creating: images/face/zhangziyi/\n", + " inflating: images/face/zhangziyi/20181206112348.png \n", + " inflating: images/face/zhangziyi/20181206112504.png \n", + " inflating: images/face/zhangziyi/20181206112529.png \n", + " inflating: images/face/zhangziyi/20181206131856.png \n", + " inflating: images/face/zhangziyi/20181206131927.png \n", + " inflating: images/face/zhangziyi/20181206131958.png \n", + " inflating: images/face/zhangziyi/20181206132022.png \n", + " inflating: images/face/zhangziyi/20181206132043.png \n", + " inflating: images/face/zhangziyi/20181206132102.png \n", + " inflating: images/face/zhangziyi/20181206132122.png \n", + " inflating: images/face/zhangziyi/20181206132141.png \n", + " inflating: images/face/zhangziyi/20181206132200.png \n", + " inflating: images/face/zhangziyi/20181206132220.png \n", + " inflating: images/face/zhangziyi/20181206132239.png \n", + " inflating: images/face/zhangziyi/20181206132255.png \n", + " inflating: images/face/zhangziyi/20181206132327.png \n", + " inflating: images/face/zhangziyi/20181206132346.png \n", + " inflating: images/face/zhangziyi/20181206132406.png \n", + " inflating: images/face/zhangziyi/20181206132425.png \n", + " inflating: images/face/zhangziyi/20181206132442.png \n", + " inflating: images/face/zhangziyi/20181206132514.png \n", + " inflating: images/face/zhangziyi/20181206132533.png \n", + " inflating: images/face/zhangziyi/20181206132554.png \n", + " inflating: images/face/zhangziyi/20181206132616.png \n", + " inflating: images/face/zhangziyi/20181206132629.png \n", + " inflating: images/face/zhangziyi/20181206132648.png \n", + " inflating: images/face/zhangziyi/20181206132700.png \n", + " inflating: images/face/zhangziyi/20181206132718.png \n", + " inflating: images/face/zhangziyi/20181206132738.png \n", + " inflating: images/face/zhangziyi/20181206132755.png \n", + " inflating: images/face/zhangziyi/20181206132812.png \n", + " inflating: images/face/zhangziyi/20181206132904.png \n", + " inflating: images/face/zhangziyi/20181206132919.png \n", + " inflating: images/face/zhangziyi/20181206132933.png \n", + " inflating: images/face/zhangziyi/20181206133002.png \n", + " inflating: images/face/zhangziyi/20181206133020.png \n", + " inflating: images/face/zhangziyi/20181206133344.png \n", + " inflating: images/face/zhangziyi/20181206133415.png \n", + " inflating: images/face/zhangziyi/20181206133521.png \n", + " inflating: images/face/zhangziyi/20181206133539.png \n", + " inflating: images/face/zhangziyi/20181206133619.png \n", + " inflating: images/face/zhangziyi/20181206133700.png \n", + " inflating: images/face/zhangziyi/20181206133720.png \n", + " inflating: images/face/zhangziyi/20181206133815.png \n", + " inflating: images/face/zhangziyi/20181206133833.png \n", + " inflating: images/face/zhangziyi/20181206133913.png \n", + " inflating: images/face/zhangziyi/20181206133927.png \n", + " inflating: images/face/zhangziyi/20181206134004.png \n", + " inflating: images/face/zhangziyi/20181206134023.png \n", + " inflating: images/face/zhangziyi/20181206134043.png \n", + " inflating: images/face/zhangziyi/20181206134103.png \n", + " inflating: images/face/zhangziyi/20181206134123.png \n", + " inflating: images/face/zhangziyi/20181206134152.png \n", + " inflating: images/face/zhangziyi/20181206134213.png \n", + " inflating: images/face/zhangziyi/20181206134229.png \n", + " inflating: images/face/zhangziyi/20181206134251.png \n", + " inflating: images/face/zhangziyi/20181206134403.png \n", + " inflating: images/face/zhangziyi/20181206134456.png \n", + " inflating: images/face/zhangziyi/20181206134529.png \n", + " inflating: images/face/zhangziyi/20181206134542.png \n", + " inflating: images/face/zhangziyi/20181206134557.png \n", + " inflating: images/face/zhangziyi/20181206134622.png \n", + " inflating: images/face/zhangziyi/20181206134640.png \n", + " inflating: images/face/zhangziyi/20181206134919.png \n", + " inflating: images/face/zhangziyi/20181206134938.png \n", + " inflating: images/face/zhangziyi/20181206135034.png \n", + " inflating: images/face/zhangziyi/20181206135108.png \n", + " inflating: images/face/zhangziyi/20181206135126.png \n", + " inflating: images/face/zhangziyi/20181206135221.png \n", + " inflating: images/face/zhangziyi/20181206135321.png \n", + " inflating: images/face/zhangziyi/20181206135607.png \n", + " inflating: images/face/zhangziyi/20181206135812.png \n", + " inflating: images/face/zhangziyi/20181206135830.png \n", + " inflating: images/face/zhangziyi/20181206135919.png \n", + " inflating: images/face/zhangziyi/20181206140000.png \n", + " inflating: images/face/zhangziyi/20181206143714.png \n", + " inflating: images/face/zhangziyi/20181206143730.png \n", + " inflating: images/face/zhangziyi/20181206143748.png \n", + " inflating: images/face/zhangziyi/20181206144418.png \n", + " inflating: images/face/zhangziyi/20181206144436.png \n", + " inflating: images/face/zhangziyi/20181206144453.png \n", + " inflating: images/face/zhangziyi/20181206144613.png \n", + " inflating: images/face/zhangziyi/20181206144652.png \n", + " inflating: images/face/zhangziyi/20181206144721.png \n", + " inflating: images/face/zhangziyi/20181206144813.png \n", + " inflating: images/face/zhangziyi/20181206144850.png \n", + " inflating: images/face/zhangziyi/20181206144938.png \n", + " inflating: images/face/zhangziyi/20181206145015.png \n", + " inflating: images/face/zhangziyi/20181206145059.png \n", + " inflating: images/face/zhangziyi/20181206145110.png \n", + " inflating: images/face/zhangziyi/20181206145122.png \n", + " inflating: images/face/zhangziyi/20181206145155.png \n", + " inflating: images/face/zhangziyi/20181206145232.png \n", + " inflating: images/face/zhangziyi/20181206145247.png \n", + " inflating: images/face/zhangziyi/20181206145308.png \n", + " inflating: images/face/zhangziyi/20181206145333.png \n", + " inflating: images/face/zhangziyi/20181206145348.png \n", + " inflating: images/face/zhangziyi/20181206145406.png \n", + " inflating: images/face/zhangziyi/20181206145427.png \n", + " inflating: images/face/zhangziyi/20181206145456.png \n", + " creating: images/face/zhaoliying/\n", + " inflating: images/face/zhaoliying/159116065710.jpg \n", + " inflating: images/face/zhaoliying/159116065712.jpg \n", + " inflating: images/face/zhaoliying/159116065716.jpg \n", + " inflating: images/face/zhaoliying/159116065718.jpg \n", + " inflating: images/face/zhaoliying/15911606572.jpg \n", + " inflating: images/face/zhaoliying/159116065721.jpg \n", + " inflating: images/face/zhaoliying/159116065722.jpg \n", + " inflating: images/face/zhaoliying/159116065726.jpg \n", + " inflating: images/face/zhaoliying/159116065729.jpg \n", + " inflating: images/face/zhaoliying/159116065730.jpg \n", + " inflating: images/face/zhaoliying/159116065733.jpg \n", + " inflating: images/face/zhaoliying/159116065736.jpg \n", + " inflating: images/face/zhaoliying/15911606574.jpg \n", + " inflating: images/face/zhaoliying/159116065740.jpg \n", + " inflating: images/face/zhaoliying/159116065741.jpg \n", + " inflating: images/face/zhaoliying/159116065742.jpg \n", + " inflating: images/face/zhaoliying/159116065743.jpg \n", + " inflating: images/face/zhaoliying/159116065744.jpg \n", + " inflating: images/face/zhaoliying/159116065745.jpg \n", + " inflating: images/face/zhaoliying/159116065751.jpg \n", + " inflating: images/face/zhaoliying/159116065753.jpg \n", + " inflating: images/face/zhaoliying/159116065754.jpg \n", + " inflating: images/face/zhaoliying/15911606577.jpg \n", + " inflating: images/face/zhaoliying/15911606578.jpg \n", + " inflating: images/face/zhaoliying/15911606579.jpg \n", + " inflating: images/face/zhaoliying/1591160658100.jpg \n", + " inflating: images/face/zhaoliying/1591160658101.jpg \n", + " inflating: images/face/zhaoliying/1591160658102.jpg \n", + " inflating: images/face/zhaoliying/1591160658103.jpg \n", + " inflating: images/face/zhaoliying/1591160658104.jpg \n", + " inflating: images/face/zhaoliying/1591160658105.jpg \n", + " inflating: images/face/zhaoliying/1591160658106.jpg \n", + " inflating: images/face/zhaoliying/1591160658108.jpg \n", + " inflating: images/face/zhaoliying/1591160658109.jpg \n", + " inflating: images/face/zhaoliying/1591160658112.jpg \n", + " inflating: images/face/zhaoliying/1591160658113.jpg \n", + " inflating: images/face/zhaoliying/1591160658114.jpg \n", + " inflating: images/face/zhaoliying/159116065857.jpg \n", + " inflating: images/face/zhaoliying/159116065859.jpg \n", + " inflating: images/face/zhaoliying/159116065860.jpg \n", + " inflating: images/face/zhaoliying/159116065861.jpg \n", + " inflating: images/face/zhaoliying/159116065862.jpg \n", + " inflating: images/face/zhaoliying/159116065866.jpg \n", + " inflating: images/face/zhaoliying/159116065867.jpg \n", + " inflating: images/face/zhaoliying/159116065870.jpg \n", + " inflating: images/face/zhaoliying/159116065871.jpg \n", + " inflating: images/face/zhaoliying/159116065873.jpg \n", + " inflating: images/face/zhaoliying/159116065874.jpg \n", + " inflating: images/face/zhaoliying/159116065875.jpg \n", + " inflating: images/face/zhaoliying/159116065876.jpg \n", + " inflating: images/face/zhaoliying/159116065877.jpg \n", + " inflating: images/face/zhaoliying/159116065878.jpg \n", + " inflating: images/face/zhaoliying/159116065879.jpg \n", + " inflating: images/face/zhaoliying/159116065883.jpg \n", + " inflating: images/face/zhaoliying/159116065886.jpg \n", + " inflating: images/face/zhaoliying/159116065888.jpg \n", + " inflating: images/face/zhaoliying/159116065890.jpg \n", + " inflating: images/face/zhaoliying/159116065891.jpg \n", + " inflating: images/face/zhaoliying/159116065892.jpg \n", + " inflating: images/face/zhaoliying/159116065893.jpg \n", + " inflating: images/face/zhaoliying/159116065894.jpg \n", + " inflating: images/face/zhaoliying/159116065896.jpg \n", + " inflating: images/face/zhaoliying/159116065898.jpg \n", + " inflating: images/face/zhaoliying/159116065899.jpg \n", + " inflating: images/face/zhaoliying/1591160659116.jpg \n", + " inflating: images/face/zhaoliying/1591160659117.jpg \n", + " inflating: images/face/zhaoliying/1591160659118.jpg \n", + " inflating: images/face/zhaoliying/1591160659119.jpg \n", + " inflating: images/face/zhaoliying/1591160659120.jpg \n", + " inflating: images/face/zhaoliying/1591160659121.jpg \n", + " inflating: images/face/zhaoliying/1591160659122.jpg \n", + " inflating: images/face/zhaoliying/1591160659124.jpg \n", + " inflating: images/face/zhaoliying/1591160659125.jpg \n", + " inflating: images/face/zhaoliying/1591160659126.jpg \n", + " inflating: images/face/zhaoliying/1591160659128.jpg \n", + " inflating: images/face/zhaoliying/1591160659130.jpg \n", + " inflating: images/face/zhaoliying/1591160659131.jpg \n", + " inflating: images/face/zhaoliying/1591160659135.jpg \n", + " inflating: images/face/zhaoliying/1591160659137.jpg \n", + " inflating: images/face/zhaoliying/1591160659138.jpg \n", + " inflating: images/face/zhaoliying/15911621701.jpg \n", + " inflating: images/face/zhaoliying/159116217110.jpg \n", + " inflating: images/face/zhaoliying/159116217111.jpg \n", + " inflating: images/face/zhaoliying/159116217112.jpg \n", + " inflating: images/face/zhaoliying/159116217113.jpg \n", + " inflating: images/face/zhaoliying/15911621712.jpg \n", + " inflating: images/face/zhaoliying/15911621713.jpg \n", + " inflating: images/face/zhaoliying/15911621714.jpg \n", + " inflating: images/face/zhaoliying/15911621716.jpg \n", + " inflating: images/face/zhaoliying/15911621717.jpg \n", + " inflating: images/face/zhaoliying/15911621718.jpg \n", + " inflating: images/face/zhaoliying/15911621719.jpg \n", + " inflating: images/face/zhaoliying/159116217214.jpg \n", + " inflating: images/face/zhaoliying/159116217215.jpg \n", + " inflating: images/face/zhaoliying/159116217216.jpg \n", + " inflating: images/face/zhaoliying/159116217217.jpg \n", + " inflating: images/face/zhaoliying/159116217218.jpg \n", + " inflating: images/face/zhaoliying/159116217219.jpg \n", + " inflating: images/face/zhaoliying/159116217220.jpg \n", + " inflating: images/face/zhaoliying/159116217221.jpg \n", + " inflating: images/face/zhaoliying/159116217223.jpg \n", + " inflating: images/face/zhaoliying/159116217224.jpg \n", + " inflating: images/face/zhaoliying/159116217225.jpg \n", + " inflating: images/face/zhaoliying/159116217226.jpg \n", + " inflating: images/face/zhaoliying/159116217227.jpg \n", + " inflating: images/face/zhaoliying/159116217228.jpg \n", + " inflating: images/face/zhaoliying/159116217229.jpg \n", + " inflating: images/face/zhaoliying/159116217330.jpg \n", + " inflating: images/face/zhaoliying/159116217331.jpg \n", + " inflating: images/face/zhaoliying/159116217332.jpg \n", + " inflating: images/face/zhaoliying/159116217333.jpg \n", + " inflating: images/face/zhaoliying/159116217334.jpg \n", + " inflating: images/face/zhaoliying/159116217335.jpg \n", + " inflating: images/face/zhaoliying/159116217336.jpg \n", + " inflating: images/face/zhaoliying/159116217337.jpg \n", + " inflating: images/face/zhaoliying/159116217338.jpg \n", + " inflating: images/face/zhaoliying/159116217339.jpg \n", + " inflating: images/face/zhaoliying/159116217340.jpg \n", + " inflating: images/face/zhaoliying/159116217341.jpg \n", + " inflating: images/face/zhaoliying/159116217342.jpg \n", + " inflating: images/face/zhaoliying/159116217343.jpg \n", + " inflating: images/face/zhaoliying/159116217444.jpg \n", + " inflating: images/face/zhaoliying/159116217445.jpg \n", + " inflating: images/face/zhaoliying/159116217446.jpg \n", + " inflating: images/face/zhaoliying/159116217447.jpg \n", + " inflating: images/face/zhaoliying/159116217448.jpg \n", + " inflating: images/face/zhaoliying/159116217449.jpg \n", + " inflating: images/face/zhaoliying/159116217450.jpg \n", + " inflating: images/face/zhaoliying/159116217452.jpg \n", + " inflating: images/face/zhaoliying/159116217453.jpg \n", + " inflating: images/face/zhaoliying/159116217454.jpg \n", + " inflating: images/face/zhaoliying/159116217455.jpg \n", + " inflating: images/face/zhaoliying/159116217456.jpg \n", + " inflating: images/face/zhaoliying/159116217457.jpg \n", + " inflating: images/face/zhaoliying/159116217558.jpg \n", + " inflating: images/face/zhaoliying/159116217559.jpg \n", + " inflating: images/face/zhaoliying/159116217560.jpg \n", + " inflating: images/face/zhaoliying/159116217561.jpg \n", + " inflating: images/face/zhaoliying/159116217562.jpg \n", + " inflating: images/face/zhaoliying/159116217563.jpg \n", + " inflating: images/face/zhaoliying/159116217564.jpg \n", + " inflating: images/face/zhaoliying/159116217565.jpg \n", + " inflating: images/face/zhaoliying/159116217566.jpg \n", + " inflating: images/face/zhaoliying/159116217567.jpg \n", + " inflating: images/face/zhaoliying/159116217568.jpg \n", + " inflating: images/face/zhaoliying/159116217669.jpg \n", + " inflating: images/face/zhaoliying/159116217670.jpg \n", + " inflating: images/face/zhaoliying/159116217671.jpg \n", + " inflating: images/face/zhaoliying/159116217672.jpg \n", + " inflating: images/face/zhaoliying/159116217673.jpg \n", + " inflating: images/face/zhaoliying/159116217674.jpg \n", + " inflating: images/face/zhaoliying/159116217675.jpg \n", + " inflating: images/face/zhaoliying/159116217676.jpg \n", + " inflating: images/face/zhaoliying/159116217677.jpg \n", + " inflating: images/face/zhaoliying/159116217678.jpg \n", + " inflating: images/face/zhaoliying/159116217679.jpg \n", + " inflating: images/face/zhaoliying/159116217680.jpg \n" + ] + } + ], + "source": [ + "!unzip /home/aistudio/data/data39602/images.zip" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 将图片的类别以及图片的路径存储到一个txt文档中" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:27:39.211730Z", + "iopub.status.busy": "2022-05-03T14:27:39.211323Z", + "iopub.status.idle": "2022-05-03T14:27:39.218007Z", + "shell.execute_reply": "2022-05-03T14:27:39.217556Z", + "shell.execute_reply.started": "2022-05-03T14:27:39.211697Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "import os\n", + "import json\n", + "\n", + "\n", + "def datagenerator(datasets_path):\n", + " total_save = {}\n", + " # all classes\n", + " classes_name = os.listdir(datasets_path)\n", + " # classes names\n", + " classes_name = sorted(classes_name)\n", + " print(\"classes name：\", classes_name)\n", + "\n", + " list_file = open('cls_train.txt', 'w')\n", + " for cls_id, type_name in enumerate(classes_name):\n", + " total_save[cls_id] = type_name\n", + " photos_path = os.path.join(datasets_path, type_name)\n", + " if not os.path.isdir(photos_path):\n", + " continue\n", + " # all images\n", + " photos_name = os.listdir(photos_path)\n", + "\n", + " for photo_name in photos_name:\n", + " # image id + image path\n", + " list_file.write(\n", + " str(cls_id) + \";\" + '%s' % (os.path.join(datasets_path, type_name, photo_name)))\n", + " list_file.write('\\n')\n", + " list_file.close()\n", + " # save json\n", + " dict_json = json.dumps(total_save)\n", + " with open('classes.json', 'w+') as file:\n", + " file.write(dict_json)\n" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T09:54:11.245844Z", + "iopub.status.busy": "2022-05-03T09:54:11.245237Z", + "iopub.status.idle": "2022-05-03T09:54:11.255881Z", + "shell.execute_reply": "2022-05-03T09:54:11.254843Z", + "shell.execute_reply.started": "2022-05-03T09:54:11.245801Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "classes name： ['dilireba', 'jiangwen', 'pengyuyan', 'zhangyan', 'zhangziyi', 'zhaoliying']\n" + ] + } + ], + "source": [ + "datasets_path = \"./images/face/\"\n", + "datagenerator(datasets_path=datasets_path)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 进行自定义数据集，构建一个三元组的数据集" + ] + }, + { + "cell_type": "code", + "execution_count": 21, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:52:48.646072Z", + "iopub.status.busy": "2022-05-03T14:52:48.645721Z", + "iopub.status.idle": "2022-05-03T14:52:48.653365Z", + "shell.execute_reply": "2022-05-03T14:52:48.652801Z", + "shell.execute_reply.started": "2022-05-03T14:52:48.646043Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "import numpy as np\n", + "from PIL import Image\n", + "\n", + "\n", + "def cvtColor(image):\n", + " if len(np.shape(image)) == 3 and np.shape(image)[2] == 3:\n", + " return image \n", + " else:\n", + " image = image.convert('RGB')\n", + " return image \n", + "\n", + "\n", + "def resize_image(image, size, letterbox_image):\n", + " iw, ih = image.size\n", + " w, h = size\n", + " if letterbox_image:\n", + " scale = min(w/iw, h/ih)\n", + " nw = int(iw*scale)\n", + " nh = int(ih*scale)\n", + "\n", + " image = image.resize((nw,nh), Image.BICUBIC)\n", + " new_image = Image.new('RGB', size, (128,128,128))\n", + " new_image.paste(image, ((w-nw)//2, (h-nh)//2))\n", + " else:\n", + " new_image = image.resize((w, h), Image.BICUBIC)\n", + " return new_image\n", + "\n", + "\n", + "def preprocess_input(image):\n", + " # 255 ---> 1\n", + " image /= 255.0 \n", + " return image\n" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:49:47.277181Z", + "iopub.status.busy": "2022-05-03T14:49:47.276832Z", + "iopub.status.idle": "2022-05-03T14:49:48.361388Z", + "shell.execute_reply": "2022-05-03T14:49:48.360537Z", + "shell.execute_reply.started": "2022-05-03T14:49:47.277153Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "import random\n", + "\n", + "import numpy as np\n", + "from PIL import Image\n", + "from paddle.io import Dataset\n", + "\n", + "\n", + "class SearchDataset(Dataset):\n", + " def __init__(self, input_shape, lines, num_classes, random):\n", + " self.input_shape = input_shape\n", + " self.lines = lines\n", + " self.length = len(lines)\n", + " self.num_classes = num_classes\n", + " self.random = random\n", + " # path and label\n", + " self.paths = []\n", + " self.labels = []\n", + " # load data\n", + " self.load_dataset()\n", + " \n", + " def __len__(self):\n", + " return self.length\n", + "\n", + " def __getitem__(self, index):\n", + " images = np.zeros((3, 3, self.input_shape[0], self.input_shape[1]))\n", + " labels = np.zeros((3))\n", + "\n", + " # 获取一个类别\n", + " c = random.randint(0, self.num_classes - 1)\n", + " selected_path = self.paths[self.labels[:] == c]\n", + " while len(selected_path) < 2:\n", + " c = random.randint(0, self.num_classes - 1)\n", + " selected_path = self.paths[self.labels[:] == c]\n", + "\n", + " # 随机选择两张\n", + " image_indexes = np.random.choice(range(0, len(selected_path)), 2)\n", + " # 处理第一张图片\n", + " # 打开图片并放入矩阵\n", + " image = cvtColor(Image.open(selected_path[image_indexes[0]]))\n", + " # 翻转图像\n", + " if self.rand() < 0.5 and self.random:\n", + " image = image.transpose(Image.FLIP_LEFT_RIGHT)\n", + " image = resize_image(image, [self.input_shape[1], self.input_shape[0]], letterbox_image = True)\n", + " # 图像预处理\n", + " image = preprocess_input(np.array(image, dtype='float32'))\n", + " image = np.transpose(image, [2, 0, 1])\n", + " # 存储图像以及标签label\n", + " images[0, :, :, :] = image\n", + " labels[0] = c\n", + "\n", + " # 处理第二张图片\n", + " image = cvtColor(Image.open(selected_path[image_indexes[1]]))\n", + " # 翻转图像\n", + " if self.rand()<.5 and self.random: \n", + " image = image.transpose(Image.FLIP_LEFT_RIGHT)\n", + " image = resize_image(image, [self.input_shape[1], self.input_shape[0]], letterbox_image = True)\n", + " image = preprocess_input(np.array(image, dtype='float32'))\n", + " image = np.transpose(image, [2, 0, 1])\n", + " images[1, :, :, :] = image\n", + " labels[1] = c\n", + "\n", + " # 取出第三个人的人脸\n", + " different_c = list(range(self.num_classes))\n", + " different_c.pop(c)\n", + " different_c_index = np.random.choice(range(0, self.num_classes - 1), 1)\n", + " current_c = different_c[different_c_index[0]]\n", + " selected_path = self.paths[self.labels == current_c]\n", + " while len(selected_path)<1:\n", + " different_c_index = np.random.choice(range(0, self.num_classes - 1), 1)\n", + " current_c = different_c[different_c_index[0]]\n", + " selected_path = self.paths[self.labels == current_c]\n", + "\n", + " # 随机选择一张\n", + " image_indexes = np.random.choice(range(0, len(selected_path)), 1)\n", + " image = cvtColor(Image.open(selected_path[image_indexes[0]]))\n", + " # 翻转图像\n", + " if self.rand()<.5 and self.random: \n", + " image = image.transpose(Image.FLIP_LEFT_RIGHT)\n", + " image = resize_image(image, [self.input_shape[1], self.input_shape[0]], letterbox_image = True)\n", + " image = preprocess_input(np.array(image, dtype='float32'))\n", + " image = np.transpose(image, [2, 0, 1])\n", + " images[2, :, :, :] = image\n", + " labels[2] = current_c\n", + "\n", + " # images:三个人脸的数据，labels：三个人脸对应的类别\n", + " return images, labels\n", + "\n", + " def rand(self, a=0, b=1):\n", + " return np.random.rand()*(b-a) + a\n", + " \n", + " def load_dataset(self):\n", + " for path in self.lines:\n", + " path_split = path.split(\";\")\n", + " # paths\n", + " self.paths.append(path_split[1].split()[0])\n", + " # labels\n", + " self.labels.append(int(path_split[0]))\n", + " self.paths = np.array(self.paths,dtype=np.object)\n", + " self.labels = np.array(self.labels)\n", + "\n", + "\n", + " # DataLoader collate_fn\n", + " def dataset_collate(self, batch):\n", + " # batch:输入一个batch的数据\n", + " images = []\n", + " labels = []\n", + " for img, label in batch:\n", + " images.append(img)\n", + " labels.append(label)\n", + " # caocat image data\n", + " images1 = np.array(images)[:, 0, :, :, :]\n", + " images2 = np.array(images)[:, 1, :, :, :]\n", + " images3 = np.array(images)[:, 2, :, :, :]\n", + " images = np.concatenate([images1, images2, images3], 0)\n", + " # concat label data\n", + " labels1 = np.array(labels)[:, 0]\n", + " labels2 = np.array(labels)[:, 1]\n", + " labels3 = np.array(labels)[:, 2]\n", + " labels = np.concatenate([labels1, labels2, labels3], 0)\n", + " return images, labels\n", + "\n" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 构建数据集并创建dataloader" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:49:52.054103Z", + "iopub.status.busy": "2022-05-03T14:49:52.053786Z", + "iopub.status.idle": "2022-05-03T14:49:52.057762Z", + "shell.execute_reply": "2022-05-03T14:49:52.057313Z", + "shell.execute_reply.started": "2022-05-03T14:49:52.054077Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "import os\n", + "import paddle\n", + "import paddle.nn as nn\n", + "from paddle.nn import functional as F\n", + "from tqdm import tqdm\n", + "import argparse\n", + "import paddle.optimizer as optim\n", + "from paddle.io import DataLoader" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:49:54.630970Z", + "iopub.status.busy": "2022-05-03T14:49:54.630644Z", + "iopub.status.idle": "2022-05-03T14:49:54.636648Z", + "shell.execute_reply": "2022-05-03T14:49:54.636195Z", + "shell.execute_reply.started": "2022-05-03T14:49:54.630945Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "input_shape = [224, 224, 3]\n", + "num_classes = 6\n", + "annotation_path = \"./cls_train.txt\"\n", + "# batchsize需要为3的倍数\n", + "batch_size = 15\n", + "with open(annotation_path, \"r\") as f:\n", + " lines = f.readlines()\n", + "num_train = len(lines)\n", + "\n", + "# build dataset\n", + "train_dataset = SearchDataset(input_shape, lines, num_classes, random=True)\n", + "# build dataloader\n", + "train_loader = DataLoader(train_dataset,\n", + " shuffle=True,\n", + " batch_size=batch_size // 3,\n", + " num_workers=4,\n", + " drop_last=True,\n", + " collate_fn=train_dataset.dataset_collate)" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:49:58.054331Z", + "iopub.status.busy": "2022-05-03T14:49:58.054008Z", + "iopub.status.idle": "2022-05-03T14:49:58.057901Z", + "shell.execute_reply": "2022-05-03T14:49:58.057472Z", + "shell.execute_reply.started": "2022-05-03T14:49:58.054308Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "number of train: 708\n" + ] + } + ], + "source": [ + "print(\"number of train:\", len(train_dataset))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 构建网络结构，此处先采用动态图结构" + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:50:01.813195Z", + "iopub.status.busy": "2022-05-03T14:50:01.812895Z", + "iopub.status.idle": "2022-05-03T14:50:01.816358Z", + "shell.execute_reply": "2022-05-03T14:50:01.815920Z", + "shell.execute_reply.started": "2022-05-03T14:50:01.813172Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "import paddle.nn as nn\n", + "import paddle\n", + "from paddle.nn import functional as F\n", + "import paddle.vision.models as models" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:28:05.502777Z", + "iopub.status.busy": "2022-05-03T14:28:05.502520Z", + "iopub.status.idle": "2022-05-03T14:28:05.513422Z", + "shell.execute_reply": "2022-05-03T14:28:05.512980Z", + "shell.execute_reply.started": "2022-05-03T14:28:05.502755Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "class OwnResnet(nn.Layer):\n", + " def __init__(self, pretrained):\n", + " super(OwnResnet, self).__init__()\n", + " self.model = models.resnet50(pretrained=pretrained)\n", + " # 删除不需要的层结构\n", + " del self.model.fc\n", + " del self.model.avgpool\n", + "\n", + " def forward(self, x):\n", + " x = self.model.conv1(x)\n", + " x = self.model.bn1(x)\n", + " x = self.model.relu(x)\n", + " x = self.model.maxpool(x)\n", + "\n", + " x = self.model.layer1(x)\n", + " x = self.model.layer2(x)\n", + " x = self.model.layer3(x)\n", + " x = self.model.layer4(x)\n", + "\n", + " return x\n", + "\n", + "\n", + "class OwnFacenet(nn.Layer):\n", + " def __init__(self, dropout_keep_prob=0.5, embedding_size=128, num_classes=None, mode=\"train\", pretrained=False):\n", + " super(OwnFacenet, self).__init__()\n", + " self.backbone = OwnResnet(pretrained=pretrained)\n", + " flat_shape = 2048\n", + " self.set_parameter_requires_grad(self.backbone, True)\n", + " self.avg = nn.AdaptiveAvgPool2D((1, 1))\n", + " self.flatten = paddle.nn.Flatten()\n", + " self.Dropout = nn.Dropout(1 - dropout_keep_prob)\n", + " self.Bottleneck = nn.Linear(flat_shape, embedding_size)\n", + " self.last_bn = nn.BatchNorm1D(embedding_size)\n", + " if mode == \"train\":\n", + " self.classifier = nn.Linear(embedding_size, num_classes)\n", + "\n", + " def set_parameter_requires_grad(self, model, feature_extracting):\n", + " if feature_extracting:\n", + " for param in model.parameters():\n", + " param.requires_grad = True\n", + "\n", + " def forward(self, x):\n", + " if self.training:\n", + " x = self.backbone(x)\n", + " x = self.avg(x)\n", + " x = self.flatten(x)\n", + " x = self.Dropout(x)\n", + " x = self.Bottleneck(x)\n", + " x = self.last_bn(x)\n", + " # 分类的结果\n", + " x1 = self.classifier(x)\n", + " # L2标准化之后的结果\n", + " x2 = F.normalize(x, p=2, axis=1)\n", + " return x1, x2\n", + " else:\n", + " x = self.backbone(x)\n", + " x = self.avg(x)\n", + " x = self.flatten(x)\n", + " x = self.Dropout(x)\n", + " x = self.Bottleneck(x)\n", + " x = self.last_bn(x)\n", + " # l2标准化后的结果\n", + " x = F.normalize(x, p=2, axis=1)\n", + " return x" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:28:13.664858Z", + "iopub.status.busy": "2022-05-03T14:28:13.664484Z", + "iopub.status.idle": "2022-05-03T14:28:22.769860Z", + "shell.execute_reply": "2022-05-03T14:28:22.769105Z", + "shell.execute_reply.started": "2022-05-03T14:28:13.664831Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "W0503 22:28:13.667455 210 device_context.cc:447] Please NOTE: device: 0, GPU Compute Capability: 8.0, Driver API Version: 11.2, Runtime API Version: 11.2\n", + "W0503 22:28:13.670624 210 device_context.cc:465] device: 0, cuDNN Version: 8.2.\n", + "100%|██████████| 151272/151272 [00:04<00:00, 35334.16it/s]\n" + ] + } + ], + "source": [ + "# build model\n", + "pretrained = True\n", + "model = OwnFacenet(num_classes=num_classes, pretrained=pretrained)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 构建损失函数，采用三元损失函数" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:50:11.043164Z", + "iopub.status.busy": "2022-05-03T14:50:11.042767Z", + "iopub.status.idle": "2022-05-03T14:50:11.049079Z", + "shell.execute_reply": "2022-05-03T14:50:11.048601Z", + "shell.execute_reply.started": "2022-05-03T14:50:11.043133Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "def triplet_loss(alpha = 0.2):\n", + " def _triplet_loss(y_pred, Batch_size):\n", + " # 获取到对应的三元特征\n", + " anchor, positive, negative = y_pred[:int(Batch_size)], y_pred[int(Batch_size):int(2*Batch_size)], y_pred[int(2*Batch_size):]\n", + "\n", + " pos_dist = paddle.sqrt(paddle.sum(paddle.pow(anchor - positive,2), axis=-1))\n", + " neg_dist = paddle.sqrt(paddle.sum(paddle.pow(anchor - negative,2), axis=-1))\n", + "\n", + " basic_loss = paddle.where(pos_dist - neg_dist + alpha > 0, pos_dist - neg_dist + alpha, paddle.to_tensor(0.0))\n", + "\n", + " loss = paddle.mean(basic_loss)\n", + " # # 产生损失\n", + " # keep_all = (neg_dist - pos_dist < alpha).cpu().numpy().flatten()\n", + " # hard_triplets = np.where(keep_all == 1)\n", + "\n", + " # print(hard_triplets)\n", + "\n", + " # pos_dist = pos_dist[hard_triplets]\n", + " # neg_dist = neg_dist[hard_triplets]\n", + "\n", + " # basic_loss = pos_dist - neg_dist + alpha\n", + " # # 取均值损失\n", + " # loss = paddle.sum(basic_loss) / paddle.max(paddle.assign([1, len(hard_triplets[0])]))\n", + " return loss\n", + " return _triplet_loss" + ] + }, + { + "cell_type": "code", + "execution_count": 9, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:50:13.434788Z", + "iopub.status.busy": "2022-05-03T14:50:13.434489Z", + "iopub.status.idle": "2022-05-03T14:50:13.437530Z", + "shell.execute_reply": "2022-05-03T14:50:13.437089Z", + "shell.execute_reply.started": "2022-05-03T14:50:13.434758Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "# build loss\n", + "loss = triplet_loss()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 构建优化器" + ] + }, + { + "cell_type": "code", + "execution_count": 10, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:50:15.523036Z", + "iopub.status.busy": "2022-05-03T14:50:15.522782Z", + "iopub.status.idle": "2022-05-03T14:50:15.531954Z", + "shell.execute_reply": "2022-05-03T14:50:15.531425Z", + "shell.execute_reply.started": "2022-05-03T14:50:15.523014Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "ename": "NameError", + "evalue": "name 'model' is not defined", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)", + "\u001b[0;32m/tmp/ipykernel_170/739400542.py\u001b[0m in \u001b[0;36m\u001b[0;34m\u001b[0m\n\u001b[1;32m 2\u001b[0m \u001b[0moptimizer_type\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'adam'\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m optimizer = {\n\u001b[0;32m----> 4\u001b[0;31m \u001b[0;34m'adam'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0moptim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mAdam\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mparameters\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparameters\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlearning_rate\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0.001\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 5\u001b[0m \u001b[0;34m'sgd'\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0moptim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mSGD\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mparameters\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mparameters\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlearning_rate\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0.01\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 6\u001b[0m }[optimizer_type]\n", + "\u001b[0;31mNameError\u001b[0m: name 'model' is not defined" + ] + } + ], + "source": [ + "# build optim\n", + "optimizer_type = 'adam'\n", + "optimizer = {\n", + " 'adam': optim.Adam(parameters = model.parameters(), learning_rate = 0.001),\n", + " 'sgd': optim.SGD(parameters = model.parameters(), learning_rate = 0.01)\n", + "}[optimizer_type]" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 开始训练" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:51:10.663455Z", + "iopub.status.busy": "2022-05-03T14:51:10.662847Z", + "iopub.status.idle": "2022-05-03T14:51:10.672958Z", + "shell.execute_reply": "2022-05-03T14:51:10.672425Z", + "shell.execute_reply.started": "2022-05-03T14:51:10.663419Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "def fit_one_epoch(model, loss, optimizer, epoch, epoch_step, gen, Epoch, Batch_size, save_period, save_dir):\n", + " # triple loss\n", + " total_triple_loss = 0\n", + " # cross entroy\n", + " total_CE_loss = 0\n", + " # total acc\n", + " total_accuracy = 0\n", + "\n", + " model.train()\n", + " with tqdm(total=epoch_step, desc=f'Epoch {epoch + 1}/{Epoch}', postfix=dict, mininterval=0.3) as pbar:\n", + " for iteration, batch in enumerate(gen):\n", + " if iteration >= epoch_step:\n", + " break\n", + " images, labels = batch\n", + " with paddle.no_grad():\n", + " images = paddle.to_tensor(images.astype('float32'))\n", + " labels = paddle.to_tensor(labels.astype('int64'))\n", + "\n", + " optimizer.clear_grad()\n", + " # 得到分类的数据以及L2标准化之前的数据\n", + " outputs2, outputs1 = model(images)\n", + " # 三元损失函数\n", + " _triplet_loss = loss(outputs1, Batch_size)\n", + " # 交叉熵损失函数\n", + " _CE_loss = nn.NLLLoss()(F.log_softmax(outputs2), labels)\n", + " # 损失相加\n", + " _loss = _triplet_loss + _CE_loss\n", + "\n", + " _loss.backward()\n", + " optimizer.step()\n", + "\n", + " with paddle.no_grad():\n", + " accuracy = paddle.mean(\n", + " (paddle.argmax(F.softmax(outputs2), axis=-1) == labels))\n", + "\n", + " total_triple_loss += _triplet_loss.item()\n", + " total_CE_loss += _CE_loss.item()\n", + " total_accuracy += accuracy.item()\n", + "\n", + " pbar.set_postfix(**{'total_triple_loss': total_triple_loss / (iteration + 1),\n", + " 'total_CE_loss': total_CE_loss / (iteration + 1),\n", + " 'accuracy': total_accuracy / (iteration + 1)})\n", + " pbar.update(1)\n", + " print('Epoch:' + str(epoch + 1) + '/' + str(Epoch))\n", + " print('Total Loss: %.4f' % ((total_triple_loss + total_CE_loss) / epoch_step))\n", + " if (epoch + 1) % save_period == 0 or epoch + 1 == Epoch:\n", + " paddle.save(model.state_dict(), os.path.join(save_dir, 'ep%02d.pth' % (epoch + 1)))" + ] + }, + { + "cell_type": "code", + "execution_count": 14, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:28:38.306112Z", + "iopub.status.busy": "2022-05-03T14:28:38.305785Z", + "iopub.status.idle": "2022-05-03T14:29:28.261628Z", + "shell.execute_reply": "2022-05-03T14:29:28.260454Z", + "shell.execute_reply.started": "2022-05-03T14:28:38.306086Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 1/10: 0%| | 0/47 [00:00]/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/nn/layer/norm.py:653: UserWarning: When training, we now always track global mean and variance.\n", + " \"When training, we now always track global mean and variance.\")\n", + "Epoch 1/10: 100%|██████████| 47/47 [00:07<00:00, 6.16it/s, accuracy=1, total_CE_loss=1.03, total_triple_loss=0.0546]\n", + "Epoch 2/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:1/10\n", + "Total Loss: 1.0842\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 2/10: 100%|██████████| 47/47 [00:04<00:00, 9.57it/s, accuracy=1, total_CE_loss=0.652, total_triple_loss=0.0527]\n", + "Epoch 3/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:2/10\n", + "Total Loss: 0.7044\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 3/10: 100%|██████████| 47/47 [00:04<00:00, 9.74it/s, accuracy=1, total_CE_loss=0.506, total_triple_loss=0.0251]\n", + "Epoch 4/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:3/10\n", + "Total Loss: 0.5315\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 4/10: 100%|██████████| 47/47 [00:04<00:00, 9.75it/s, accuracy=1, total_CE_loss=0.284, total_triple_loss=0.016] \n", + "Epoch 5/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:4/10\n", + "Total Loss: 0.2998\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 5/10: 100%|██████████| 47/47 [00:04<00:00, 9.83it/s, accuracy=1, total_CE_loss=0.363, total_triple_loss=0.0143] \n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:5/10\n", + "Total Loss: 0.3772\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 6/10: 100%|██████████| 47/47 [00:04<00:00, 10.46it/s, accuracy=1, total_CE_loss=0.372, total_triple_loss=0.0155]\n", + "Epoch 7/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:6/10\n", + "Total Loss: 0.3878\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 7/10: 100%|██████████| 47/47 [00:04<00:00, 10.30it/s, accuracy=1, total_CE_loss=0.299, total_triple_loss=0.0083] \n", + "Epoch 8/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:7/10\n", + "Total Loss: 0.3072\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 8/10: 100%|██████████| 47/47 [00:04<00:00, 10.52it/s, accuracy=1, total_CE_loss=0.261, total_triple_loss=0.0169] \n", + "Epoch 9/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:8/10\n", + "Total Loss: 0.2779\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 9/10: 100%|██████████| 47/47 [00:04<00:00, 10.49it/s, accuracy=1, total_CE_loss=0.2, total_triple_loss=0.00426] \n", + "Epoch 10/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:9/10\n", + "Total Loss: 0.2044\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 10/10: 100%|██████████| 47/47 [00:04<00:00, 10.58it/s, accuracy=1, total_CE_loss=0.243, total_triple_loss=0.0139]\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:10/10\n", + "Total Loss: 0.2573\n" + ] + } + ], + "source": [ + "epoch_step = num_train // batch_size\n", + "epochs = 10\n", + "save_period = 5\n", + "save_dir = \"./output/\"\n", + "# train\n", + "for epoch in range(epochs):\n", + " fit_one_epoch(model, loss, optimizer,\n", + " epoch, epoch_step, train_loader,\n", + " epochs, batch_size // 3, save_period, save_dir)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 预测过程，将图像库中的人脸转换为高维度特征中" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T13:49:01.828720Z", + "iopub.status.busy": "2022-05-03T13:49:01.828369Z", + "iopub.status.idle": "2022-05-03T13:49:24.397577Z", + "shell.execute_reply": "2022-05-03T13:49:24.397038Z", + "shell.execute_reply.started": "2022-05-03T13:49:01.828693Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "100%|██████████| 708/708 [00:22<00:00, 31.47it/s]\n" + ] + } + ], + "source": [ + "import json\n", + "model.eval()\n", + "total_embedding = {}\n", + "with open(\"./cls_train.txt\", \"r\") as f:\n", + " lines = f.readlines()\n", + "\n", + "for line in tqdm(lines):\n", + " path = line.split(\";\")[1].split()[0]\n", + " image_data = Image.open(path)\n", + " with paddle.no_grad():\n", + " image_data = resize_image(image_data, [input_shape[1], input_shape[0]], letterbox_image=True)\n", + "\n", + " photo_data = paddle.to_tensor(\n", + " np.expand_dims(np.transpose(preprocess_input(np.array(image_data, np.float32)), (2, 0, 1)), 0))\n", + "\n", + " output = model(photo_data).cpu().numpy().tolist()\n", + "\n", + " total_embedding[path] = output\n", + "\n", + "# 转化为json格式文件\n", + "dict_json = json.dumps(total_embedding)\n", + "# save\n", + "with open('image_feature.json', 'w+') as file:\n", + " file.write(dict_json)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 输入一张测试图片进行测试" + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T13:54:02.233029Z", + "iopub.status.busy": "2022-05-03T13:54:02.232675Z", + "iopub.status.idle": "2022-05-03T13:54:06.996105Z", + "shell.execute_reply": "2022-05-03T13:54:06.995530Z", + "shell.execute_reply.started": "2022-05-03T13:54:02.232999Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdin", + "output_type": "stream", + "text": [ + "Input image filename: ./test.png\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "{'./images/face/dilireba/1591160437114.jpg': array([1.52943838]), './images/face/dilireba/159116043766.jpg': array([1.81909941]), './images/face/dilireba/159116043761.jpg': array([1.72572565]), './images/face/dilireba/159116043652.jpg': array([1.62632892]), './images/face/dilireba/159116043631.jpg': array([1.49347721]), './images/face/dilireba/159116043787.jpg': array([1.54048517]), './images/face/dilireba/159116043638.jpg': array([1.72115597]), './images/face/dilireba/159116043618.jpg': array([1.67191251]), './images/face/dilireba/159116043791.jpg': array([1.57484631]), './images/face/dilireba/159116043629.jpg': array([1.56656123]), './images/face/dilireba/15911604352.jpg': array([1.59859238]), './images/face/dilireba/1591160438117.jpg': array([1.67207121]), './images/face/dilireba/159116043783.jpg': array([1.68600151]), './images/face/dilireba/159116043765.jpg': array([1.67974763]), './images/face/dilireba/1591160437103.jpg': array([1.6821898]), './images/face/dilireba/159116043634.jpg': array([1.70097263]), './images/face/dilireba/1591160437108.jpg': array([1.70638834]), './images/face/dilireba/1591160437109.jpg': array([1.78146652]), './images/face/dilireba/1591160438123.jpg': array([1.66361889]), './images/face/dilireba/159116043644.jpg': array([1.68887578]), './images/face/dilireba/159116043763.jpg': array([1.74190645]), './images/face/dilireba/1591160438121.jpg': array([1.70675565]), './images/face/dilireba/159116043627.jpg': array([1.56431934]), './images/face/dilireba/159116043788.jpg': array([1.44203147]), './images/face/dilireba/159116043779.jpg': array([1.69098115]), './images/face/dilireba/159116043796.jpg': array([1.73975495]), './images/face/dilireba/159116043786.jpg': array([1.66647533]), './images/face/dilireba/1591160438119.jpg': array([1.71579342]), './images/face/dilireba/159116043651.jpg': array([1.59390031]), './images/face/dilireba/159116043620.jpg': array([1.73887183]), './images/face/dilireba/159116043610.jpg': array([1.62272366]), './images/face/dilireba/1591160437111.jpg': array([1.56988452]), './images/face/dilireba/1591160438124.jpg': array([1.71312583]), './images/face/dilireba/1591160438122.jpg': array([1.60557072]), './images/face/dilireba/159116043624.jpg': array([1.48538616]), './images/face/dilireba/159116043640.jpg': array([1.59245396]), './images/face/dilireba/159116043641.jpg': array([1.59179558]), './images/face/dilireba/159116043637.jpg': array([1.67911132]), './images/face/dilireba/159116043656.jpg': array([1.59573242]), './images/face/dilireba/159116043642.jpg': array([1.80563757]), './images/face/dilireba/159116043794.jpg': array([1.60968038]), './images/face/dilireba/159116043649.jpg': array([1.71867427]), './images/face/dilireba/159116043776.jpg': array([1.7447762]), './images/face/dilireba/159116043767.jpg': array([1.63657115]), './images/face/dilireba/159116043643.jpg': array([1.59670168]), './images/face/dilireba/1591160437105.jpg': array([1.7392782]), './images/face/dilireba/159116043613.jpg': array([1.50833338]), './images/face/dilireba/1591160437104.jpg': array([1.61545643]), './images/face/dilireba/159116043657.jpg': array([1.55892586]), './images/face/dilireba/159116043770.jpg': array([1.71549738]), './images/face/dilireba/159116043764.jpg': array([1.60978589]), './images/face/dilireba/1591160438116.jpg': array([1.62507194]), './images/face/dilireba/159116043658.jpg': array([1.72788412]), './images/face/dilireba/159116043653.jpg': array([1.78939847]), './images/face/dilireba/159116043799.jpg': array([1.62991133]), './images/face/dilireba/159116043784.jpg': array([1.71323944]), './images/face/dilireba/159116043659.jpg': array([1.8155689]), './images/face/dilireba/1591160437110.jpg': array([1.71703573]), './images/face/dilireba/159116043616.jpg': array([1.6314609]), './images/face/dilireba/159116043628.jpg': array([1.84156401]), './images/face/dilireba/15911604354.jpg': array([1.75486224]), './images/face/dilireba/159116043655.jpg': array([1.64817878]), './images/face/dilireba/159116043793.jpg': array([1.67631544]), './images/face/dilireba/159116043797.jpg': array([1.80858585]), './images/face/dilireba/159116043781.jpg': array([1.68854601]), './images/face/dilireba/1591160437102.jpg': array([1.6969666]), './images/face/dilireba/159116043632.jpg': array([1.74814438]), './images/face/dilireba/159116043646.jpg': array([1.73473525]), './images/face/dilireba/159116043625.jpg': array([1.62579826]), './images/face/dilireba/15911604367.jpg': array([1.537668]), './images/face/dilireba/159116043771.jpg': array([1.54202493]), './images/face/dilireba/159116043772.jpg': array([1.61429913]), './images/face/dilireba/1591160438118.jpg': array([1.57284546]), './images/face/dilireba/15911604353.jpg': array([1.66065944]), './images/face/dilireba/15911604368.jpg': array([1.70666132]), './images/face/dilireba/159116043777.jpg': array([1.646693]), './images/face/dilireba/159116043612.jpg': array([1.66327532]), './images/face/dilireba/15911604366.jpg': array([1.4983944]), './images/face/dilireba/159116043611.jpg': array([1.65398957]), './images/face/dilireba/159116043639.jpg': array([1.75573135]), './images/face/dilireba/159116043633.jpg': array([1.68875467]), './images/face/dilireba/1591160437101.jpg': array([1.74704563]), './images/face/dilireba/159116043650.jpg': array([1.66247571]), './images/face/dilireba/159116043768.jpg': array([1.67769875]), './images/face/dilireba/159116043660.jpg': array([1.70547071]), './images/face/dilireba/1591160437100.jpg': array([1.75637874]), './images/face/dilireba/159116043623.jpg': array([1.74993659]), './images/face/dilireba/159116043773.jpg': array([1.67728618]), './images/face/dilireba/15911604355.jpg': array([1.75746382]), './images/face/dilireba/159116043626.jpg': array([1.7759604]), './images/face/dilireba/159116043775.jpg': array([1.48714481]), './images/face/dilireba/159116043785.jpg': array([1.7139492]), './images/face/dilireba/1591160438125.jpg': array([1.61817334]), './images/face/dilireba/159116043798.jpg': array([1.60751246]), './images/face/dilireba/159116043622.jpg': array([1.7662543]), './images/face/dilireba/159116043769.jpg': array([1.83472159]), './images/face/dilireba/1591160438115.jpg': array([1.71429537]), './images/face/dilireba/159116043654.jpg': array([1.79191572]), './images/face/dilireba/159116043635.jpg': array([1.74386829]), './images/face/dilireba/1591160437112.jpg': array([1.58541011]), './images/face/dilireba/1591160437107.jpg': array([1.56765911]), './images/face/dilireba/159116043789.jpg': array([1.65139744]), './images/face/dilireba/159116043648.jpg': array([1.53366293]), './images/face/dilireba/159116043614.jpg': array([1.790321]), './images/face/dilireba/159116043647.jpg': array([1.7981514]), './images/face/dilireba/159116043615.jpg': array([1.57558158]), './images/face/dilireba/159116043792.jpg': array([1.77592819]), './images/face/dilireba/15911604369.jpg': array([1.65681215]), './images/face/dilireba/159116043645.jpg': array([1.67782463]), './images/face/dilireba/159116043762.jpg': array([1.6038118]), './images/face/dilireba/1591160438120.jpg': array([1.68891728]), './images/face/dilireba/159116043636.jpg': array([1.71430617]), './images/face/jiangwen/13d309cf0-f929-11e8-855f-005056c00008.jpg': array([1.53408933]), './images/face/jiangwen/1e2c7540-f929-11e8-b467-005056c00008.jpg': array([1.46895388]), './images/face/jiangwen/8900b980-f929-11e8-b547-005056c00008.jpg': array([1.65570251]), './images/face/jiangwen/4f2308d2-f929-11e8-b45b-005056c00008.jpg': array([1.53157442]), './images/face/jiangwen/13bc1a3a2-f929-11e8-bdf1-005056c00008.jpeg': array([1.52548363]), './images/face/jiangwen/1c2128eb0-f929-11e8-8a05-005056c00008.jpg': array([1.5787679]), './images/face/jiangwen/18207c00-f929-11e8-bee9-005056c00008.jpg': array([1.54180684]), './images/face/jiangwen/b91699a2-f929-11e8-a82b-005056c00008.jpg': array([1.54215832]), './images/face/jiangwen/40040840-f929-11e8-b9ec-005056c00008.jpg': array([1.51078592]), './images/face/jiangwen/11fe84670-f929-11e8-8e1b-005056c00008.jpg': array([1.40068775]), './images/face/jiangwen/ea6c3292-f928-11e8-b3ee-005056c00008.jpg': array([1.58965581]), './images/face/jiangwen/1ba5952d2-f929-11e8-bcf4-005056c00008.jpg': array([1.55967683]), './images/face/jiangwen/6c874bc0-f929-11e8-bad9-005056c00008.jpg': array([1.55609984]), './images/face/jiangwen/b18ece00-f929-11e8-8e00-005056c00008.jpg': array([1.57136487]), './images/face/jiangwen/191270f12-f929-11e8-8bc6-005056c00008.jpeg': array([1.57976453]), './images/face/jiangwen/1413cfd70-f929-11e8-a030-005056c00008.jpg': array([1.51078592]), './images/face/jiangwen/aff912d2-f929-11e8-8807-005056c00008.jpg': array([1.43728046]), './images/face/jiangwen/741224a2-f929-11e8-9e25-005056c00008.jpg': array([1.67166416]), './images/face/jiangwen/118207c00-f929-11e8-bee9-005056c00008.jpg': array([1.54180684]), './images/face/jiangwen/11b9203e2-f929-11e8-970c-005056c00008.jpg': array([1.65648994]), './images/face/jiangwen/b9969f12-f929-11e8-8e46-005056c00008.jpg': array([1.58404831]), './images/face/jiangwen/f97205d2-f928-11e8-9d91-005056c00008.jpg': array([1.56811734]), './images/face/jiangwen/1eade4870-f929-11e8-befb-005056c00008.jpg': array([1.56605017]), './images/face/jiangwen/18e42b3d2-f929-11e8-ad0c-005056c00008.jpg': array([1.59749534]), './images/face/jiangwen/1fe84670-f929-11e8-8e1b-005056c00008.jpg': array([1.40068775]), './images/face/jiangwen/170bedaf0-f929-11e8-b1ba-005056c00008.jpeg': array([1.56263477]), './images/face/jiangwen/18f6fc222-f929-11e8-86ae-005056c00008.jpg': array([1.51786015]), './images/face/jiangwen/1fc638070-f928-11e8-9d12-005056c00008.jpg': array([1.45677522]), './images/face/jiangwen/13ac6eff0-f929-11e8-bd75-005056c00008.jpg': array([1.6416277]), './images/face/jiangwen/1cc97270-f929-11e8-8f86-005056c00008.jpeg': array([1.50425536]), './images/face/jiangwen/1bebb2652-f929-11e8-a1e8-005056c00008.jpg': array([1.61556215]), './images/face/jiangwen/70bedaf0-f929-11e8-b1ba-005056c00008.jpeg': array([1.56263477]), './images/face/jiangwen/14f2308d2-f929-11e8-b45b-005056c00008.jpeg': array([1.53157442]), './images/face/jiangwen/f9c8b152-f928-11e8-b6ca-005056c00008.jpg': array([1.61597777]), './images/face/jiangwen/192c57cd2-f929-11e8-855d-005056c00008.jpg': array([1.5996909]), './images/face/jiangwen/1ad8ed12-f929-11e8-ab1f-005056c00008.jpg': array([1.53823607]), './images/face/jiangwen/91270f12-f929-11e8-8bc6-005056c00008.jpeg': array([1.57976453]), './images/face/jiangwen/faba8b62-f928-11e8-a0bc-005056c00008.jpg': array([1.63895669]), './images/face/jiangwen/0acb8d12-f929-11e8-ac67-005056c00008.jpg': array([1.56120733]), './images/face/jiangwen/13b938ec0-f929-11e8-85af-005056c00008.jpg': array([1.62119738]), './images/face/jiangwen/11bc5be12-f929-11e8-860a-005056c00008.jpg': array([1.57334068]), './images/face/jiangwen/0b1937e2-f929-11e8-8a8a-005056c00008.jpg': array([1.60418853]), './images/face/jiangwen/185e2140-f929-11e8-9b08-005056c00008.jpeg': array([1.60071491]), './images/face/jiangwen/fb42ce30-f928-11e8-b69d-005056c00008.jpg': array([1.63584761]), './images/face/jiangwen/1b91699a2-f929-11e8-a82b-005056c00008.jpg': array([1.54215832]), './images/face/jiangwen/14047c800-f929-11e8-9bbd-005056c00008.jpg': array([1.59915011]), './images/face/jiangwen/94b19562-f929-11e8-88e0-005056c00008.jpg': array([1.61001398]), './images/face/jiangwen/11cc97270-f929-11e8-8f86-005056c00008.jpeg': array([1.50425536]), './images/face/jiangwen/1b9203e2-f929-11e8-970c-005056c00008.jpg': array([1.65648994]), './images/face/jiangwen/18900b980-f929-11e8-b547-005056c00008.jpg': array([1.65570251]), './images/face/jiangwen/1a2935b00-f929-11e8-8ef8-005056c00008.jpg': array([1.61162726]), './images/face/jiangwen/1bd622600-f929-11e8-8623-005056c00008.jpg': array([1.41541236]), './images/face/jiangwen/1d3134792-f929-11e8-8194-005056c00008.jpg': array([1.64392526]), './images/face/jiangwen/b055d8d2-f929-11e8-a6e5-005056c00008.jpg': array([1.53620118]), './images/face/jiangwen/4047c800-f929-11e8-9bbd-005056c00008.jpg': array([1.59915011]), './images/face/jiangwen/eade4870-f929-11e8-befb-005056c00008.jpg': array([1.56605017]), './images/face/jiangwen/3a0d63f0-f929-11e8-9dce-005056c00008.png': array([1.4150194]), './images/face/jiangwen/11e78d7f0-f929-11e8-b323-005056c00008.jpg': array([1.52242032]), './images/face/jiangwen/e9fde062-f928-11e8-acbd-005056c00008.jpg': array([1.59704973]), './images/face/jiangwen/3d309cf0-f929-11e8-855f-005056c00008.jpg': array([1.53408933]), './images/face/jiangwen/16c874bc0-f929-11e8-bad9-005056c00008.jpeg': array([1.55609984]), './images/face/jiangwen/1741224a2-f929-11e8-9e25-005056c00008.jpg': array([1.67166416]), './images/face/jiangwen/d3134792-f929-11e8-8194-005056c00008.jpg': array([1.64392526]), './images/face/jiangwen/3f521d62-f929-11e8-a72b-005056c00008.jpeg': array([1.52370707]), './images/face/jiangwen/1fb8be522-f928-11e8-8016-005056c00008.jpg': array([1.6073134]), './images/face/jiangwen/fc638070-f928-11e8-9d12-005056c00008.jpg': array([1.45677522]), './images/face/jiangwen/930285d2-f929-11e8-a039-005056c00008.jpg': array([1.68123373]), './images/face/jiangwen/1eb9ffa2-f929-11e8-a78c-005056c00008.jpg': array([1.52311424]), './images/face/jiangwen/1b20584a2-f929-11e8-8681-005056c00008.png': array([1.57751174]), './images/face/jiangwen/1f896b422-f929-11e8-b4fb-005056c00008.jpg': array([1.53582145]), './images/face/jiangwen/14eb97192-f929-11e8-a1ff-005056c00008.jpg': array([1.5941048]), './images/face/jiangwen/1f9c8b152-f928-11e8-b6ca-005056c00008.jpg': array([1.61597777]), './images/face/jiangwen/b20584a2-f929-11e8-8681-005056c00008.png': array([1.57751174]), './images/face/jiangwen/1185e2140-f929-11e8-9b08-005056c00008.jpeg': array([1.60071491]), './images/face/jiangwen/ea95b392-f928-11e8-a520-005056c00008.jpg': array([1.58534174]), './images/face/jiangwen/1905985e2-f929-11e8-a7e1-005056c00008.png': array([1.57676868]), './images/face/jiangwen/11ad8ed12-f929-11e8-ab1f-005056c00008.jpg': array([1.53823607]), './images/face/jiangwen/1eb0ae392-f928-11e8-81ce-005056c00008.jpg': array([1.49289666]), './images/face/jiangwen/1b055d8d2-f929-11e8-a6e5-005056c00008.jpg': array([1.53620118]), './images/face/jiangwen/11c749bb0-f929-11e8-bfb9-005056c00008.jpg': array([1.60151004]), './images/face/jiangwen/1fa115312-f928-11e8-91fb-005056c00008.jpg': array([1.5715635]), './images/face/jiangwen/1b9969f12-f929-11e8-8e46-005056c00008.jpg': array([1.58404831]), './images/face/jiangwen/3b4791a2-f929-11e8-8555-005056c00008.jpg': array([1.51122026]), './images/face/jiangwen/3bc1a3a2-f929-11e8-bdf1-005056c00008.jpeg': array([1.52548363]), './images/face/jiangwen/1e9fde062-f928-11e8-acbd-005056c00008.jpg': array([1.59704973]), './images/face/jiangwen/c2128eb0-f929-11e8-8a05-005056c00008.jpg': array([1.5787679]), './images/face/jiangwen/905985e2-f929-11e8-a7e1-005056c00008.png': array([1.57676868]), './images/face/jiangwen/fb8be522-f928-11e8-8016-005056c00008.jpg': array([1.6073134]), './images/face/jiangwen/bebb2652-f929-11e8-a1e8-005056c00008.jpg': array([1.61556215]), './images/face/jiangwen/1930285d2-f929-11e8-a039-005056c00008.jpg': array([1.68123373]), './images/face/jiangwen/1b56b44e2-f929-11e8-995c-005056c00008.jpg': array([1.53978328]), './images/face/jiangwen/3b938ec0-f929-11e8-85af-005056c00008.jpg': array([1.62119738]), './images/face/jiangwen/bd622600-f929-11e8-8623-005056c00008.jpg': array([1.41541236]), './images/face/jiangwen/b56b44e2-f929-11e8-995c-005056c00008.jpg': array([1.53978328]), './images/face/jiangwen/18b60540-f929-11e8-aefd-005056c00008.jpg': array([1.57537575]), './images/face/jiangwen/d3725780-f929-11e8-b066-005056c00008.jpg': array([1.62143664]), './images/face/jiangwen/18358d170-f929-11e8-b2a3-005056c00008.jpg': array([1.65570251]), './images/face/jiangwen/8e42b3d2-f929-11e8-ad0c-005056c00008.jpg': array([1.59749534]), './images/face/jiangwen/194b19562-f929-11e8-88e0-005056c00008.jpg': array([1.61001398]), './images/face/jiangwen/4eb97192-f929-11e8-a1ff-005056c00008.jpg': array([1.5941048]), './images/face/jiangwen/1d3725780-f929-11e8-b066-005056c00008.jpg': array([1.62143664]), './images/face/jiangwen/413cfd70-f929-11e8-a030-005056c00008.jpg': array([1.51078592]), './images/face/jiangwen/1c749bb0-f929-11e8-bfb9-005056c00008.jpg': array([1.60151004]), './images/face/jiangwen/1ea6c3292-f928-11e8-b3ee-005056c00008.jpg': array([1.58965581]), './images/face/jiangwen/11e2c7540-f929-11e8-b467-005056c00008.jpg': array([1.46895388]), './images/face/jiangwen/1b74c60f0-f929-11e8-8eea-005056c00008.jpg': array([1.5572305]), './images/face/jiangwen/1ea95b392-f928-11e8-a520-005056c00008.jpg': array([1.58534174]), './images/face/jiangwen/10b1937e2-f929-11e8-8a8a-005056c00008.jpg': array([1.60418853]), './images/face/jiangwen/118b60540-f929-11e8-aefd-005056c00008.jpg': array([1.57537575]), './images/face/jiangwen/1e78d7f0-f929-11e8-b323-005056c00008.jpg': array([1.52242032]), './images/face/jiangwen/13f521d62-f929-11e8-a72b-005056c00008.jpeg': array([1.52370707]), './images/face/jiangwen/41625fc0-f929-11e8-bcf4-005056c00008.jpg': array([1.59915011]), './images/face/jiangwen/140040840-f929-11e8-b9ec-005056c00008.jpg': array([1.51078592]), './images/face/jiangwen/1faba8b62-f928-11e8-a0bc-005056c00008.jpeg': array([1.63895669]), './images/face/jiangwen/b74c60f0-f929-11e8-8eea-005056c00008.jpg': array([1.5572305]), './images/face/jiangwen/1de1c040-f929-11e8-b9f1-005056c00008.jpeg': array([1.5882164]), './images/face/jiangwen/3ac6eff0-f929-11e8-bd75-005056c00008.jpg': array([1.6416277]), './images/face/jiangwen/11eb9ffa2-f929-11e8-a78c-005056c00008.jpg': array([1.52311424]), './images/face/jiangwen/f896b422-f929-11e8-b4fb-005056c00008.jpg': array([1.53582145]), './images/face/jiangwen/141625fc0-f929-11e8-bcf4-005056c00008.jpg': array([1.59915011]), './images/face/jiangwen/1b18ece00-f929-11e8-8e00-005056c00008.jpg': array([1.57136487]), './images/face/jiangwen/1bc5be12-f929-11e8-860a-005056c00008.jpg': array([1.57334068]), './images/face/jiangwen/11de1c040-f929-11e8-b9f1-005056c00008.jpeg': array([1.5882164]), './images/face/jiangwen/8358d170-f929-11e8-b2a3-005056c00008.jpg': array([1.65570251]), './images/face/jiangwen/13b4791a2-f929-11e8-8555-005056c00008.jpg': array([1.51122026]), './images/face/jiangwen/13a0d63f0-f929-11e8-9dce-005056c00008.png': array([1.4150194]), './images/face/jiangwen/fa115312-f928-11e8-91fb-005056c00008.jpg': array([1.5715635]), './images/face/jiangwen/10acb8d12-f929-11e8-ac67-005056c00008.jpg': array([1.56120733]), './images/face/jiangwen/8f6fc222-f929-11e8-86ae-005056c00008.jpg': array([1.51786015]), './images/face/jiangwen/a2935b00-f929-11e8-8ef8-005056c00008.jpg': array([1.61162726]), './images/face/jiangwen/eb0ae392-f928-11e8-81ce-005056c00008.jpg': array([1.49289666]), './images/face/jiangwen/1fb42ce30-f928-11e8-b69d-005056c00008.jpg': array([1.63584761]), './images/face/jiangwen/92c57cd2-f929-11e8-855d-005056c00008.jpg': array([1.5996909]), './images/face/pengyuyan/20181206162816.png': array([1.26991212]), './images/face/pengyuyan/20181206162749.png': array([1.22698742]), './images/face/pengyuyan/20181206161903.png': array([1.15296163]), './images/face/pengyuyan/20181206162836.png': array([1.32709804]), './images/face/pengyuyan/20181206161127.png': array([1.2018372]), './images/face/pengyuyan/20181206161502.png': array([1.25224711]), './images/face/pengyuyan/20181206164423.png': array([1.39950838]), './images/face/pengyuyan/20181206164450.png': array([1.26418181]), './images/face/pengyuyan/20181206165131.png': array([1.3496142]), './images/face/pengyuyan/20181206162850.png': array([1.32544673]), './images/face/pengyuyan/20181206162625.png': array([1.26772531]), './images/face/pengyuyan/20181206163439.png': array([1.24822719]), './images/face/pengyuyan/20181206162423.png': array([1.38640564]), './images/face/pengyuyan/20181206165108.png': array([1.26731603]), './images/face/pengyuyan/20181206165249.png': array([1.306695]), './images/face/pengyuyan/20181206161848.png': array([1.15010582]), './images/face/pengyuyan/20181206161312.png': array([1.4259354]), './images/face/pengyuyan/20181206161634.png': array([1.24339201]), './images/face/pengyuyan/20181206162654.png': array([1.36345067]), './images/face/pengyuyan/20181206162454.png': array([1.19541545]), './images/face/pengyuyan/20181206162157.png': array([1.26477038]), './images/face/pengyuyan/20181206164707.png': array([1.36513975]), './images/face/pengyuyan/20181206161828.png': array([1.29521617]), './images/face/pengyuyan/20181206162254.png': array([1.2783777]), './images/face/pengyuyan/20181206162641.png': array([1.29685858]), './images/face/pengyuyan/20181206163250.png': array([1.41669799]), './images/face/pengyuyan/20181206163359.png': array([1.28546783]), './images/face/pengyuyan/20181206162507.png': array([1.35187183]), './images/face/pengyuyan/20181206164840.png': array([1.34755347]), './images/face/pengyuyan/20181206161233.png': array([1.21267176]), './images/face/pengyuyan/20181206164513.png': array([1.19208464]), './images/face/pengyuyan/20181206161702.png': array([1.38108185]), './images/face/pengyuyan/20181206162932.png': array([1.35069361]), './images/face/pengyuyan/20181206162904.png': array([1.23554556]), './images/face/pengyuyan/20181206162112.png': array([1.34479659]), './images/face/pengyuyan/20181206162540.png': array([1.30361458]), './images/face/pengyuyan/20181206165238.png': array([1.27102907]), './images/face/pengyuyan/20181206161053.png': array([1.11449058]), './images/face/pengyuyan/20181206164718.png': array([1.37601363]), './images/face/pengyuyan/20181206161530.png': array([1.36035882]), './images/face/pengyuyan/20181206164804.png': array([1.34744799]), './images/face/pengyuyan/20181206162442.png': array([1.368956]), './images/face/pengyuyan/20181206161917.png': array([1.22383694]), './images/face/pengyuyan/20181206162941.png': array([1.23955441]), './images/face/pengyuyan/20181206161115.png': array([1.15380278]), './images/face/pengyuyan/20181206161954.png': array([1.47989941]), './images/face/pengyuyan/20181206162613.png': array([1.30917719]), './images/face/pengyuyan/20181206161041.png': array([1.2648942]), './images/face/pengyuyan/20181206162956.png': array([1.44249117]), './images/face/pengyuyan/20181206163344.png': array([1.48437485]), './images/face/pengyuyan/20181206164640.png': array([1.37174855]), './images/face/pengyuyan/20181206165057.png': array([1.26825132]), './images/face/pengyuyan/20181206162915.png': array([1.27567884]), './images/face/pengyuyan/20181206164539.png': array([1.27363206]), './images/face/pengyuyan/20181206161806.png': array([1.37481485]), './images/face/pengyuyan/20181206162351.png': array([1.24989007]), './images/face/pengyuyan/20181206162411.png': array([1.17732127]), './images/face/pengyuyan/20181206165143.png': array([1.19142067]), './images/face/pengyuyan/20181206164436.png': array([1.28505961]), './images/face/pengyuyan/20181206163331.png': array([1.31533359]), './images/face/pengyuyan/20181206162707.png': array([1.4083768]), './images/face/pengyuyan/20181206161734.png': array([1.31083678]), './images/face/pengyuyan/20181206162027.png': array([1.20208685]), './images/face/pengyuyan/20181206165014.png': array([1.60046897]), './images/face/pengyuyan/20181206164742.png': array([1.29124669]), './images/face/pengyuyan/20181206161943.png': array([1.45211956]), './images/face/pengyuyan/20181206164912.png': array([1.41580418]), './images/face/pengyuyan/20181206165157.png': array([1.291113]), './images/face/pengyuyan/20181206162732.png': array([1.35806649]), './images/face/pengyuyan/20181206165211.png': array([1.33622974]), './images/face/pengyuyan/20181206162340.png': array([1.22575567]), './images/face/pengyuyan/20181206164727.png': array([1.04373706]), './images/face/pengyuyan/20181206163011.png': array([1.26341517]), './images/face/pengyuyan/20181206164827.png': array([1.22541939]), './images/face/pengyuyan/20181206161153.png': array([1.39494155]), './images/face/pengyuyan/20181206162804.png': array([1.25037493]), './images/face/pengyuyan/20181206164655.png': array([1.34506992]), './images/face/pengyuyan/20181206162042.png': array([1.32444487]), './images/face/pengyuyan/20181206162057.png': array([1.16532863]), './images/face/pengyuyan/20181206162552.png': array([1.31126654]), './images/face/pengyuyan/20181206162717.png': array([1.1722449]), './images/face/pengyuyan/20181206163224.png': array([1.1185643]), './images/face/pengyuyan/20181206161331.png': array([1.25963304]), './images/face/pengyuyan/20181206162211.png': array([1.27617245]), './images/face/pengyuyan/20181206164550.png': array([1.29826014]), './images/face/pengyuyan/20181206160937.png': array([1.40927825]), './images/face/pengyuyan/20181206165118.png': array([1.3481805]), './images/face/pengyuyan/20181206161554.png': array([1.33268144]), './images/face/pengyuyan/20181206162010.png': array([1.37818208]), './images/face/pengyuyan/20181206161255.png': array([0.91146088]), './images/face/pengyuyan/20181206161755.png': array([1.2777044]), './images/face/pengyuyan/20181206164526.png': array([1.28971848]), './images/face/pengyuyan/20181206161541.png': array([1.38400307]), './images/face/pengyuyan/20181206161448.png': array([1.31689675]), './images/face/pengyuyan/20181206164853.png': array([1.29253535]), './images/face/pengyuyan/20181206164628.png': array([1.40452742]), './images/face/pengyuyan/20181206163412.png': array([1.38392538]), './images/face/pengyuyan/20181206163306.png': array([1.47526546]), './images/face/pengyuyan/20181206163452.png': array([1.28898951]), './images/face/pengyuyan/20181206162311.png': array([1.32877471]), './images/face/pengyuyan/20181206164816.png': array([1.27161833]), './images/face/pengyuyan/20181206161648.png': array([1.44395119]), './images/face/pengyuyan/20181206161516.png': array([1.24544545]), './images/face/pengyuyan/20181206162602.png': array([1.24971361]), './images/face/pengyuyan/20181206163238.png': array([1.38014875]), './images/face/pengyuyan/20181206163318.png': array([1.32134958]), './images/face/pengyuyan/20181206162234.png': array([1.2738346]), './images/face/pengyuyan/20181206165044.png': array([1.30355176]), './images/face/pengyuyan/20181206162223.png': array([1.39589218]), './images/face/pengyuyan/20181206162522.png': array([1.15465787]), './images/face/pengyuyan/20181206164753.png': array([1.21832952]), './images/face/pengyuyan/20181206161607.png': array([1.26553888]), './images/face/pengyuyan/20181206163213.png': array([1.34851548]), './images/face/pengyuyan/20181206164502.png': array([1.18552278]), './images/face/zhangyan/微信图片_20200610094752.jpg': array([1.40027353]), './images/face/zhangyan/微信图片_20200610094810.jpg': array([1.58403678]), './images/face/zhangyan/微信图片_20200610094832.jpg': array([1.65434208]), './images/face/zhangyan/微信图片_20200610094902.jpg': array([1.31581952]), './images/face/zhangyan/微信图片_20200610094925.jpg': array([1.30575897]), './images/face/zhangyan/微信图片_20200610095028.jpg': array([1.58880399]), './images/face/zhangyan/微信图片_20200610094952.jpg': array([1.49125238]), './images/face/zhangyan/微信图片_20200610094742.jpg': array([1.272124]), './images/face/zhangyan/微信图片_20200610095046.jpg': array([1.35806002]), './images/face/zhangyan/微信图片_20200610094954.jpg': array([1.52103177]), './images/face/zhangyan/微信图片_20200610095018.jpg': array([1.26943521]), './images/face/zhangyan/微信图片_20200610094931.jpg': array([1.31800173]), './images/face/zhangyan/微信图片_20200610094755.jpg': array([1.49868196]), './images/face/zhangyan/微信图片_20200610094907.jpg': array([1.23562851]), './images/face/zhangyan/微信图片_20200610095116.jpg': array([1.45205302]), './images/face/zhangyan/微信图片_20200610094816.jpg': array([1.41599596]), './images/face/zhangyan/微信图片_20200610094813.jpg': array([1.44563089]), './images/face/zhangyan/微信图片_20200610094733.jpg': array([1.43045284]), './images/face/zhangyan/微信图片_20200610094637.jpg': array([1.42664555]), './images/face/zhangyan/微信图片_20200610095057.jpg': array([1.4036199]), './images/face/zhangyan/微信图片_20200610095107.jpg': array([1.36908389]), './images/face/zhangyan/微信图片_20200610095110.jpg': array([1.44176337]), './images/face/zhangyan/微信图片_20200610094854.jpg': array([1.3750708]), './images/face/zhangyan/微信图片_20200610094843.jpg': array([1.10185349]), './images/face/zhangyan/微信图片_20200610094957.jpg': array([1.42322438]), './images/face/zhangyan/微信图片_20200610095102.jpg': array([1.4726391]), './images/face/zhangyan/微信图片_20200610094821.jpg': array([1.48223118]), './images/face/zhangyan/微信图片_20200610095023.jpg': array([1.37674752]), './images/face/zhangyan/微信图片_20200610094739.jpg': array([1.45321288]), './images/face/zhangyan/微信图片_20200610094910.jpg': array([1.29580394]), './images/face/zhangyan/微信图片_20200610094927.jpg': array([1.28625453]), './images/face/zhangyan/微信图片_20200610094745.jpg': array([1.25934079]), './images/face/zhangyan/微信图片_20200610094818.jpg': array([1.45372353]), './images/face/zhangyan/微信图片_20200610094838.jpg': array([1.61909876]), './images/face/zhangyan/微信图片_20200610095059.jpg': array([1.38998636]), './images/face/zhangyan/微信图片_20200610094723.jpg': array([1.38340825]), './images/face/zhangyan/微信图片_20200610094905.jpg': array([1.26632732]), './images/face/zhangyan/微信图片_20200610095052.jpg': array([1.37002863]), './images/face/zhangyan/微信图片_20200610095002.jpg': array([1.42453864]), './images/face/zhangyan/微信图片_20200610094947.jpg': array([1.34753013]), './images/face/zhangyan/微信图片_20200610095049.jpg': array([1.42709705]), './images/face/zhangyan/微信图片_20200610095013.jpg': array([1.2665734]), './images/face/zhangyan/微信图片_20200610094827.jpg': array([1.45358197]), './images/face/zhangyan/微信图片_20200610094940.jpg': array([1.34940905]), './images/face/zhangyan/微信图片_20200610095025.jpg': array([1.5754618]), './images/face/zhangyan/微信图片_20200610094936.jpg': array([1.23344318]), './images/face/zhangyan/微信图片_20200610094807.jpg': array([1.51964492]), './images/face/zhangyan/微信图片_20200610094829.jpg': array([1.66925551]), './images/face/zhangyan/微信图片_20200610094727.jpg': array([1.63673877]), './images/face/zhangyan/微信图片_20200610094945.jpg': array([1.46576578]), './images/face/zhangyan/微信图片_20200610094710.jpg': array([1.46857931]), './images/face/zhangyan/微信图片_20200610094841.jpg': array([1.17614694]), './images/face/zhangyan/微信图片_20200610095104.jpg': array([1.36848266]), './images/face/zhangyan/微信图片_20200610094707.jpg': array([1.48325496]), './images/face/zhangyan/微信图片_20200610094942.jpg': array([1.49676983]), './images/face/zhangyan/微信图片_20200610094848.jpg': array([1.2163845]), './images/face/zhangyan/微信图片_20200610095044.jpg': array([1.3638586]), './images/face/zhangyan/微信图片_20200610094949.jpg': array([1.33669862]), './images/face/zhangyan/微信图片_20200610095033.jpg': array([1.34660268]), './images/face/zhangyan/微信图片_20200610094922.jpg': array([1.30191265]), './images/face/zhangyan/微信图片_20200610094920.jpg': array([1.27146845]), './images/face/zhangyan/微信图片_20200610094714.jpg': array([1.40402483]), './images/face/zhangyan/微信图片_20200610095030.jpg': array([1.6164128]), './images/face/zhangyan/微信图片_20200610094720.jpg': array([1.33625933]), './images/face/zhangyan/微信图片_20200610095054.jpg': array([1.36998378]), './images/face/zhangyan/微信图片_20200610095020.jpg': array([1.44767905]), './images/face/zhangyan/微信图片_20200610095010.jpg': array([1.30471948]), './images/face/zhangyan/微信图片_20200610094852.jpg': array([1.27815612]), './images/face/zhangyan/微信图片_20200610095039.jpg': array([1.29049665]), './images/face/zhangyan/微信图片_20200610094730.jpg': array([1.41347737]), './images/face/zhangyan/微信图片_20200610095118.jpg': array([1.45073988]), './images/face/zhangyan/微信图片_20200610095036.jpg': array([1.4174847]), './images/face/zhangyan/微信图片_20200610094917.jpg': array([1.33239932]), './images/face/zhangyan/微信图片_20200610094749.jpg': array([1.45273752]), './images/face/zhangyan/微信图片_20200610095113.jpg': array([1.4225342]), './images/face/zhangyan/微信图片_20200610094736.jpg': array([1.41597866]), './images/face/zhangyan/微信图片_20200610094846.jpg': array([0.96936816]), './images/face/zhangyan/微信图片_20200610094804.jpg': array([1.3588735]), './images/face/zhangyan/微信图片_20200610094913.jpg': array([1.31100633]), './images/face/zhangyan/微信图片_20200610095016.jpg': array([1.40377529]), './images/face/zhangyan/微信图片_20200610095041.jpg': array([1.32200666]), './images/face/zhangyan/微信图片_20200610095004.jpg': array([1.47245747]), './images/face/zhangyan/微信图片_20200610095007.jpg': array([1.36417812]), './images/face/zhangyan/微信图片_20200610094759.jpg': array([1.47242467]), './images/face/zhangyan/微信图片_20200610094933.jpg': array([1.30643071]), './images/face/zhangyan/微信图片_20200610094824.jpg': array([1.46212279]), './images/face/zhangyan/微信图片_20200610094801.jpg': array([1.53770813]), './images/face/zhangyan/微信图片_20200610094835.jpg': array([1.64134314]), './images/face/zhangyan/微信图片_20200610094900.jpg': array([1.20902341]), './images/face/zhangyan/微信图片_20200610094717.jpg': array([1.33425928]), './images/face/zhangyan/微信图片_20200610095000.jpg': array([1.33434164]), './images/face/zhangyan/微信图片_20200610094857.jpg': array([1.26627066]), './images/face/zhangziyi/20181206134403.png': array([0.88353688]), './images/face/zhangziyi/20181206132442.png': array([0.48116018]), './images/face/zhangziyi/20181206134529.png': array([0.99082223]), './images/face/zhangziyi/20181206132200.png': array([0.30981416]), './images/face/zhangziyi/20181206133927.png': array([0.28675244]), './images/face/zhangziyi/20181206134456.png': array([1.02641146]), './images/face/zhangziyi/20181206144652.png': array([0.26957104]), './images/face/zhangziyi/20181206132255.png': array([0.3939704]), './images/face/zhangziyi/20181206143714.png': array([0.34643213]), './images/face/zhangziyi/20181206131856.png': array([0.12617998]), './images/face/zhangziyi/20181206144938.png': array([0.27521766]), './images/face/zhangziyi/20181206145333.png': array([0.36129121]), './images/face/zhangziyi/20181206145456.png': array([0.19581287]), './images/face/zhangziyi/20181206144418.png': array([0.31410392]), './images/face/zhangziyi/20181206132102.png': array([0.67569057]), './images/face/zhangziyi/20181206145232.png': array([0.82238368]), './images/face/zhangziyi/20181206144721.png': array([0.44187809]), './images/face/zhangziyi/20181206112504.png': array([0.40999551]), './images/face/zhangziyi/20181206132122.png': array([0.53674142]), './images/face/zhangziyi/20181206132648.png': array([0.54507757]), './images/face/zhangziyi/20181206134557.png': array([0.98732268]), './images/face/zhangziyi/20181206132700.png': array([0.33713975]), './images/face/zhangziyi/20181206134004.png': array([0.55389936]), './images/face/zhangziyi/20181206133619.png': array([0.21859729]), './images/face/zhangziyi/20181206133521.png': array([0.51047612]), './images/face/zhangziyi/20181206134043.png': array([0.1938993]), './images/face/zhangziyi/20181206134640.png': array([0.39580052]), './images/face/zhangziyi/20181206132514.png': array([0.24444075]), './images/face/zhangziyi/20181206132327.png': array([0.50561499]), './images/face/zhangziyi/20181206135607.png': array([0.34922424]), './images/face/zhangziyi/20181206132141.png': array([0.26472148]), './images/face/zhangziyi/20181206132616.png': array([0.23440749]), './images/face/zhangziyi/20181206134251.png': array([0.26223524]), './images/face/zhangziyi/20181206140000.png': array([0.35654955]), './images/face/zhangziyi/20181206133415.png': array([0.13295161]), './images/face/zhangziyi/20181206133833.png': array([0.35204641]), './images/face/zhangziyi/20181206112529.png': array([0.29755109]), './images/face/zhangziyi/20181206134622.png': array([0.26536308]), './images/face/zhangziyi/20181206144453.png': array([0.34620007]), './images/face/zhangziyi/20181206132933.png': array([0.31895432]), './images/face/zhangziyi/20181206143748.png': array([0.24935962]), './images/face/zhangziyi/20181206134229.png': array([0.68330467]), './images/face/zhangziyi/20181206135812.png': array([0.18084772]), './images/face/zhangziyi/20181206132738.png': array([0.17173088]), './images/face/zhangziyi/20181206132220.png': array([0.50930203]), './images/face/zhangziyi/20181206143730.png': array([0.24941719]), './images/face/zhangziyi/20181206132022.png': array([0.16599039]), './images/face/zhangziyi/20181206132239.png': array([0.10651628]), './images/face/zhangziyi/20181206135919.png': array([0.36832603]), './images/face/zhangziyi/20181206132755.png': array([0.19518674]), './images/face/zhangziyi/20181206134123.png': array([0.23307154]), './images/face/zhangziyi/20181206134213.png': array([0.37357194]), './images/face/zhangziyi/20181206133002.png': array([0.59603765]), './images/face/zhangziyi/20181206132533.png': array([0.462161]), './images/face/zhangziyi/20181206134542.png': array([0.56110268]), './images/face/zhangziyi/20181206145348.png': array([0.54998463]), './images/face/zhangziyi/20181206132425.png': array([0.17082641]), './images/face/zhangziyi/20181206132406.png': array([0.18143326]), './images/face/zhangziyi/20181206144613.png': array([0.17878332]), './images/face/zhangziyi/20181206132919.png': array([0.80354679]), './images/face/zhangziyi/20181206145015.png': array([0.28710437]), './images/face/zhangziyi/20181206135034.png': array([0.60284176]), './images/face/zhangziyi/20181206145427.png': array([0.23263349]), './images/face/zhangziyi/20181206134938.png': array([0.32406395]), './images/face/zhangziyi/20181206145247.png': array([0.34272185]), './images/face/zhangziyi/20181206145110.png': array([0.27744102]), './images/face/zhangziyi/20181206134919.png': array([0.48750972]), './images/face/zhangziyi/20181206144436.png': array([0.80489776]), './images/face/zhangziyi/20181206135830.png': array([0.23264912]), './images/face/zhangziyi/20181206112348.png': array([0.]), './images/face/zhangziyi/20181206135321.png': array([0.4125828]), './images/face/zhangziyi/20181206134103.png': array([0.52651775]), './images/face/zhangziyi/20181206145122.png': array([0.53984661]), './images/face/zhangziyi/20181206145406.png': array([0.14730191]), './images/face/zhangziyi/20181206132346.png': array([0.60278177]), './images/face/zhangziyi/20181206134152.png': array([0.75190147]), './images/face/zhangziyi/20181206133539.png': array([0.26393235]), './images/face/zhangziyi/20181206134023.png': array([0.28628948]), './images/face/zhangziyi/20181206132629.png': array([0.1992753]), './images/face/zhangziyi/20181206135221.png': array([0.32202104]), './images/face/zhangziyi/20181206135108.png': array([0.33252744]), './images/face/zhangziyi/20181206133700.png': array([0.9417706]), './images/face/zhangziyi/20181206132904.png': array([0.21572748]), './images/face/zhangziyi/20181206132718.png': array([0.4360616]), './images/face/zhangziyi/20181206144813.png': array([0.98963036]), './images/face/zhangziyi/20181206133913.png': array([0.53738866]), './images/face/zhangziyi/20181206145059.png': array([0.43137224]), './images/face/zhangziyi/20181206145308.png': array([0.1833224]), './images/face/zhangziyi/20181206133720.png': array([0.69095899]), './images/face/zhangziyi/20181206145155.png': array([0.96426342]), './images/face/zhangziyi/20181206133344.png': array([0.3303623]), './images/face/zhangziyi/20181206132812.png': array([0.35305147]), './images/face/zhangziyi/20181206131958.png': array([0.45496102]), './images/face/zhangziyi/20181206144850.png': array([1.03903532]), './images/face/zhangziyi/20181206135126.png': array([0.2094741]), './images/face/zhangziyi/20181206132043.png': array([0.32931265]), './images/face/zhangziyi/20181206133020.png': array([0.67039429]), './images/face/zhangziyi/20181206132554.png': array([0.33953602]), './images/face/zhangziyi/20181206133815.png': array([0.94777137]), './images/face/zhangziyi/20181206131927.png': array([0.18944839]), './images/face/zhaoliying/159116217559.jpg': array([1.51032588]), './images/face/zhaoliying/159116217562.jpg': array([1.57409008]), './images/face/zhaoliying/159116065886.jpg': array([1.52132373]), './images/face/zhaoliying/159116217337.jpg': array([1.34621146]), './images/face/zhaoliying/159116065729.jpg': array([1.67140987]), './images/face/zhaoliying/159116217111.jpg': array([1.20376896]), './images/face/zhaoliying/159116217563.jpg': array([1.6045596]), './images/face/zhaoliying/159116065744.jpg': array([1.53007807]), './images/face/zhaoliying/15911621712.jpg': array([1.32116419]), './images/face/zhaoliying/159116217113.jpg': array([1.59309382]), './images/face/zhaoliying/159116065876.jpg': array([1.15263871]), './images/face/zhaoliying/159116217449.jpg': array([1.66641968]), './images/face/zhaoliying/159116217333.jpg': array([1.61282228]), './images/face/zhaoliying/1591160659131.jpg': array([1.49011479]), './images/face/zhaoliying/159116065859.jpg': array([1.45963945]), './images/face/zhaoliying/159116217224.jpg': array([1.37889482]), './images/face/zhaoliying/159116065888.jpg': array([1.24148988]), './images/face/zhaoliying/159116217450.jpg': array([1.78684558]), './images/face/zhaoliying/159116217446.jpg': array([1.42994789]), './images/face/zhaoliying/159116217457.jpg': array([1.28625844]), './images/face/zhaoliying/159116065857.jpg': array([1.44118384]), './images/face/zhaoliying/159116217456.jpg': array([1.41714971]), './images/face/zhaoliying/15911621701.jpg': array([1.3293017]), './images/face/zhaoliying/159116065893.jpg': array([1.59209381]), './images/face/zhaoliying/159116065860.jpg': array([1.35538393]), './images/face/zhaoliying/159116217332.jpg': array([1.55363855]), './images/face/zhaoliying/159116217447.jpg': array([1.55834994]), './images/face/zhaoliying/1591160659125.jpg': array([1.29717671]), './images/face/zhaoliying/159116217214.jpg': array([1.49815306]), './images/face/zhaoliying/159116217219.jpg': array([1.37445788]), './images/face/zhaoliying/159116217676.jpg': array([1.42876776]), './images/face/zhaoliying/159116065736.jpg': array([1.76417703]), './images/face/zhaoliying/159116217672.jpg': array([1.32805748]), './images/face/zhaoliying/15911606578.jpg': array([1.32477088]), './images/face/zhaoliying/159116065742.jpg': array([1.26204985]), './images/face/zhaoliying/1591160658108.jpg': array([1.66337937]), './images/face/zhaoliying/15911621714.jpg': array([1.21178382]), './images/face/zhaoliying/15911621716.jpg': array([1.42017997]), './images/face/zhaoliying/15911621717.jpg': array([1.50684582]), './images/face/zhaoliying/159116217339.jpg': array([1.45362436]), './images/face/zhaoliying/1591160659124.jpg': array([1.67008903]), './images/face/zhaoliying/159116065867.jpg': array([1.27894006]), './images/face/zhaoliying/159116065741.jpg': array([1.41954609]), './images/face/zhaoliying/159116065712.jpg': array([1.13437303]), './images/face/zhaoliying/1591160659121.jpg': array([1.230206]), './images/face/zhaoliying/1591160659118.jpg': array([1.1902327]), './images/face/zhaoliying/159116065745.jpg': array([1.33896254]), './images/face/zhaoliying/159116065862.jpg': array([1.46772534]), './images/face/zhaoliying/15911606572.jpg': array([1.34194905]), './images/face/zhaoliying/159116065870.jpg': array([1.71904612]), './images/face/zhaoliying/159116217669.jpg': array([1.63016196]), './images/face/zhaoliying/1591160659126.jpg': array([1.63986365]), './images/face/zhaoliying/159116065899.jpg': array([1.60089009]), './images/face/zhaoliying/159116065861.jpg': array([1.28740701]), './images/face/zhaoliying/1591160659138.jpg': array([1.37530156]), './images/face/zhaoliying/159116217565.jpg': array([1.27443067]), './images/face/zhaoliying/159116065879.jpg': array([1.43015383]), './images/face/zhaoliying/159116217341.jpg': array([1.43121653]), './images/face/zhaoliying/159116065740.jpg': array([1.40934162]), './images/face/zhaoliying/15911621718.jpg': array([1.27835556]), './images/face/zhaoliying/159116217455.jpg': array([1.61926035]), './images/face/zhaoliying/159116065730.jpg': array([1.63267324]), './images/face/zhaoliying/159116217335.jpg': array([1.28582327]), './images/face/zhaoliying/1591160658109.jpg': array([1.75511553]), './images/face/zhaoliying/159116217221.jpg': array([1.50280885]), './images/face/zhaoliying/159116217560.jpg': array([1.56552606]), './images/face/zhaoliying/159116217677.jpg': array([1.7485893]), './images/face/zhaoliying/159116217331.jpg': array([1.53786172]), './images/face/zhaoliying/1591160659119.jpg': array([1.69479436]), './images/face/zhaoliying/1591160659122.jpg': array([1.59929588]), './images/face/zhaoliying/1591160658112.jpg': array([1.29815591]), './images/face/zhaoliying/159116065875.jpg': array([1.46539253]), './images/face/zhaoliying/159116065866.jpg': array([1.74589133]), './images/face/zhaoliying/159116065877.jpg': array([1.42855831]), './images/face/zhaoliying/1591160658101.jpg': array([1.56977396]), './images/face/zhaoliying/159116217228.jpg': array([1.38106009]), './images/face/zhaoliying/159116065878.jpg': array([1.34379831]), './images/face/zhaoliying/159116065710.jpg': array([1.76290442]), './images/face/zhaoliying/159116065743.jpg': array([1.54934715]), './images/face/zhaoliying/159116217223.jpg': array([1.21678337]), './images/face/zhaoliying/159116065716.jpg': array([1.57497229]), './images/face/zhaoliying/15911621719.jpg': array([1.4788245]), './images/face/zhaoliying/159116217680.jpg': array([1.21674219]), './images/face/zhaoliying/1591160658114.jpg': array([1.51667047]), './images/face/zhaoliying/159116217217.jpg': array([1.33673288]), './images/face/zhaoliying/1591160658104.jpg': array([1.51705552]), './images/face/zhaoliying/159116217445.jpg': array([1.45157413]), './images/face/zhaoliying/159116217218.jpg': array([1.48400377]), './images/face/zhaoliying/1591160659130.jpg': array([1.42493385]), './images/face/zhaoliying/159116217229.jpg': array([1.39802823]), './images/face/zhaoliying/159116217568.jpg': array([1.37545971]), './images/face/zhaoliying/15911606579.jpg': array([1.49436256]), './images/face/zhaoliying/159116217226.jpg': array([1.63498776]), './images/face/zhaoliying/159116065883.jpg': array([1.53083003]), './images/face/zhaoliying/159116217452.jpg': array([1.6720442]), './images/face/zhaoliying/159116217216.jpg': array([1.27804824]), './images/face/zhaoliying/159116217215.jpg': array([1.36444103]), './images/face/zhaoliying/1591160659116.jpg': array([1.37484662]), './images/face/zhaoliying/159116065754.jpg': array([1.65349632]), './images/face/zhaoliying/159116065874.jpg': array([1.47650631]), './images/face/zhaoliying/159116065873.jpg': array([1.52960284]), './images/face/zhaoliying/159116217334.jpg': array([1.52184688]), './images/face/zhaoliying/159116217678.jpg': array([1.66856643]), './images/face/zhaoliying/15911621713.jpg': array([1.48458757]), './images/face/zhaoliying/159116217336.jpg': array([1.30323252]), './images/face/zhaoliying/159116217567.jpg': array([1.28236027]), './images/face/zhaoliying/159116217112.jpg': array([1.27972987]), './images/face/zhaoliying/159116217448.jpg': array([1.42138489]), './images/face/zhaoliying/159116065751.jpg': array([1.49239388]), './images/face/zhaoliying/159116065871.jpg': array([1.48546285]), './images/face/zhaoliying/1591160659135.jpg': array([1.78301455]), './images/face/zhaoliying/159116217453.jpg': array([1.36210108]), './images/face/zhaoliying/159116217674.jpg': array([1.48425985]), './images/face/zhaoliying/1591160658103.jpg': array([1.39185723]), './images/face/zhaoliying/159116065898.jpg': array([1.63837884]), './images/face/zhaoliying/159116217338.jpg': array([1.5125434]), './images/face/zhaoliying/1591160659137.jpg': array([1.56223026]), './images/face/zhaoliying/159116217673.jpg': array([1.56268764]), './images/face/zhaoliying/1591160658102.jpg': array([1.65509142]), './images/face/zhaoliying/159116065726.jpg': array([1.46286181]), './images/face/zhaoliying/159116065718.jpg': array([1.58915727]), './images/face/zhaoliying/1591160658106.jpg': array([1.24936241]), './images/face/zhaoliying/1591160659120.jpg': array([1.42216457]), './images/face/zhaoliying/159116065896.jpg': array([1.33050469]), './images/face/zhaoliying/159116065721.jpg': array([1.55970006]), './images/face/zhaoliying/1591160658105.jpg': array([1.15990559]), './images/face/zhaoliying/159116065753.jpg': array([1.61850907]), './images/face/zhaoliying/159116217340.jpg': array([1.12549528]), './images/face/zhaoliying/159116217679.jpg': array([1.40564991]), './images/face/zhaoliying/159116065892.jpg': array([1.59906314]), './images/face/zhaoliying/159116217670.jpg': array([1.66880214]), './images/face/zhaoliying/159116217342.jpg': array([1.6334337]), './images/face/zhaoliying/159116065733.jpg': array([1.50316427]), './images/face/zhaoliying/159116217566.jpg': array([1.66268042]), './images/face/zhaoliying/1591160658113.jpg': array([1.3455708]), './images/face/zhaoliying/159116217343.jpg': array([1.33418995]), './images/face/zhaoliying/159116217675.jpg': array([1.52138626]), './images/face/zhaoliying/15911606577.jpg': array([1.60599294]), './images/face/zhaoliying/159116217561.jpg': array([1.6042454]), './images/face/zhaoliying/159116217110.jpg': array([1.75560387]), './images/face/zhaoliying/15911606574.jpg': array([1.22118916]), './images/face/zhaoliying/159116217225.jpg': array([1.55550983]), './images/face/zhaoliying/159116217227.jpg': array([1.58113151]), './images/face/zhaoliying/159116217444.jpg': array([1.49004941]), './images/face/zhaoliying/1591160658100.jpg': array([1.5078325]), './images/face/zhaoliying/159116065890.jpg': array([1.70108838]), './images/face/zhaoliying/1591160659117.jpg': array([1.28917726]), './images/face/zhaoliying/159116217558.jpg': array([1.33290413]), './images/face/zhaoliying/159116217671.jpg': array([1.14405126]), './images/face/zhaoliying/1591160659128.jpg': array([1.20121083]), './images/face/zhaoliying/159116217454.jpg': array([1.45329425]), './images/face/zhaoliying/159116217330.jpg': array([1.62202366]), './images/face/zhaoliying/159116217564.jpg': array([1.19868704]), './images/face/zhaoliying/159116217220.jpg': array([1.77561535]), './images/face/zhaoliying/159116065722.jpg': array([1.60813004]), './images/face/zhaoliying/159116065891.jpg': array([1.49938677]), './images/face/zhaoliying/159116065894.jpg': array([1.3454066])}\n" + ] + } + ], + "source": [ + "image_test_path = input('Input image filename:')\n", + "image_test = Image.open(image_test_path)\n", + "with paddle.no_grad():\n", + " image_test = resize_image(image_test, [input_shape[1], input_shape[0]], letterbox_image=True)\n", + "\n", + " photo_test = paddle.to_tensor(\n", + " np.expand_dims(np.transpose(preprocess_input(np.array(image_test, np.float32)), (2, 0, 1)), 0))\n", + "\n", + " photo_test = photo_test.cuda()\n", + " output_test = model(photo_test).cpu().numpy()\n", + "\n", + "total_diff = {}\n", + "for key,value in total_embedding.items():\n", + " value = np.array(value)\n", + " total_diff[key] = np.linalg.norm(output_test - value, axis=1)\n", + "\n", + "print(total_diff)\n" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 将与之最像的三张图展示出来" + ] + }, + { + "cell_type": "code", + "execution_count": 26, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T13:54:56.437929Z", + "iopub.status.busy": "2022-05-03T13:54:56.437647Z", + "iopub.status.idle": "2022-05-03T13:54:56.925512Z", + "shell.execute_reply": "2022-05-03T13:54:56.925059Z", + "shell.execute_reply.started": "2022-05-03T13:54:56.437905Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "import matplotlib.pyplot as plt\n", + "\n", + "sorted_total_diff = sorted(total_diff.items(),key=lambda x:x[1])\n", + "use_data = sorted_total_diff[:3]\n", + "\n", + "fig = plt.figure(figsize=(20, 8))\n", + "# src image\n", + "src_data_path = image_test_path\n", + "img = Image.open(src_data_path)\n", + "ax = fig.add_subplot(2, 3, 1)\n", + "ax.imshow(img)\n", + "for i in range(len(use_data)):\n", + " sub_data = use_data[i]\n", + " sub_data_path = sub_data[0]\n", + " img = Image.open(sub_data_path)\n", + " ax = fig.add_subplot(2, 3, i + 4)\n", + " ax.imshow(img)\n", + "\n", + "plt.show()" + ] + }, + { + "cell_type": "code", + "execution_count": 27, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T13:56:25.910771Z", + "iopub.status.busy": "2022-05-03T13:56:25.910445Z", + "iopub.status.idle": "2022-05-03T13:56:25.915943Z", + "shell.execute_reply": "2022-05-03T13:56:25.915513Z", + "shell.execute_reply.started": "2022-05-03T13:56:25.910743Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "data": { + "text/plain": [ + "[('./images/face/zhangziyi/20181206112348.png', array([0.])),\n", + " ('./images/face/zhangziyi/20181206132239.png', array([0.10651628])),\n", + " ('./images/face/zhangziyi/20181206131856.png', array([0.12617998]))]" + ] + }, + "execution_count": 27, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "use_data" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 输出类别名称" + ] + }, + { + "cell_type": "code", + "execution_count": 28, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T13:57:58.464809Z", + "iopub.status.busy": "2022-05-03T13:57:58.464485Z", + "iopub.status.idle": "2022-05-03T13:57:58.469942Z", + "shell.execute_reply": "2022-05-03T13:57:58.469500Z", + "shell.execute_reply.started": "2022-05-03T13:57:58.464785Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "{'zhangziyi': 5}\n", + "this is zhangziyi\n" + ] + } + ], + "source": [ + "import os\n", + "classes_dict = {}\n", + "for i in range(5):\n", + " sub_key = os.path.split(os.path.split(sorted_total_diff[i][0])[0])[1]\n", + " count = classes_dict.get(sub_key, 0)\n", + " count += 1\n", + " classes_dict[sub_key] = count\n", + "\n", + "print(classes_dict)\n", + "sorted_classes_dict = sorted(classes_dict.items(),key=lambda x:x[1])\n", + "print(\"this is\", sorted_classes_dict[-1][0])" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 使用 paddle.jit.to_static 实现动转静" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 改写网络" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:52:01.931554Z", + "iopub.status.busy": "2022-05-03T14:52:01.931163Z", + "iopub.status.idle": "2022-05-03T14:52:02.019329Z", + "shell.execute_reply": "2022-05-03T14:52:02.018821Z", + "shell.execute_reply.started": "2022-05-03T14:52:01.931512Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "class OwnResnet2(nn.Layer):\n", + " def __init__(self, pretrained):\n", + " super(OwnResnet2, self).__init__()\n", + " self.model = models.resnet50(pretrained=pretrained)\n", + " # 删除不需要的层结构\n", + " del self.model.fc\n", + " del self.model.avgpool\n", + "\n", + " def forward(self, x):\n", + " x = self.model.conv1(x)\n", + " x = self.model.bn1(x)\n", + " x = self.model.relu(x)\n", + " x = self.model.maxpool(x)\n", + "\n", + " x = self.model.layer1(x)\n", + " x = self.model.layer2(x)\n", + " x = self.model.layer3(x)\n", + " x = self.model.layer4(x)\n", + "\n", + " return x\n", + "\n", + "\n", + "class OwnFacenet2(nn.Layer):\n", + " def __init__(self, dropout_keep_prob=0.5, embedding_size=128, num_classes=None, pretrained=False):\n", + " super(OwnFacenet2, self).__init__()\n", + " self.backbone = OwnResnet2(pretrained=pretrained)\n", + " flat_shape = 2048\n", + " self.set_parameter_requires_grad(self.backbone, True)\n", + " self.avg = nn.AdaptiveAvgPool2D((1, 1))\n", + " self.flatten = paddle.nn.Flatten()\n", + " self.Dropout = nn.Dropout(1 - dropout_keep_prob)\n", + " self.Bottleneck = nn.Linear(flat_shape, embedding_size)\n", + " self.last_bn = nn.BatchNorm1D(embedding_size)\n", + " # if mode == \"train\":\n", + " self.classifier = nn.Linear(embedding_size, num_classes)\n", + "\n", + " def set_parameter_requires_grad(self, model, feature_extracting):\n", + " if feature_extracting:\n", + " for param in model.parameters():\n", + " param.requires_grad = True\n", + "\n", + " @paddle.jit.to_static()\n", + " def forward(self, x):\n", + " if self.training:\n", + " x = self.backbone(x)\n", + " x = self.avg(x)\n", + " x = self.flatten(x)\n", + " x = self.Dropout(x)\n", + " x = self.Bottleneck(x)\n", + " x = self.last_bn(x)\n", + " # 分类的结果\n", + " x1 = self.classifier(x)\n", + " # L2标准化之后的结果\n", + " x2 = F.normalize(x, p=2, axis=1)\n", + " return x1, x2\n", + " else:\n", + " x = self.backbone(x)\n", + " x = self.avg(x)\n", + " x = self.flatten(x)\n", + " x = self.Dropout(x)\n", + " x = self.Bottleneck(x)\n", + " x = self.last_bn(x)\n", + " # l2标准化后的结果\n", + " x = F.normalize(x, p=2, axis=1)\n", + " return x" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:52:05.168183Z", + "iopub.status.busy": "2022-05-03T14:52:05.167779Z", + "iopub.status.idle": "2022-05-03T14:52:15.305984Z", + "shell.execute_reply": "2022-05-03T14:52:15.305423Z", + "shell.execute_reply.started": "2022-05-03T14:52:05.168151Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "W0503 22:52:05.170619 170 device_context.cc:447] Please NOTE: device: 0, GPU Compute Capability: 8.0, Driver API Version: 11.2, Runtime API Version: 11.2\n", + "W0503 22:52:05.173643 170 device_context.cc:465] device: 0, cuDNN Version: 8.2.\n", + "100%|██████████| 151272/151272 [00:05<00:00, 28325.70it/s]\n" + ] + } + ], + "source": [ + "# build model\n", + "pretrained = True\n", + "model_2 = OwnFacenet2(num_classes=num_classes, pretrained=pretrained)" + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:52:16.985615Z", + "iopub.status.busy": "2022-05-03T14:52:16.985337Z", + "iopub.status.idle": "2022-05-03T14:52:16.989813Z", + "shell.execute_reply": "2022-05-03T14:52:16.989420Z", + "shell.execute_reply.started": "2022-05-03T14:52:16.985595Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "# build optim\n", + "optimizer_type = 'adam'\n", + "optimizer = {\n", + " 'adam': optim.Adam(parameters = model_2.parameters(), learning_rate = 0.001),\n", + " 'sgd': optim.SGD(parameters = model_2.parameters(), learning_rate = 0.01)\n", + "}[optimizer_type]" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:52:20.604446Z", + "iopub.status.busy": "2022-05-03T14:52:20.604195Z", + "iopub.status.idle": "2022-05-03T14:52:22.026671Z", + "shell.execute_reply": "2022-05-03T14:52:22.026138Z", + "shell.execute_reply.started": "2022-05-03T14:52:20.604423Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddle/fluid/layers/utils.py:77: DeprecationWarning: Using or importing the ABCs from 'collections' instead of from 'collections.abc' is deprecated, and in 3.8 it will stop working\n", + " return (isinstance(seq, collections.Sequence) and\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "-------------------------------------------------------------------------------\n", + " Layer (type) Input Shape Output Shape Param # \n", + "===============================================================================\n", + " Conv2D-1 [[10, 3, 224, 224]] [10, 64, 112, 112] 9,408 \n", + " BatchNorm2D-1 [[10, 64, 112, 112]] [10, 64, 112, 112] 256 \n", + " ReLU-1 [[10, 64, 112, 112]] [10, 64, 112, 112] 0 \n", + " MaxPool2D-1 [[10, 64, 112, 112]] [10, 64, 56, 56] 0 \n", + " Conv2D-3 [[10, 64, 56, 56]] [10, 64, 56, 56] 4,096 \n", + " BatchNorm2D-3 [[10, 64, 56, 56]] [10, 64, 56, 56] 256 \n", + " ReLU-2 [[10, 256, 56, 56]] [10, 256, 56, 56] 0 \n", + " Conv2D-4 [[10, 64, 56, 56]] [10, 64, 56, 56] 36,864 \n", + " BatchNorm2D-4 [[10, 64, 56, 56]] [10, 64, 56, 56] 256 \n", + " Conv2D-5 [[10, 64, 56, 56]] [10, 256, 56, 56] 16,384 \n", + " BatchNorm2D-5 [[10, 256, 56, 56]] [10, 256, 56, 56] 1,024 \n", + " Conv2D-2 [[10, 64, 56, 56]] [10, 256, 56, 56] 16,384 \n", + " BatchNorm2D-2 [[10, 256, 56, 56]] [10, 256, 56, 56] 1,024 \n", + " BottleneckBlock-1 [[10, 64, 56, 56]] [10, 256, 56, 56] 76,288 \n", + " Conv2D-6 [[10, 256, 56, 56]] [10, 64, 56, 56] 16,384 \n", + " BatchNorm2D-6 [[10, 64, 56, 56]] [10, 64, 56, 56] 256 \n", + " ReLU-3 [[10, 256, 56, 56]] [10, 256, 56, 56] 0 \n", + " Conv2D-7 [[10, 64, 56, 56]] [10, 64, 56, 56] 36,864 \n", + " BatchNorm2D-7 [[10, 64, 56, 56]] [10, 64, 56, 56] 256 \n", + " Conv2D-8 [[10, 64, 56, 56]] [10, 256, 56, 56] 16,384 \n", + " BatchNorm2D-8 [[10, 256, 56, 56]] [10, 256, 56, 56] 1,024 \n", + " BottleneckBlock-2 [[10, 256, 56, 56]] [10, 256, 56, 56] 71,168 \n", + " Conv2D-9 [[10, 256, 56, 56]] [10, 64, 56, 56] 16,384 \n", + " BatchNorm2D-9 [[10, 64, 56, 56]] [10, 64, 56, 56] 256 \n", + " ReLU-4 [[10, 256, 56, 56]] [10, 256, 56, 56] 0 \n", + " Conv2D-10 [[10, 64, 56, 56]] [10, 64, 56, 56] 36,864 \n", + " BatchNorm2D-10 [[10, 64, 56, 56]] [10, 64, 56, 56] 256 \n", + " Conv2D-11 [[10, 64, 56, 56]] [10, 256, 56, 56] 16,384 \n", + " BatchNorm2D-11 [[10, 256, 56, 56]] [10, 256, 56, 56] 1,024 \n", + " BottleneckBlock-3 [[10, 256, 56, 56]] [10, 256, 56, 56] 71,168 \n", + " Conv2D-13 [[10, 256, 56, 56]] [10, 128, 56, 56] 32,768 \n", + " BatchNorm2D-13 [[10, 128, 56, 56]] [10, 128, 56, 56] 512 \n", + " ReLU-5 [[10, 512, 28, 28]] [10, 512, 28, 28] 0 \n", + " Conv2D-14 [[10, 128, 56, 56]] [10, 128, 28, 28] 147,456 \n", + " BatchNorm2D-14 [[10, 128, 28, 28]] [10, 128, 28, 28] 512 \n", + " Conv2D-15 [[10, 128, 28, 28]] [10, 512, 28, 28] 65,536 \n", + " BatchNorm2D-15 [[10, 512, 28, 28]] [10, 512, 28, 28] 2,048 \n", + " Conv2D-12 [[10, 256, 56, 56]] [10, 512, 28, 28] 131,072 \n", + " BatchNorm2D-12 [[10, 512, 28, 28]] [10, 512, 28, 28] 2,048 \n", + " BottleneckBlock-4 [[10, 256, 56, 56]] [10, 512, 28, 28] 381,952 \n", + " Conv2D-16 [[10, 512, 28, 28]] [10, 128, 28, 28] 65,536 \n", + " BatchNorm2D-16 [[10, 128, 28, 28]] [10, 128, 28, 28] 512 \n", + " ReLU-6 [[10, 512, 28, 28]] [10, 512, 28, 28] 0 \n", + " Conv2D-17 [[10, 128, 28, 28]] [10, 128, 28, 28] 147,456 \n", + " BatchNorm2D-17 [[10, 128, 28, 28]] [10, 128, 28, 28] 512 \n", + " Conv2D-18 [[10, 128, 28, 28]] [10, 512, 28, 28] 65,536 \n", + " BatchNorm2D-18 [[10, 512, 28, 28]] [10, 512, 28, 28] 2,048 \n", + " BottleneckBlock-5 [[10, 512, 28, 28]] [10, 512, 28, 28] 281,600 \n", + " Conv2D-19 [[10, 512, 28, 28]] [10, 128, 28, 28] 65,536 \n", + " BatchNorm2D-19 [[10, 128, 28, 28]] [10, 128, 28, 28] 512 \n", + " ReLU-7 [[10, 512, 28, 28]] [10, 512, 28, 28] 0 \n", + " Conv2D-20 [[10, 128, 28, 28]] [10, 128, 28, 28] 147,456 \n", + " BatchNorm2D-20 [[10, 128, 28, 28]] [10, 128, 28, 28] 512 \n", + " Conv2D-21 [[10, 128, 28, 28]] [10, 512, 28, 28] 65,536 \n", + " BatchNorm2D-21 [[10, 512, 28, 28]] [10, 512, 28, 28] 2,048 \n", + " BottleneckBlock-6 [[10, 512, 28, 28]] [10, 512, 28, 28] 281,600 \n", + " Conv2D-22 [[10, 512, 28, 28]] [10, 128, 28, 28] 65,536 \n", + " BatchNorm2D-22 [[10, 128, 28, 28]] [10, 128, 28, 28] 512 \n", + " ReLU-8 [[10, 512, 28, 28]] [10, 512, 28, 28] 0 \n", + " Conv2D-23 [[10, 128, 28, 28]] [10, 128, 28, 28] 147,456 \n", + " BatchNorm2D-23 [[10, 128, 28, 28]] [10, 128, 28, 28] 512 \n", + " Conv2D-24 [[10, 128, 28, 28]] [10, 512, 28, 28] 65,536 \n", + " BatchNorm2D-24 [[10, 512, 28, 28]] [10, 512, 28, 28] 2,048 \n", + " BottleneckBlock-7 [[10, 512, 28, 28]] [10, 512, 28, 28] 281,600 \n", + " Conv2D-26 [[10, 512, 28, 28]] [10, 256, 28, 28] 131,072 \n", + " BatchNorm2D-26 [[10, 256, 28, 28]] [10, 256, 28, 28] 1,024 \n", + " ReLU-9 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 0 \n", + " Conv2D-27 [[10, 256, 28, 28]] [10, 256, 14, 14] 589,824 \n", + " BatchNorm2D-27 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " Conv2D-28 [[10, 256, 14, 14]] [10, 1024, 14, 14] 262,144 \n", + " BatchNorm2D-28 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 4,096 \n", + " Conv2D-25 [[10, 512, 28, 28]] [10, 1024, 14, 14] 524,288 \n", + " BatchNorm2D-25 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 4,096 \n", + " BottleneckBlock-8 [[10, 512, 28, 28]] [10, 1024, 14, 14] 1,517,568 \n", + " Conv2D-29 [[10, 1024, 14, 14]] [10, 256, 14, 14] 262,144 \n", + " BatchNorm2D-29 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " ReLU-10 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 0 \n", + " Conv2D-30 [[10, 256, 14, 14]] [10, 256, 14, 14] 589,824 \n", + " BatchNorm2D-30 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " Conv2D-31 [[10, 256, 14, 14]] [10, 1024, 14, 14] 262,144 \n", + " BatchNorm2D-31 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 4,096 \n", + " BottleneckBlock-9 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 1,120,256 \n", + " Conv2D-32 [[10, 1024, 14, 14]] [10, 256, 14, 14] 262,144 \n", + " BatchNorm2D-32 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " ReLU-11 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 0 \n", + " Conv2D-33 [[10, 256, 14, 14]] [10, 256, 14, 14] 589,824 \n", + " BatchNorm2D-33 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " Conv2D-34 [[10, 256, 14, 14]] [10, 1024, 14, 14] 262,144 \n", + " BatchNorm2D-34 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 4,096 \n", + "BottleneckBlock-10 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 1,120,256 \n", + " Conv2D-35 [[10, 1024, 14, 14]] [10, 256, 14, 14] 262,144 \n", + " BatchNorm2D-35 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " ReLU-12 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 0 \n", + " Conv2D-36 [[10, 256, 14, 14]] [10, 256, 14, 14] 589,824 \n", + " BatchNorm2D-36 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " Conv2D-37 [[10, 256, 14, 14]] [10, 1024, 14, 14] 262,144 \n", + " BatchNorm2D-37 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 4,096 \n", + "BottleneckBlock-11 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 1,120,256 \n", + " Conv2D-38 [[10, 1024, 14, 14]] [10, 256, 14, 14] 262,144 \n", + " BatchNorm2D-38 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " ReLU-13 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 0 \n", + " Conv2D-39 [[10, 256, 14, 14]] [10, 256, 14, 14] 589,824 \n", + " BatchNorm2D-39 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " Conv2D-40 [[10, 256, 14, 14]] [10, 1024, 14, 14] 262,144 \n", + " BatchNorm2D-40 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 4,096 \n", + "BottleneckBlock-12 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 1,120,256 \n", + " Conv2D-41 [[10, 1024, 14, 14]] [10, 256, 14, 14] 262,144 \n", + " BatchNorm2D-41 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " ReLU-14 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 0 \n", + " Conv2D-42 [[10, 256, 14, 14]] [10, 256, 14, 14] 589,824 \n", + " BatchNorm2D-42 [[10, 256, 14, 14]] [10, 256, 14, 14] 1,024 \n", + " Conv2D-43 [[10, 256, 14, 14]] [10, 1024, 14, 14] 262,144 \n", + " BatchNorm2D-43 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 4,096 \n", + "BottleneckBlock-13 [[10, 1024, 14, 14]] [10, 1024, 14, 14] 1,120,256 \n", + " Conv2D-45 [[10, 1024, 14, 14]] [10, 512, 14, 14] 524,288 \n", + " BatchNorm2D-45 [[10, 512, 14, 14]] [10, 512, 14, 14] 2,048 \n", + " ReLU-15 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 0 \n", + " Conv2D-46 [[10, 512, 14, 14]] [10, 512, 7, 7] 2,359,296 \n", + " BatchNorm2D-46 [[10, 512, 7, 7]] [10, 512, 7, 7] 2,048 \n", + " Conv2D-47 [[10, 512, 7, 7]] [10, 2048, 7, 7] 1,048,576 \n", + " BatchNorm2D-47 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 8,192 \n", + " Conv2D-44 [[10, 1024, 14, 14]] [10, 2048, 7, 7] 2,097,152 \n", + " BatchNorm2D-44 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 8,192 \n", + "BottleneckBlock-14 [[10, 1024, 14, 14]] [10, 2048, 7, 7] 6,049,792 \n", + " Conv2D-48 [[10, 2048, 7, 7]] [10, 512, 7, 7] 1,048,576 \n", + " BatchNorm2D-48 [[10, 512, 7, 7]] [10, 512, 7, 7] 2,048 \n", + " ReLU-16 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 0 \n", + " Conv2D-49 [[10, 512, 7, 7]] [10, 512, 7, 7] 2,359,296 \n", + " BatchNorm2D-49 [[10, 512, 7, 7]] [10, 512, 7, 7] 2,048 \n", + " Conv2D-50 [[10, 512, 7, 7]] [10, 2048, 7, 7] 1,048,576 \n", + " BatchNorm2D-50 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 8,192 \n", + "BottleneckBlock-15 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 4,468,736 \n", + " Conv2D-51 [[10, 2048, 7, 7]] [10, 512, 7, 7] 1,048,576 \n", + " BatchNorm2D-51 [[10, 512, 7, 7]] [10, 512, 7, 7] 2,048 \n", + " ReLU-17 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 0 \n", + " Conv2D-52 [[10, 512, 7, 7]] [10, 512, 7, 7] 2,359,296 \n", + " BatchNorm2D-52 [[10, 512, 7, 7]] [10, 512, 7, 7] 2,048 \n", + " Conv2D-53 [[10, 512, 7, 7]] [10, 2048, 7, 7] 1,048,576 \n", + " BatchNorm2D-53 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 8,192 \n", + "BottleneckBlock-16 [[10, 2048, 7, 7]] [10, 2048, 7, 7] 4,468,736 \n", + " OwnResnet2-1 [[10, 3, 224, 224]] [10, 2048, 7, 7] 23,561,152 \n", + "AdaptiveAvgPool2D-2 [[10, 2048, 7, 7]] [10, 2048, 1, 1] 0 \n", + " Flatten-1 [[10, 2048, 1, 1]] [10, 2048] 0 \n", + " Dropout-1 [[10, 2048]] [10, 2048] 0 \n", + " Linear-2 [[10, 2048]] [10, 128] 262,272 \n", + " BatchNorm1D-1 [[10, 128]] [10, 128] 512 \n", + "===============================================================================\n", + "Total params: 70,936,576\n", + "Trainable params: 70,936,576\n", + "Non-trainable params: 0\n", + "-------------------------------------------------------------------------------\n", + "Input size (MB): 5.74\n", + "Forward/backward pass size (MB): 2622.75\n", + "Params size (MB): 270.60\n", + "Estimated Total Size (MB): 2899.10\n", + "-------------------------------------------------------------------------------\n", + "\n", + "{'total_params': 70936576, 'trainable_params': 70936576}\n" + ] + } + ], + "source": [ + "model_info = paddle.summary(model_2, (10, 3, 224, 224))\n", + "print(model_info)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 模型训练" + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:53:40.159127Z", + "iopub.status.busy": "2022-05-03T14:53:40.158762Z", + "iopub.status.idle": "2022-05-03T14:54:22.421742Z", + "shell.execute_reply": "2022-05-03T14:54:22.420529Z", + "shell.execute_reply.started": "2022-05-03T14:53:40.159095Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 1/10: 100%|██████████| 47/47 [00:03<00:00, 11.83it/s, accuracy=1, total_CE_loss=0.977, total_triple_loss=0.0426]\n", + "Epoch 2/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:1/10\n", + "Total Loss: 1.0193\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 2/10: 100%|██████████| 47/47 [00:03<00:00, 11.89it/s, accuracy=1, total_CE_loss=0.492, total_triple_loss=0.02] \n", + "Epoch 3/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:2/10\n", + "Total Loss: 0.5124\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 3/10: 100%|██████████| 47/47 [00:04<00:00, 11.73it/s, accuracy=1, total_CE_loss=0.393, total_triple_loss=0.0158]\n", + "Epoch 4/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:3/10\n", + "Total Loss: 0.4087\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 4/10: 100%|██████████| 47/47 [00:03<00:00, 11.79it/s, accuracy=1, total_CE_loss=0.542, total_triple_loss=0.0312]\n", + "Epoch 5/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:4/10\n", + "Total Loss: 0.5735\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 5/10: 100%|██████████| 47/47 [00:04<00:00, 11.74it/s, accuracy=1, total_CE_loss=0.289, total_triple_loss=0.0133] \n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:5/10\n", + "Total Loss: 0.3026\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 6/10: 100%|██████████| 47/47 [00:04<00:00, 11.04it/s, accuracy=1, total_CE_loss=0.323, total_triple_loss=0.00919]\n", + "Epoch 7/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:6/10\n", + "Total Loss: 0.3324\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 7/10: 100%|██████████| 47/47 [00:04<00:00, 10.16it/s, accuracy=1, total_CE_loss=0.286, total_triple_loss=0.0301]\n", + "Epoch 8/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:7/10\n", + "Total Loss: 0.3166\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 8/10: 100%|██████████| 47/47 [00:04<00:00, 10.94it/s, accuracy=1, total_CE_loss=0.343, total_triple_loss=0.0116]\n", + "Epoch 9/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:8/10\n", + "Total Loss: 0.3548\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 9/10: 100%|██████████| 47/47 [00:04<00:00, 10.90it/s, accuracy=1, total_CE_loss=0.385, total_triple_loss=0.0291]\n", + "Epoch 10/10: 0%| | 0/47 [00:00]" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:9/10\n", + "Total Loss: 0.4138\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Epoch 10/10: 100%|██████████| 47/47 [00:04<00:00, 10.92it/s, accuracy=1, total_CE_loss=0.251, total_triple_loss=0.0129]\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Epoch:10/10\n", + "Total Loss: 0.2637\n" + ] + } + ], + "source": [ + "epoch_step = num_train // batch_size\n", + "epochs = 10\n", + "save_period = 5\n", + "save_dir = \"./output2/\"\n", + "# train\n", + "for epoch in range(epochs):\n", + " fit_one_epoch(model_2, loss, optimizer,\n", + " epoch, epoch_step, train_loader,\n", + " epochs, batch_size // 3, save_period, save_dir)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 使用 paddle.jit.save 保存动转静模型" + ] + }, + { + "cell_type": "code", + "execution_count": 24, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-03T14:55:33.972791Z", + "iopub.status.busy": "2022-05-03T14:55:33.972398Z", + "iopub.status.idle": "2022-05-03T14:55:34.782325Z", + "shell.execute_reply": "2022-05-03T14:55:34.781592Z", + "shell.execute_reply.started": "2022-05-03T14:55:33.972764Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Tue May 03 22:55:33 Dynamic-to-Static WARNING: Current function: forward(x), input_spec: None has more than one cached programs: 2, the last traced progam will be return by default.\n" + ] + } + ], + "source": [ + "paddle.jit.save(model_2, 'model')" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "scrolled": true + }, + "outputs": [], + "source": [] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "请点击[此处](https://ai.baidu.com/docs#/AIStudio_Project_Notebook/a38e5576)查看本环境基本用法.
\n", + "Please click [here ](https://ai.baidu.com/docs#/AIStudio_Project_Notebook/a38e5576) for more detailed instructions. " + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "py35-paddle1.2.0" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.7.4" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} diff --git a/docs/practices/jit/Image denoising-jit.ipynb b/docs/practices/jit/Image denoising-jit.ipynb new file mode 100644 index 00000000000..84aea6db54b --- /dev/null +++ b/docs/practices/jit/Image denoising-jit.ipynb @@ -0,0 +1,1297 @@ +{ + "cells": [ + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T08:34:20.989158Z", + "iopub.status.busy": "2022-05-04T08:34:20.988840Z", + "iopub.status.idle": "2022-05-04T08:34:21.185186Z", + "shell.execute_reply": "2022-05-04T08:34:21.184481Z", + "shell.execute_reply.started": "2022-05-04T08:34:20.989130Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "# 查看当前挂载的数据集目录, 该目录下的变更重启环境后会自动还原\n", + "# View dataset directory. \n", + "# This directory will be recovered automatically after resetting environment. \n", + "!ls /home/aistudio/data" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T08:34:23.044833Z", + "iopub.status.busy": "2022-05-04T08:34:23.044557Z", + "iopub.status.idle": "2022-05-04T08:34:23.241258Z", + "shell.execute_reply": "2022-05-04T08:34:23.240528Z", + "shell.execute_reply.started": "2022-05-04T08:34:23.044805Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "# 查看工作区文件, 该目录下的变更将会持久保存. 请及时清理不必要的文件, 避免加载过慢.\n", + "# View personal work directory. \n", + "# All changes under this directory will be kept even after reset. \n", + "# Please clean unnecessary files in time to speed up environment loading. \n", + "!ls /home/aistudio/work" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T08:34:26.119091Z", + "iopub.status.busy": "2022-05-04T08:34:26.118786Z", + "iopub.status.idle": "2022-05-04T08:34:28.698866Z", + "shell.execute_reply": "2022-05-04T08:34:28.698104Z", + "shell.execute_reply.started": "2022-05-04T08:34:26.119062Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple\n", + "Collecting beautifulsoup4\n", + " Downloading https://pypi.tuna.tsinghua.edu.cn/packages/9c/d8/909c4089dbe4ade9f9705f143c9f13f065049a9d5e7d34c828aefdd0a97c/beautifulsoup4-4.11.1-py3-none-any.whl (128 kB)\n", + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m128.2/128.2 KB\u001b[0m \u001b[31m1.7 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0ma \u001b[36m0:00:01\u001b[0m\n", + "\u001b[?25hCollecting soupsieve>1.2\n", + " Downloading https://pypi.tuna.tsinghua.edu.cn/packages/16/e3/4ad79882b92617e3a4a0df1960d6bce08edfb637737ac5c3f3ba29022e25/soupsieve-2.3.2.post1-py3-none-any.whl (37 kB)\n", + "Installing collected packages: soupsieve, beautifulsoup4\n", + "Successfully installed beautifulsoup4-4.11.1 soupsieve-2.3.2.post1\n" + ] + } + ], + "source": [ + "# 如果需要进行持久化安装, 需要使用持久化路径, 如下方代码示例:\n", + "# If a persistence installation is required, \n", + "# you need to use the persistence path as the following: \n", + "!mkdir /home/aistudio/external-libraries\n", + "!pip install beautifulsoup4 -t /home/aistudio/external-libraries" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T08:34:29.817313Z", + "iopub.status.busy": "2022-05-04T08:34:29.817070Z", + "iopub.status.idle": "2022-05-04T08:34:29.820288Z", + "shell.execute_reply": "2022-05-04T08:34:29.819897Z", + "shell.execute_reply.started": "2022-05-04T08:34:29.817291Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "# 同时添加如下代码, 这样每次环境(kernel)启动的时候只要运行下方代码即可: \n", + "# Also add the following code, \n", + "# so that every time the environment (kernel) starts, \n", + "# just run the following code: \n", + "import sys \n", + "sys.path.append('/home/aistudio/external-libraries')" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:36:51.914676Z", + "iopub.status.busy": "2022-05-04T08:36:51.914263Z", + "iopub.status.idle": "2022-05-04T08:36:53.054344Z", + "shell.execute_reply": "2022-05-04T08:36:53.053446Z", + "shell.execute_reply.started": "2022-05-04T08:36:51.914643Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "import paddle\r\n", + "from paddle.vision.transforms import Compose, Normalize\r\n", + "import math" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 数据集的构建" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 加载mnist数据集 " + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:38:11.644012Z", + "iopub.status.busy": "2022-05-04T08:38:11.643694Z", + "iopub.status.idle": "2022-05-04T08:38:17.776802Z", + "shell.execute_reply": "2022-05-04T08:38:17.776025Z", + "shell.execute_reply.started": "2022-05-04T08:38:11.643982Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "item 162/2421 [=>............................] - ETA: 1s - 752us/item" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Cache file /home/aistudio/.cache/paddle/dataset/mnist/train-images-idx3-ubyte.gz not found, downloading https://dataset.bj.bcebos.com/mnist/train-images-idx3-ubyte.gz \n", + "Begin to download\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "item 8/8 [============================>.] - ETA: 0s - 2ms/item" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "\n", + "Download finished\n", + "Cache file /home/aistudio/.cache/paddle/dataset/mnist/train-labels-idx1-ubyte.gz not found, downloading https://dataset.bj.bcebos.com/mnist/train-labels-idx1-ubyte.gz \n", + "Begin to download\n", + "\n", + "Download finished\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "item 207/403 [==============>...............] - ETA: 0s - 707us/item" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Cache file /home/aistudio/.cache/paddle/dataset/mnist/t10k-images-idx3-ubyte.gz not found, downloading https://dataset.bj.bcebos.com/mnist/t10k-images-idx3-ubyte.gz \n", + "Begin to download\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "item 2/2 [===========================>..] - ETA: 0s - 1ms/item " + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "\n", + "Download finished\n", + "Cache file /home/aistudio/.cache/paddle/dataset/mnist/t10k-labels-idx1-ubyte.gz not found, downloading https://dataset.bj.bcebos.com/mnist/t10k-labels-idx1-ubyte.gz \n", + "Begin to download\n", + "\n", + "Download finished\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "训练集样本量: 60000，验证集样本量: 10000\n" + ] + } + ], + "source": [ + "batch_size=256\r\n", + "\r\n", + "\r\n", + "# 使用transform对数据集做归一化\r\n", + "transform = Compose([Normalize(mean=[0],\r\n", + " std=[255],\r\n", + " data_format='CHW')])\r\n", + "\r\n", + "# 训练数据集\r\n", + "train_datasets = paddle.vision.datasets.MNIST(mode='train',transform=transform)\r\n", + "\r\n", + "# 测试数据集\r\n", + "test_datasets = paddle.vision.datasets.MNIST(mode='test',transform=transform)\r\n", + "\r\n", + "print('训练集样本量: {}，验证集样本量: {}'.format(len(train_datasets), len(test_datasets)))\r\n" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:39:57.363747Z", + "iopub.status.busy": "2022-05-04T08:39:57.363340Z", + "iopub.status.idle": "2022-05-04T08:39:57.372499Z", + "shell.execute_reply": "2022-05-04T08:39:57.371974Z", + "shell.execute_reply.started": "2022-05-04T08:39:57.363714Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "data": { + "text/plain": [ + "(1, 28, 28)" + ] + }, + "execution_count": 8, + "metadata": {}, + "output_type": "execute_result" + } + ], + "source": [ + "train_datasets[0][0].shape" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 图像进行加噪" + ] + }, + { + "cell_type": "code", + "execution_count": 11, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:42:45.123608Z", + "iopub.status.busy": "2022-05-04T08:42:45.123321Z", + "iopub.status.idle": "2022-05-04T08:42:51.167900Z", + "shell.execute_reply": "2022-05-04T08:42:51.167346Z", + "shell.execute_reply.started": "2022-05-04T08:42:45.123583Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "训练数据的长度是60000 测试数据的长度是10000\n", + "训练数据的shape是(1, 28, 28) 测试数据的shape是(1, 28, 28)\n", + "添加噪声后数据长度是60000 测试数据的长度是10000\n", + "添加噪声后数据的shape是(1, 28, 28) 测试数据的shape是(1, 28, 28)\n" + ] + } + ], + "source": [ + "import numpy as np\r\n", + "#1、将数据格式进行转换\r\n", + "x_train = [i[0] for i in train_datasets]\r\n", + "x_test = [i[0] for i in test_datasets]\r\n", + "\r\n", + "x_train = np.reshape(x_train,(len(x_train),1,28,28))\r\n", + "x_test = np.reshape(x_test,(len(x_test),1,28,28))\r\n", + "print('训练数据的长度是{} 测试数据的长度是{}'.format(len(x_train),len(x_test)))\r\n", + "print('训练数据的shape是{} 测试数据的shape是{}'.format(x_train[0].shape,x_test[0].shape))\r\n", + "\r\n", + "\r\n", + "#2、添加随机白噪声（均值是0，方差为1）\r\n", + "noise_factor = 0.5\r\n", + "x_train_noise = x_train+noise_factor*np.random.normal(loc=0.0,scale=1.0,size=x_train.shape)\r\n", + "x_test_noise = x_test+noise_factor*np.random.normal(loc=0.0,scale=1.0,size=x_test.shape)\r\n", + "#限制加完噪声的数值取值范围\r\n", + "x_train_noise = np.clip(x_train_noise,0,1) \r\n", + "x_test_noise = np.clip(x_test_noise,0,1)\r\n", + "\r\n", + "print('添加噪声后数据长度是{} 测试数据的长度是{}'.format(len(x_train_noise),len(x_test_noise)))\r\n", + "print('添加噪声后数据的shape是{} 测试数据的shape是{}'.format(x_train_noise[0].shape,x_test_noise[0].shape))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 构建dataset，输入为噪声图片，标签为原图" + ] + }, + { + "cell_type": "code", + "execution_count": 12, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:44:49.216924Z", + "iopub.status.busy": "2022-05-04T08:44:49.216628Z", + "iopub.status.idle": "2022-05-04T08:44:49.223109Z", + "shell.execute_reply": "2022-05-04T08:44:49.222703Z", + "shell.execute_reply.started": "2022-05-04T08:44:49.216899Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "import paddle\r\n", + "import json\r\n", + "import gzip\r\n", + "import numpy as np\r\n", + "\r\n", + "# 创建一个类MnistDataset，继承paddle.io.Dataset 这个类\r\n", + "class MnistDataset(paddle.io.Dataset):\r\n", + " def __init__(self, mode):\r\n", + " super(MnistDataset, self).__init__()\r\n", + " # 读取到的数据区分训练集，验证集，测试集\r\n", + " if mode == 'train':\r\n", + " # 获得训练数据集\r\n", + " imgs, labels = x_train_noise, x_train\r\n", + " elif mode == 'test':\r\n", + " # 获得测试数据集\r\n", + " imgs, labels = x_test_noise,x_test\r\n", + " else:\r\n", + " raise Exception(\"mode can only be one of ['train', 'valid', 'eval']\")\r\n", + "\r\n", + " # 校验数据\r\n", + " imgs_length = len(imgs)\r\n", + " assert len(imgs) == len(labels), \\\r\n", + " \"length of train_imgs({}) should be the same as train_labels({})\".format(len(imgs), len(labels))\r\n", + "\r\n", + " self.imgs = imgs \r\n", + " self.labels = labels\r\n", + "\r\n", + " def __getitem__(self, idx):\r\n", + " img = np.array(self.imgs[idx]).astype('float32') \r\n", + " label = np.array(self.labels[idx]).astype('float32')\r\n", + "\r\n", + " return img, label\r\n", + "\r\n", + " def __len__(self):\r\n", + " return len(self.imgs)\r\n" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 数据可视化" + ] + }, + { + "cell_type": "code", + "execution_count": 15, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:47:58.812349Z", + "iopub.status.busy": "2022-05-04T08:47:58.812056Z", + "iopub.status.idle": "2022-05-04T08:47:59.230391Z", + "shell.execute_reply": "2022-05-04T08:47:59.229971Z", + "shell.execute_reply.started": "2022-05-04T08:47:58.812324Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "import warnings \r\n", + "warnings.filterwarnings('ignore')\r\n", + "\r\n", + "import matplotlib.pyplot as plt\r\n", + "%matplotlib inline\r\n", + "\r\n", + "n = 10 \r\n", + "\r\n", + "plt.figure(figsize=(20,2)) \r\n", + "for i in range(n): \r\n", + " ax = plt.subplot(2,n,i+1) \r\n", + " plt.imshow(x_train[i][0].reshape(28,28)) \r\n", + " plt.gray() \r\n", + " ax.get_xaxis().set_visible(False) \r\n", + " ax.get_yaxis().set_visible(False)\r\n", + "\r\n", + "for i in range(n): \r\n", + " ax = plt.subplot(2,n,10+i+1) \r\n", + " plt.imshow(x_train_noise[i][0].reshape(28,28)) \r\n", + " plt.gray() \r\n", + " ax.get_xaxis().set_visible(False) \r\n", + " ax.get_yaxis().set_visible(False)\r\n", + "\r\n", + "plt.show()\r\n", + "\r\n" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 构建dataloader" + ] + }, + { + "cell_type": "code", + "execution_count": 18, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:55:46.773880Z", + "iopub.status.busy": "2022-05-04T08:55:46.773599Z", + "iopub.status.idle": "2022-05-04T08:55:46.776941Z", + "shell.execute_reply": "2022-05-04T08:55:46.776545Z", + "shell.execute_reply.started": "2022-05-04T08:55:46.773855Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "train_dataset = MnistDataset(mode='train')\r\n", + "train_loader = paddle.io.DataLoader(train_dataset, batch_size=batch_size, shuffle=True)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 模型的构建：构建自编码器，此处采用动态图" + ] + }, + { + "cell_type": "code", + "execution_count": 16, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:52:43.914724Z", + "iopub.status.busy": "2022-05-04T08:52:43.914422Z", + "iopub.status.idle": "2022-05-04T08:52:43.923754Z", + "shell.execute_reply": "2022-05-04T08:52:43.923336Z", + "shell.execute_reply.started": "2022-05-04T08:52:43.914699Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "import paddle\r\n", + "import paddle.nn.functional as F\r\n", + "\r\n", + "class autoencoder(paddle.nn.Layer):\r\n", + " def __init__(self):\r\n", + " super(autoencoder, self).__init__()\r\n", + " # encoder部分\r\n", + " # （1， 28， 28） ---> （32， 28， 28）\r\n", + " self.conv1 = paddle.nn.Conv2D(in_channels=1, out_channels=32, kernel_size=(3,3), stride=1, padding=1)\r\n", + " # （32， 28， 28）--- > (32, 14, 14)\r\n", + " self.max_pool1 = paddle.nn.MaxPool2D(kernel_size=2, stride=2)\r\n", + " # （32， 14, 14)--- > (64, 14, 14)\r\n", + " self.conv2 = paddle.nn.Conv2D(in_channels=32, out_channels=64, kernel_size=(3,3), stride=1,padding=1)\r\n", + " # (64, 14, 14)--- > (64, 7, 7)\r\n", + " self.max_pool2 = paddle.nn.MaxPool2D(kernel_size=2, stride=2)\r\n", + " \r\n", + " # decoder部分\r\n", + " # (64, 7, 7) ----> \r\n", + " self.conv3 = paddle.nn.Conv2D(in_channels=64, out_channels=64, kernel_size=(3,3), stride=1, padding=1)\r\n", + " self.up_pool3 = paddle.nn.Upsample(size=[14,14]) #不改变channel，只是将其hw翻倍\r\n", + " self.conv4 = paddle.nn.Conv2D(in_channels=64, out_channels=32, kernel_size=(3,3), stride=1,padding=1)\r\n", + " self.up_pool4 = paddle.nn.Upsample(size=[28,28])\r\n", + " self.conv5 = paddle.nn.Conv2D(in_channels=32, out_channels=1, kernel_size=(3,3), stride=1, padding=1)\r\n", + " \r\n", + " \r\n", + " def forward(self, x):\r\n", + " # encoder部分\r\n", + " x = self.conv1(x)\r\n", + " x = F.relu(x)\r\n", + " x = self.max_pool1(x)\r\n", + " x = self.conv2(x)\r\n", + " x = F.relu(x)\r\n", + " x = self.max_pool2(x)\r\n", + " #decoder部分\r\n", + " x = self.conv3(x)\r\n", + " x = F.relu(x)\r\n", + " x = self.up_pool3(x)\r\n", + " x = self.conv4(x)\r\n", + " x = F.relu(x)\r\n", + " x = self.up_pool4(x)\r\n", + " x = self.conv5(x)\r\n", + " x = F.sigmoid(x)\r\n", + " return x" + ] + }, + { + "cell_type": "code", + "execution_count": 17, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:52:56.554027Z", + "iopub.status.busy": "2022-05-04T08:52:56.553741Z", + "iopub.status.idle": "2022-05-04T08:53:00.767506Z", + "shell.execute_reply": "2022-05-04T08:53:00.766954Z", + "shell.execute_reply.started": "2022-05-04T08:52:56.554003Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "W0504 16:52:56.556185 163 device_context.cc:447] Please NOTE: device: 0, GPU Compute Capability: 8.0, Driver API Version: 11.2, Runtime API Version: 11.2\n", + "W0504 16:52:56.559684 163 device_context.cc:465] device: 0, cuDNN Version: 8.2.\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "---------------------------------------------------------------------------\n", + " Layer (type) Input Shape Output Shape Param # \n", + "===========================================================================\n", + " Conv2D-1 [[1, 1, 28, 28]] [1, 32, 28, 28] 320 \n", + " MaxPool2D-1 [[1, 32, 28, 28]] [1, 32, 14, 14] 0 \n", + " Conv2D-2 [[1, 32, 14, 14]] [1, 64, 14, 14] 18,496 \n", + " MaxPool2D-2 [[1, 64, 14, 14]] [1, 64, 7, 7] 0 \n", + " Conv2D-3 [[1, 64, 7, 7]] [1, 64, 7, 7] 36,928 \n", + " Upsample-1 [[1, 64, 7, 7]] [1, 64, 14, 14] 0 \n", + " Conv2D-4 [[1, 64, 14, 14]] [1, 32, 14, 14] 18,464 \n", + " Upsample-2 [[1, 32, 14, 14]] [1, 32, 28, 28] 0 \n", + " Conv2D-5 [[1, 32, 28, 28]] [1, 1, 28, 28] 289 \n", + "===========================================================================\n", + "Total params: 74,497\n", + "Trainable params: 74,497\n", + "Non-trainable params: 0\n", + "---------------------------------------------------------------------------\n", + "Input size (MB): 0.00\n", + "Forward/backward pass size (MB): 0.72\n", + "Params size (MB): 0.28\n", + "Estimated Total Size (MB): 1.01\n", + "---------------------------------------------------------------------------\n", + "\n", + "{'total_params': 74497, 'trainable_params': 74497}\n" + ] + } + ], + "source": [ + "model = autoencoder()\r\n", + "#查看模型结构\r\n", + "params_info = paddle.summary(model, (1, 1, 28, 28))\r\n", + "print(params_info)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 构建优化器" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T08:56:37.561502Z", + "iopub.status.busy": "2022-05-04T08:56:37.561204Z", + "iopub.status.idle": "2022-05-04T08:56:37.564706Z", + "shell.execute_reply": "2022-05-04T08:56:37.564251Z", + "shell.execute_reply.started": "2022-05-04T08:56:37.561476Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "optim = paddle.optimizer.Adam(learning_rate=0.001, parameters=model.parameters())" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 训练" + ] + }, + { + "cell_type": "code", + "execution_count": 22, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:01:50.695787Z", + "iopub.status.busy": "2022-05-04T09:01:50.695418Z", + "iopub.status.idle": "2022-05-04T09:01:50.702122Z", + "shell.execute_reply": "2022-05-04T09:01:50.701503Z", + "shell.execute_reply.started": "2022-05-04T09:01:50.695757Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "def train(model, optim, epochs): \r\n", + " model.train()\r\n", + " for epoch in range(epochs):\r\n", + " for batch_id, data in enumerate(train_loader,1):\r\n", + " x_data = data[0]\r\n", + " y_data = data[1]\r\n", + " predicts = model(x_data)\r\n", + " # y_data为0， 1可以采用二分类交叉熵损失按数\r\n", + " loss = F.binary_cross_entropy(predicts, y_data) \r\n", + " # 计算损失\r\n", + " loss.backward()\r\n", + " optim.step()\r\n", + " optim.clear_grad()\r\n", + "\r\n", + " if batch_id % 50 == 0:\r\n", + " opt_lr = optim.get_lr()\r\n", + " print(\"epoch: {}, batch_id: {}, loss is: {}, opt_lr: {}\".format(epoch, batch_id, loss.numpy(), opt_lr))\r\n" + ] + }, + { + "cell_type": "code", + "execution_count": 23, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:01:53.560327Z", + "iopub.status.busy": "2022-05-04T09:01:53.559994Z", + "iopub.status.idle": "2022-05-04T09:02:32.744321Z", + "shell.execute_reply": "2022-05-04T09:02:32.743708Z", + "shell.execute_reply.started": "2022-05-04T09:01:53.560297Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "epoch: 0, batch_id: 50, loss is: [0.1508055], opt_lr: 0.001\n", + "epoch: 0, batch_id: 100, loss is: [0.13356395], opt_lr: 0.001\n", + "epoch: 0, batch_id: 150, loss is: [0.12579954], opt_lr: 0.001\n", + "epoch: 0, batch_id: 200, loss is: [0.11945431], opt_lr: 0.001\n", + "epoch: 1, batch_id: 50, loss is: [0.11757692], opt_lr: 0.001\n", + "epoch: 1, batch_id: 100, loss is: [0.11732341], opt_lr: 0.001\n", + "epoch: 1, batch_id: 150, loss is: [0.11405985], opt_lr: 0.001\n", + "epoch: 1, batch_id: 200, loss is: [0.11205357], opt_lr: 0.001\n", + "epoch: 2, batch_id: 50, loss is: [0.10728515], opt_lr: 0.001\n", + "epoch: 2, batch_id: 100, loss is: [0.10778838], opt_lr: 0.001\n", + "epoch: 2, batch_id: 150, loss is: [0.10553826], opt_lr: 0.001\n", + "epoch: 2, batch_id: 200, loss is: [0.10809629], opt_lr: 0.001\n", + "epoch: 3, batch_id: 50, loss is: [0.10743915], opt_lr: 0.001\n", + "epoch: 3, batch_id: 100, loss is: [0.10226806], opt_lr: 0.001\n", + "epoch: 3, batch_id: 150, loss is: [0.1021505], opt_lr: 0.001\n", + "epoch: 3, batch_id: 200, loss is: [0.10333781], opt_lr: 0.001\n", + "epoch: 4, batch_id: 50, loss is: [0.10274198], opt_lr: 0.001\n", + "epoch: 4, batch_id: 100, loss is: [0.10416618], opt_lr: 0.001\n", + "epoch: 4, batch_id: 150, loss is: [0.10226612], opt_lr: 0.001\n", + "epoch: 4, batch_id: 200, loss is: [0.10335997], opt_lr: 0.001\n", + "epoch: 5, batch_id: 50, loss is: [0.102478], opt_lr: 0.001\n", + "epoch: 5, batch_id: 100, loss is: [0.09874876], opt_lr: 0.001\n", + "epoch: 5, batch_id: 150, loss is: [0.10439914], opt_lr: 0.001\n", + "epoch: 5, batch_id: 200, loss is: [0.10197291], opt_lr: 0.001\n", + "epoch: 6, batch_id: 50, loss is: [0.09828829], opt_lr: 0.001\n", + "epoch: 6, batch_id: 100, loss is: [0.09923309], opt_lr: 0.001\n", + "epoch: 6, batch_id: 150, loss is: [0.09882569], opt_lr: 0.001\n", + "epoch: 6, batch_id: 200, loss is: [0.10150312], opt_lr: 0.001\n", + "epoch: 7, batch_id: 50, loss is: [0.09955291], opt_lr: 0.001\n", + "epoch: 7, batch_id: 100, loss is: [0.09907445], opt_lr: 0.001\n", + "epoch: 7, batch_id: 150, loss is: [0.09910685], opt_lr: 0.001\n", + "epoch: 7, batch_id: 200, loss is: [0.0992258], opt_lr: 0.001\n", + "epoch: 8, batch_id: 50, loss is: [0.09876722], opt_lr: 0.001\n", + "epoch: 8, batch_id: 100, loss is: [0.09823684], opt_lr: 0.001\n", + "epoch: 8, batch_id: 150, loss is: [0.0986691], opt_lr: 0.001\n", + "epoch: 8, batch_id: 200, loss is: [0.09885979], opt_lr: 0.001\n", + "epoch: 9, batch_id: 50, loss is: [0.10165662], opt_lr: 0.001\n", + "epoch: 9, batch_id: 100, loss is: [0.09757827], opt_lr: 0.001\n", + "epoch: 9, batch_id: 150, loss is: [0.09765131], opt_lr: 0.001\n", + "epoch: 9, batch_id: 200, loss is: [0.09868667], opt_lr: 0.001\n", + "epoch: 10, batch_id: 50, loss is: [0.09821317], opt_lr: 0.001\n", + "epoch: 10, batch_id: 100, loss is: [0.09776197], opt_lr: 0.001\n", + "epoch: 10, batch_id: 150, loss is: [0.0996934], opt_lr: 0.001\n", + "epoch: 10, batch_id: 200, loss is: [0.09697092], opt_lr: 0.001\n", + "epoch: 11, batch_id: 50, loss is: [0.09950332], opt_lr: 0.001\n", + "epoch: 11, batch_id: 100, loss is: [0.0986447], opt_lr: 0.001\n", + "epoch: 11, batch_id: 150, loss is: [0.09873365], opt_lr: 0.001\n", + "epoch: 11, batch_id: 200, loss is: [0.0978436], opt_lr: 0.001\n", + "epoch: 12, batch_id: 50, loss is: [0.09742838], opt_lr: 0.001\n", + "epoch: 12, batch_id: 100, loss is: [0.09709799], opt_lr: 0.001\n", + "epoch: 12, batch_id: 150, loss is: [0.09535895], opt_lr: 0.001\n", + "epoch: 12, batch_id: 200, loss is: [0.09824618], opt_lr: 0.001\n", + "epoch: 13, batch_id: 50, loss is: [0.09578139], opt_lr: 0.001\n", + "epoch: 13, batch_id: 100, loss is: [0.09841461], opt_lr: 0.001\n", + "epoch: 13, batch_id: 150, loss is: [0.09561975], opt_lr: 0.001\n", + "epoch: 13, batch_id: 200, loss is: [0.09720887], opt_lr: 0.001\n", + "epoch: 14, batch_id: 50, loss is: [0.09563191], opt_lr: 0.001\n", + "epoch: 14, batch_id: 100, loss is: [0.09563148], opt_lr: 0.001\n", + "epoch: 14, batch_id: 150, loss is: [0.09583557], opt_lr: 0.001\n", + "epoch: 14, batch_id: 200, loss is: [0.09642026], opt_lr: 0.001\n", + "epoch: 15, batch_id: 50, loss is: [0.09888563], opt_lr: 0.001\n", + "epoch: 15, batch_id: 100, loss is: [0.09592318], opt_lr: 0.001\n", + "epoch: 15, batch_id: 150, loss is: [0.09575748], opt_lr: 0.001\n", + "epoch: 15, batch_id: 200, loss is: [0.09646571], opt_lr: 0.001\n", + "epoch: 16, batch_id: 50, loss is: [0.09500609], opt_lr: 0.001\n", + "epoch: 16, batch_id: 100, loss is: [0.09510812], opt_lr: 0.001\n", + "epoch: 16, batch_id: 150, loss is: [0.09582922], opt_lr: 0.001\n", + "epoch: 16, batch_id: 200, loss is: [0.09704416], opt_lr: 0.001\n", + "epoch: 17, batch_id: 50, loss is: [0.0948493], opt_lr: 0.001\n", + "epoch: 17, batch_id: 100, loss is: [0.09397411], opt_lr: 0.001\n", + "epoch: 17, batch_id: 150, loss is: [0.09457557], opt_lr: 0.001\n", + "epoch: 17, batch_id: 200, loss is: [0.09863666], opt_lr: 0.001\n", + "epoch: 18, batch_id: 50, loss is: [0.09370035], opt_lr: 0.001\n", + "epoch: 18, batch_id: 100, loss is: [0.09644823], opt_lr: 0.001\n", + "epoch: 18, batch_id: 150, loss is: [0.09582626], opt_lr: 0.001\n", + "epoch: 18, batch_id: 200, loss is: [0.09611063], opt_lr: 0.001\n", + "epoch: 19, batch_id: 50, loss is: [0.09449808], opt_lr: 0.001\n", + "epoch: 19, batch_id: 100, loss is: [0.09640694], opt_lr: 0.001\n", + "epoch: 19, batch_id: 150, loss is: [0.09285444], opt_lr: 0.001\n", + "epoch: 19, batch_id: 200, loss is: [0.09234895], opt_lr: 0.001\n" + ] + } + ], + "source": [ + "epochs = 20\r\n", + "train(model, optim, epochs)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 预测" + ] + }, + { + "cell_type": "code", + "execution_count": 24, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:05:36.561586Z", + "iopub.status.busy": "2022-05-04T09:05:36.558161Z", + "iopub.status.idle": "2022-05-04T09:05:36.565264Z", + "shell.execute_reply": "2022-05-04T09:05:36.564800Z", + "shell.execute_reply.started": "2022-05-04T09:05:36.561502Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "test_dataset = MnistDataset(mode='test')\r\n", + "# 使用paddle.io.DataLoader 定义DataLoader对象用于加载Python生成器产生的数据，\r\n", + "test_loader = paddle.io.DataLoader(test_dataset, batch_size=16, shuffle=False)" + ] + }, + { + "cell_type": "code", + "execution_count": 51, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:30:19.712354Z", + "iopub.status.busy": "2022-05-04T09:30:19.712068Z", + "iopub.status.idle": "2022-05-04T09:30:19.721057Z", + "shell.execute_reply": "2022-05-04T09:30:19.720638Z", + "shell.execute_reply.started": "2022-05-04T09:30:19.712329Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + " for batch_id, data in enumerate(test_loader):\r\n", + " x_data = data[0]\r\n", + " y_data = data[1]\r\n", + " predicts = model(x_data)\r\n", + " break" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 可视化原图与网络去噪后图像" + ] + }, + { + "cell_type": "code", + "execution_count": 52, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:30:21.980561Z", + "iopub.status.busy": "2022-05-04T09:30:21.980214Z", + "iopub.status.idle": "2022-05-04T09:30:22.402297Z", + "shell.execute_reply": "2022-05-04T09:30:22.401837Z", + "shell.execute_reply.started": "2022-05-04T09:30:21.980530Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "import warnings \r\n", + "warnings.filterwarnings('ignore')\r\n", + "\r\n", + "import matplotlib.pyplot as plt\r\n", + "%matplotlib inline\r\n", + "\r\n", + "n = 10 \r\n", + "\r\n", + "x_data = x_data.numpy()\r\n", + "predicts = predicts.numpy()\r\n", + "\r\n", + "plt.figure(figsize=(20,2)) \r\n", + "for i in range(n): \r\n", + " ax = plt.subplot(2,n,i+1) \r\n", + " plt.imshow(x_data[i][0].reshape(28,28)) \r\n", + " plt.gray() \r\n", + " ax.get_xaxis().set_visible(False) \r\n", + " ax.get_yaxis().set_visible(False)\r\n", + "\r\n", + "for i in range(n): \r\n", + " ax = plt.subplot(2,n,10+i+1) \r\n", + " plt.imshow(predicts[i].reshape(28,28)) \r\n", + " plt.gray() \r\n", + " ax.get_xaxis().set_visible(False) \r\n", + " ax.get_yaxis().set_visible(False)\r\n", + "\r\n", + "plt.show()\r\n", + "\r\n" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# 使用 paddle.jit.to_static 实现动转静" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 改写组网代码" + ] + }, + { + "cell_type": "code", + "execution_count": 42, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:25:21.924312Z", + "iopub.status.busy": "2022-05-04T09:25:21.924023Z", + "iopub.status.idle": "2022-05-04T09:25:21.933480Z", + "shell.execute_reply": "2022-05-04T09:25:21.933029Z", + "shell.execute_reply.started": "2022-05-04T09:25:21.924287Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "class autoencoder2(paddle.nn.Layer):\r\n", + " def __init__(self):\r\n", + " super(autoencoder2, self).__init__()\r\n", + " # encoder部分\r\n", + " # （1， 28， 28） ---> （32， 28， 28）\r\n", + " self.conv1 = paddle.nn.Conv2D(in_channels=1, out_channels=32, kernel_size=(3,3), stride=1, padding=1)\r\n", + " # （32， 28， 28）--- > (32, 14, 14)\r\n", + " self.max_pool1 = paddle.nn.MaxPool2D(kernel_size=2, stride=2)\r\n", + " # （32， 14, 14)--- > (64, 14, 14)\r\n", + " self.conv2 = paddle.nn.Conv2D(in_channels=32, out_channels=64, kernel_size=(3,3), stride=1,padding=1)\r\n", + " # (64, 14, 14)--- > (64, 7, 7)\r\n", + " self.max_pool2 = paddle.nn.MaxPool2D(kernel_size=2, stride=2)\r\n", + " \r\n", + " # decoder部分\r\n", + " # (64, 7, 7) ----> \r\n", + " self.conv3 = paddle.nn.Conv2D(in_channels=64, out_channels=64, kernel_size=(3,3), stride=1, padding=1)\r\n", + " self.up_pool3 = paddle.nn.Upsample(size=[14,14]) #不改变channel，只是将其hw翻倍\r\n", + " self.conv4 = paddle.nn.Conv2D(in_channels=64, out_channels=32, kernel_size=(3,3), stride=1,padding=1)\r\n", + " self.up_pool4 = paddle.nn.Upsample(size=[28,28])\r\n", + " self.conv5 = paddle.nn.Conv2D(in_channels=32, out_channels=1, kernel_size=(3,3), stride=1, padding=1)\r\n", + " \r\n", + " \r\n", + " # 在forward 前添加 paddle.jit.to_static 装饰器\r\n", + " @paddle.jit.to_static()\r\n", + " def forward(self, x):\r\n", + " # encoder部分\r\n", + " x = self.conv1(x)\r\n", + " x = F.relu(x)\r\n", + " x = self.max_pool1(x)\r\n", + " x = self.conv2(x)\r\n", + " x = F.relu(x)\r\n", + " x = self.max_pool2(x)\r\n", + " #decoder部分\r\n", + " x = self.conv3(x)\r\n", + " x = F.relu(x)\r\n", + " x = self.up_pool3(x)\r\n", + " x = self.conv4(x)\r\n", + " x = F.relu(x)\r\n", + " x = self.up_pool4(x)\r\n", + " x = self.conv5(x)\r\n", + " x = F.sigmoid(x)\r\n", + " return x" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 模型训练" + ] + }, + { + "cell_type": "code", + "execution_count": 43, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:27:07.124631Z", + "iopub.status.busy": "2022-05-04T09:27:07.124322Z", + "iopub.status.idle": "2022-05-04T09:27:40.046876Z", + "shell.execute_reply": "2022-05-04T09:27:40.046183Z", + "shell.execute_reply.started": "2022-05-04T09:27:07.124602Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "epoch: 0, batch_id: 50, loss is: [0.169617], opt_lr: 0.001\n", + "epoch: 0, batch_id: 100, loss is: [0.14395738], opt_lr: 0.001\n", + "epoch: 0, batch_id: 150, loss is: [0.1290719], opt_lr: 0.001\n", + "epoch: 0, batch_id: 200, loss is: [0.12327737], opt_lr: 0.001\n", + "epoch: 1, batch_id: 50, loss is: [0.11703932], opt_lr: 0.001\n", + "epoch: 1, batch_id: 100, loss is: [0.11644541], opt_lr: 0.001\n", + "epoch: 1, batch_id: 150, loss is: [0.11610189], opt_lr: 0.001\n", + "epoch: 1, batch_id: 200, loss is: [0.11325552], opt_lr: 0.001\n", + "epoch: 2, batch_id: 50, loss is: [0.11371018], opt_lr: 0.001\n", + "epoch: 2, batch_id: 100, loss is: [0.10730042], opt_lr: 0.001\n", + "epoch: 2, batch_id: 150, loss is: [0.10588661], opt_lr: 0.001\n", + "epoch: 2, batch_id: 200, loss is: [0.10843735], opt_lr: 0.001\n", + "epoch: 3, batch_id: 50, loss is: [0.10706727], opt_lr: 0.001\n", + "epoch: 3, batch_id: 100, loss is: [0.10239325], opt_lr: 0.001\n", + "epoch: 3, batch_id: 150, loss is: [0.10862175], opt_lr: 0.001\n", + "epoch: 3, batch_id: 200, loss is: [0.10543257], opt_lr: 0.001\n", + "epoch: 4, batch_id: 50, loss is: [0.10374369], opt_lr: 0.001\n", + "epoch: 4, batch_id: 100, loss is: [0.10380919], opt_lr: 0.001\n", + "epoch: 4, batch_id: 150, loss is: [0.1029813], opt_lr: 0.001\n", + "epoch: 4, batch_id: 200, loss is: [0.10477626], opt_lr: 0.001\n", + "epoch: 5, batch_id: 50, loss is: [0.10353363], opt_lr: 0.001\n", + "epoch: 5, batch_id: 100, loss is: [0.10166034], opt_lr: 0.001\n", + "epoch: 5, batch_id: 150, loss is: [0.10359956], opt_lr: 0.001\n", + "epoch: 5, batch_id: 200, loss is: [0.10143657], opt_lr: 0.001\n", + "epoch: 6, batch_id: 50, loss is: [0.10366039], opt_lr: 0.001\n", + "epoch: 6, batch_id: 100, loss is: [0.10135973], opt_lr: 0.001\n", + "epoch: 6, batch_id: 150, loss is: [0.1002941], opt_lr: 0.001\n", + "epoch: 6, batch_id: 200, loss is: [0.10337035], opt_lr: 0.001\n", + "epoch: 7, batch_id: 50, loss is: [0.10042988], opt_lr: 0.001\n", + "epoch: 7, batch_id: 100, loss is: [0.1004205], opt_lr: 0.001\n", + "epoch: 7, batch_id: 150, loss is: [0.09875569], opt_lr: 0.001\n", + "epoch: 7, batch_id: 200, loss is: [0.10202026], opt_lr: 0.001\n", + "epoch: 8, batch_id: 50, loss is: [0.10035688], opt_lr: 0.001\n", + "epoch: 8, batch_id: 100, loss is: [0.0982463], opt_lr: 0.001\n", + "epoch: 8, batch_id: 150, loss is: [0.10255557], opt_lr: 0.001\n", + "epoch: 8, batch_id: 200, loss is: [0.10166503], opt_lr: 0.001\n", + "epoch: 9, batch_id: 50, loss is: [0.09931719], opt_lr: 0.001\n", + "epoch: 9, batch_id: 100, loss is: [0.09736487], opt_lr: 0.001\n", + "epoch: 9, batch_id: 150, loss is: [0.09799257], opt_lr: 0.001\n", + "epoch: 9, batch_id: 200, loss is: [0.09900901], opt_lr: 0.001\n", + "epoch: 10, batch_id: 50, loss is: [0.10127259], opt_lr: 0.001\n", + "epoch: 10, batch_id: 100, loss is: [0.10058917], opt_lr: 0.001\n", + "epoch: 10, batch_id: 150, loss is: [0.09551379], opt_lr: 0.001\n", + "epoch: 10, batch_id: 200, loss is: [0.09726325], opt_lr: 0.001\n", + "epoch: 11, batch_id: 50, loss is: [0.09861174], opt_lr: 0.001\n", + "epoch: 11, batch_id: 100, loss is: [0.09999968], opt_lr: 0.001\n", + "epoch: 11, batch_id: 150, loss is: [0.09981595], opt_lr: 0.001\n", + "epoch: 11, batch_id: 200, loss is: [0.09690353], opt_lr: 0.001\n", + "epoch: 12, batch_id: 50, loss is: [0.09662779], opt_lr: 0.001\n", + "epoch: 12, batch_id: 100, loss is: [0.09695677], opt_lr: 0.001\n", + "epoch: 12, batch_id: 150, loss is: [0.09773625], opt_lr: 0.001\n", + "epoch: 12, batch_id: 200, loss is: [0.09750137], opt_lr: 0.001\n", + "epoch: 13, batch_id: 50, loss is: [0.0978588], opt_lr: 0.001\n", + "epoch: 13, batch_id: 100, loss is: [0.09882198], opt_lr: 0.001\n", + "epoch: 13, batch_id: 150, loss is: [0.09647948], opt_lr: 0.001\n", + "epoch: 13, batch_id: 200, loss is: [0.0956523], opt_lr: 0.001\n", + "epoch: 14, batch_id: 50, loss is: [0.09674709], opt_lr: 0.001\n", + "epoch: 14, batch_id: 100, loss is: [0.09855855], opt_lr: 0.001\n", + "epoch: 14, batch_id: 150, loss is: [0.09678999], opt_lr: 0.001\n", + "epoch: 14, batch_id: 200, loss is: [0.09646296], opt_lr: 0.001\n", + "epoch: 15, batch_id: 50, loss is: [0.09774074], opt_lr: 0.001\n", + "epoch: 15, batch_id: 100, loss is: [0.0990114], opt_lr: 0.001\n", + "epoch: 15, batch_id: 150, loss is: [0.09859668], opt_lr: 0.001\n", + "epoch: 15, batch_id: 200, loss is: [0.09847274], opt_lr: 0.001\n", + "epoch: 16, batch_id: 50, loss is: [0.0930793], opt_lr: 0.001\n", + "epoch: 16, batch_id: 100, loss is: [0.09622699], opt_lr: 0.001\n", + "epoch: 16, batch_id: 150, loss is: [0.0949804], opt_lr: 0.001\n", + "epoch: 16, batch_id: 200, loss is: [0.09686655], opt_lr: 0.001\n", + "epoch: 17, batch_id: 50, loss is: [0.09775889], opt_lr: 0.001\n", + "epoch: 17, batch_id: 100, loss is: [0.09961951], opt_lr: 0.001\n", + "epoch: 17, batch_id: 150, loss is: [0.09541321], opt_lr: 0.001\n", + "epoch: 17, batch_id: 200, loss is: [0.09536967], opt_lr: 0.001\n", + "epoch: 18, batch_id: 50, loss is: [0.09764759], opt_lr: 0.001\n", + "epoch: 18, batch_id: 100, loss is: [0.09770085], opt_lr: 0.001\n", + "epoch: 18, batch_id: 150, loss is: [0.09733106], opt_lr: 0.001\n", + "epoch: 18, batch_id: 200, loss is: [0.0958437], opt_lr: 0.001\n", + "epoch: 19, batch_id: 50, loss is: [0.09469792], opt_lr: 0.001\n", + "epoch: 19, batch_id: 100, loss is: [0.09562233], opt_lr: 0.001\n", + "epoch: 19, batch_id: 150, loss is: [0.09384999], opt_lr: 0.001\n", + "epoch: 19, batch_id: 200, loss is: [0.0972498], opt_lr: 0.001\n" + ] + } + ], + "source": [ + "epochs = 20\r\n", + "model_2 = autoencoder2()\r\n", + "optim_2 = paddle.optimizer.Adam(learning_rate=0.001, parameters=model_2.parameters())\r\n", + "train(model_2, optim_2, epochs)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 使用 paddle.jit.save 保存动转静模型" + ] + }, + { + "cell_type": "code", + "execution_count": 44, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:28:16.745681Z", + "iopub.status.busy": "2022-05-04T09:28:16.745391Z", + "iopub.status.idle": "2022-05-04T09:28:16.772081Z", + "shell.execute_reply": "2022-05-04T09:28:16.771638Z", + "shell.execute_reply.started": "2022-05-04T09:28:16.745658Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Wed May 04 17:28:16 Dynamic-to-Static WARNING: Current function: forward(x), input_spec: None has more than one cached programs: 2, the last traced progam will be return by default.\n" + ] + } + ], + "source": [ + "paddle.jit.save(model_2, 'model')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 使用 paddle.jit.load 加载动转静模型" + ] + }, + { + "cell_type": "code", + "execution_count": 45, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:28:44.010140Z", + "iopub.status.busy": "2022-05-04T09:28:44.009848Z", + "iopub.status.idle": "2022-05-04T09:28:44.024315Z", + "shell.execute_reply": "2022-05-04T09:28:44.023881Z", + "shell.execute_reply.started": "2022-05-04T09:28:44.010115Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "model_2 = paddle.jit.load('model')" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 预测" + ] + }, + { + "cell_type": "code", + "execution_count": 49, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:30:03.149250Z", + "iopub.status.busy": "2022-05-04T09:30:03.148950Z", + "iopub.status.idle": "2022-05-04T09:30:03.158770Z", + "shell.execute_reply": "2022-05-04T09:30:03.158138Z", + "shell.execute_reply.started": "2022-05-04T09:30:03.149226Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + " for batch_id, data in enumerate(test_loader):\r\n", + " x_data = data[0]\r\n", + " y_data = data[1]\r\n", + " predicts = model_2(x_data)\r\n", + " break" + ] + }, + { + "cell_type": "code", + "execution_count": 50, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T09:30:05.299936Z", + "iopub.status.busy": "2022-05-04T09:30:05.299631Z", + "iopub.status.idle": "2022-05-04T09:30:05.921781Z", + "shell.execute_reply": "2022-05-04T09:30:05.921343Z", + "shell.execute_reply.started": "2022-05-04T09:30:05.299912Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "data": { + "image/png": "iVBORw0KGgoAAAANSUhEUgAABEQAAACCCAYAAABPex2ZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzsnXmAjvX6xq/ZzVhmMINmGvve2KmJLC20iKzZZS9LZKsUCkmOJVSWEkqWCGXJcX46QmSNE7LmoOxrsiTL8/vjOdf9fN/nfWaipOW5P/9g5p13Xu/3vb/bfd3XHWJZFhRFURRFURRFURRFUfxE6B/9AhRFURRFURRFURRFUW41eiGiKIqiKIqiKIqiKIrv0AsRRVEURVEURVEURVF8h16IKIqiKIqiKIqiKIriO/RCRFEURVEURVEURVEU36EXIoqiKIqiKIqiKIqi+A69EFEURVEURVEURVEUxXfohYiiKIqiKIqiKIqiKL5DL0QURVEURVEURVEURfEdeiGiKIqiKIqiKIqiKIrvCL+RB4eEhFjmv8uVK4c9e/YAAH744QcAQNasWQEAp0+fTvN5SpUqBQDYtm0brly5Yr+QcPul3HbbbQCA7777DgAQExODq1evAgCyZcsGAEhMTAQA7N+/HydOnAh47ri4OABAhgwZAABHjhxBwYIFAUBeq2VZIdf9n/6bwTE0xyl//vwAgNDQwPuxPXv2IDY2FoAzvvHx8QH/tixLxjBHjhwAgJw5cwIAvv32WwDAhQsX5Ofc45UzZ04cPXoUABAREQEAuHz5csBjoqOjkTdvXgDA9u3b+eUTlmUl3Nj//u+DOxYBIHv27ACAkydPXvfzlC1bFoA9nhwvUqRIEQDArl27AADFixfHvn37AADnz59P8zndnxk+z86dO5GQYA/Z8ePHAfg7FjNmzGjFxcXh2LFjACBxBACZMmUK+PPIkSPyPY4z4+Ts2bMA7Pi9du2a5++KiYkBAOTLl0/GmWPB+dX8/XxNjEnLsj9uhQoVMmOQ+DoWo6KirIwZM8qalz179huKQc6bfM+94HzNMS9SpAg2btwY8Bh3bJlkzJgRgBO3RYsWxY4dOwDY6/i+fftw4sQJ38ZidHS0lTlzZhk3M464LnrFFmORj+F7X65cORw+fBiA856HhNhv75kzZwDY484x557m0KFDQb/DPZ8mJyfLz/MzYIy9r2MxIiLCioqKQtGiRQEAGzduRKFChQAAu3fvDnp8WFgYAGff+f333wd8v1y5cti5cycA4Ny5c56/s1ixYrKvccdeVFSU/I4LFy4EfC+9ePXzuhgeHm5FRkbKenTt2jVZowj399euXZO/Mxb/+9//Bjw2NjZWzgD79+8H4Iw7955xcXGyRnI8ONdmypRJxi4yMhIA8NNPPwX8DnMfy73qvn37fB2LmTNntuLj42XPeDNwnxHMvWVajyGlS5fG5s2b033+jBkzBu1t/RyLkZGRVnR0tKxBZ86cwY8//ng9PwcA+PnnnwEE7nG4VnJPwufj+T48PFzWRfe+JSoqSr526tSpgN9Zrlw5AN73ArjOPWoIN7rXg/sQFh0djYsXL6b5+Ntvvx1A8CJDMmfOLG9G27ZtATgXIxMmTJDHPf300wCAN954A4BzSDAXKG4M+KYkJSUBsBerpUuXAgAWLFiA7t27Y/fu3b79gCclJVkdO3ZEv379AABZsmSRA1Xz5s0BAB988AEAoHz58tiwYUO6z5eamoo1a9YAsCccwLk06dq1KwCgdu3a8niO4ccffwwA+Pe//42RI0cCALp37w4AePnllwEA77//PgBg7969Xr96o2VZ5a/jv/y3pGjRotbEiRNRuXJlAECtWrWwYMECAPa4AfjFsTOZPXs2GjZseN2Pf/DBBwEAS5YsAQBUr14do0ePBuDEcOHChYN+jvMNDwd+XmxKlixpLVq0CLlz55avcfPATZUJD1jcvJOxY8cCAB5++GHky5cPANCoUSP+DgDApUuXAAADBw4MuiAmefPmxapVqwA48+ewYcMAAL169QLgjJsLX8ei1+UkMTbHAIC+ffvilVdeSff56tevj/Xr1wNwDr+86OdcWaRIEdkE8gKaj8mSJYts+CdPngwAqFOnjnwPcC7RTPwcixxDr9h49NFHAQADBgwAADz++ONBl8ec1/7v//4PgL034tzI7w0ePDjgZyIjI2WuZFxVqFABADBjxoygOHcTGRkpG84WLVoAAKZOnerrWIyKirJy5cqFAwcOpPkYro/vvPMOBg4cCMA5YJcoUQKAc3kya9asoAPSO++8AwBo3769fI1Jvv/85z8AnL3PmDFj0nwdHLP4+Hi8/vrrAd/TWAzeKwBA//79AUDGrU6dOrjzzjsBAC+88AIA59zRs2dPAM6+0osVK1YAAKZPn47x48cDcJJEXG+zZcuGbdu2Bfzc/PnzAThnk59//hmjRo0C4Ky98Pm6yD1qx44dAQBbt24Negzf86eeekq+Vr16dQBOYu+rr76S73Xo0AGAc37gwdmct/m5WbZsGQBnzVy4cGGar/Wll14CAOTKlUteb926dbFs2TKcPn3a97FYvHhxAMDEiRNRsWJFAM78x/kQAN5++20Azji5GT9+PJ577jkAQEpKCgAnSW5eavFCm5+LZ555BgDwwAMPSAKfv4u0bNkSAPDJJ59I8oA/P378+OuKRS2ZURRFURRFURRFURTFd/wmhUjHjh2xadMmABCVwKeffgrAvq2lPIk3QVu2bAHg3PzFx8eLlIaZMJbKmPJh3rxSTUJ1Q8WKFbF69eqA18jbf2bYmjdvLtmw+Ph4zJ8/39fSYI4hM5c9e/aUMVu8eHHAY6tVq4bU1FQAkCx2p06dADi397GxsfL+uiWgrVu3BmCrd77++mvP12NKt92kJyOGz2/fOY4cFzMjxqxhs2bNANjqj1/Dhx9+CMDJeJhzBbM2U6ZMAWDHdN26dQFAbnDJoEGDADhxa+LnTFjBggWtYcOGoV69ekHfu+eeewAAX3zxxXU/X5MmTTBjxox0HxMVFSVqkXHjxgFw1HibN28WBRczX25q166NF198EQBQs2ZNAMCJEyd8HYsxMTFWkSJFAuS4zGrdddddAOy5FLAVBMxAMRPG8acKp3DhwpJhYdaDarlXX30VgL2+cr6lkoAKvTVr1kiZG6HKwStLVqBAAXz//ff46aeffBuL2bJls6pXry6y9+XLl8u6w3WIWJaVllJK5sp58+bJPPzss88CcFR1XG8HDRqEOXPmAHDUAoz7F198EQ8//DAApxyH8Uq10bJly/Dll18CCMii+zoWw8LCrIwZMwbIuqnCMUsCAVv91rt3b8/n4XiMHTsWmTNnTvd35syZU0oaOQ5UxZ45c0ZUCJSEU8VHVXN0dLSs43v37sWVK1dw7do138ai+5xhWRbuu+8+AM4+x63QMuH8ynPGRx99JDJ7Sv95zmD52r333ivzd5UqVQA4c/ewYcMkLnlOodLOVLKTJk2aAABmzJjh61jMli2bdf/994v6okSJEhIDVatWBWDPk4Tr3/Dhwz2fb/To0aK8cs+/jz32GABbHUB4duE6efDgQVEnGCoeAI4iwTyLnDlzBtWqVcOmTZs0Fg1YMl2sWLGAr5uqVeKl8mJVCfehVLZeL2mp7zi2lSpVkrsG7lVxneviDV2IxMfHW7Vq1ZJDUHpERUXhm2++AWBvuLwYP368SFq4SXjkkUcAAEOGDAEA9OnTRw7vLMXgJUhkZGSaQcQLmtTU1KASGz8fwtwf8O7du8uiQHn1/fffDwD47LPPUKlSJQAQKT3h54YLCwCUKVMGgF0GAzilL+akR7p06SKPYc0n5cLk8ccfB2DLVolRhuXrxSY8PNzKnDmzjF379u1FItinTx8AzsTRrl07zJ07FwDQtGlTAMBrr70GIHCiYokNLxUJ5faUKQKOHPHee++VrzHOKTenpI4/N378eCkX4OZDY9Fh1KhRIg1kCeG7774r33eXIHr58nCRmD59OgBnHiQHDx6UwzIvLlkqldYhDwD+9a9/AbBljdOmTXM/t69jkePIi4nDhw/LppyHWJIpUya89957AOyxAJyyJB68165dKzJfxgnnVB7STT8Lxh0PCbVr1xZJN0ufWANPn5NMmTLJenjXXXdh69atOHfunO9jketacnKyxJnXpi4tJk6cCMCW9NeoUQOAMy6cM1maevToUTmocY5kgujq1atyoGP5Br0wKFE+efKkyIaZlPD75WS2bNmsGjVqyGV+WFhYUFkS5fVbt25Fu3btADjjxs05L6NmzJghB+O1a9f+4u//xz/+AcAZKyYJTLw+T7w8Y+29n9fF22+/3ercubN87s+cORO0h+QF744dO4JKP/memz/D0gxe+LNEiWU2b731lsypd999NwBnHzVkyBCJZR7iWKrMZI+WkgYTHR1t5c2bVw7QuXPnlvMC/V542bhv376gMuybBS8kedFl0qpVKwBOYs+yLCntqVOnju8TBblz57aeffZZdOvWDYC9LnH+5LrGxDfPj79EWuvpzJkzAdixyQtmev7QFyYhIUEuOXj24b85vitWrBBBBZNNX3zxhZbMKIqiKIqiKIqiKIqieHFDCpHk5GSrR48e6NGjBwD7poe3uDRQZFZ4y5YtkkGkIzONq8ybV96+8+aWBkle8LXydtaUk/PWimZylO6kpqaiTZs2AByjFz/fvicnJ1vdunUTqahpnMr3jvLA5cuXi3kVS15YfsEMZO7cuUUZwgyKewy7du0aJG8ybwfTcoXmTf99993nJeH39e17eHi4FRsbK07L99xzz3WVV1BVQMknb+Xz588vcnxmkZm9ZCY5Pj5eyl54i+8uszKh7Pjzzz8HYKuJmIH75JNP0KFDB+zYscO3sRgREWFlz55dZPomVBtQETB16lQpn3BjZsQoD33zzTcBOCorZj0uX74spnFuFi9eLJnR9AyxKQfn723durWvYzFfvnzWgAEDxIx69+7dQc76fM/mz5+fZobEbcAKOIap7s9Iy5YtRenDdTY6OhoAAozO3W7vNBqcOXOmqI8eeughAP5eF6kQ4VpYrlw5eT/ZfYJz7fz58wOMwgFHQUklFQ1Yvahfvz4ASLmMFyVLlpSsJVWzTz75JADHTO7w4cNStrZo0SL+qK9jsXz58taGDRsCYssoYQAQbB4PBGctDTO+NLu3UWXQvXv3AFUt4MRy8+bNg0yUWdpG1QFVJYCtmJ0wYQIOHTrk21gMCwuzMmXKJGcKL3N4L+Wyu7zQq5yYcH3jWKRnuAk4nx2Wf3Du5c+b5ZI0Oe/UqZOvYzFLlixWamqqGE2nR8GCBYOUPobxvnzNqzQGcNQFNWrUEANcjjvPpBcuXJCSHSoQeC6hGfLs2bPlvGpZFs9Hvo7F6OhoMZZOSEgI6opFNTLV54AzZpwjeWeQlJQkyljOeywpJWXLlg0w0gUCzVp5Bpk6dSoAZ4/0Cwo+VYgoiqIoiqIoiqIoiqJ4EX4jDz579iyWLFmCxo0bA/Cum+PtmpnlojKEbUFr1aol33OrCpgJY2Zz3bp1iIqKAuDU2bIdlgnNxZghJefOnQvwoPA733//PXr37i0qnVdffVXGkTVYrHVesmSJKAjcmBky3gbSyNZtonnfffdJxps3s6Rt27bImjUrAKdlMjM6rK2vV6+etHpym+j6latXr+LUqVMSW1988UVQe+qVK1cCABo0aCBxxXo/Gp/S66Bnz55iOEaFCL0lmGkpUaKEfFbo+8I63bp160rdHuvgeYNrwrq/Dz74ICjj5jcKFiyIKVOmSL272XqRc5zXHEuFB2/R6bc0b948GSvewtNHho9lG17A8V2iDxPr6fnagGCFSKtWrUTJ1aBBAwDOZ8qvnDlzBh9//LFkDffv3y9ZC2Yr//nPf8rjOS/SoI+ZLMbPokWLRNXoVobQ3+f9998XpQGhMmTatGmiAKTCiIowZl0PHjwobbJbt24tij+/Q8VqfHy87GXohcUxdKtDACdjzbVw4sSJEoNUxXHuozKkTp06sua5Dadff/11qX3nmulWnfzf//0fHnjggYCvXY/Pyd+ZU6dOScYSACZNmiSxRKgMeemll0SFQ585Qs+d7Nmzi18FVXdUIpjtXKkMcbe8f//998VHiPHJbClVBU8//TQ++ugjAPb+lYpZv3Lt2jWcPXtWzhBxcXESO4TtkVetWiVj5c40T5o0CQCCYgRwPLrSU3KRVatWSXybqiAgUBnCsde51CZr1qyoX7++mBkvW7YsyBeQXnbPP/+8/BzXTs6/HOuRI0fKuYVq9qtXrwJwvF22bdsmXoT0WqKyLyoqSuZH7mF47qRy0tyz+n0uBexYNNuOHz9+HEWKFAHgVGHwvGhCRQfnur59+wJAgFqO5wzCvelXX30VpA4yqjvke/SG4jzBeYCKTsCJU3db87RQhYiiKIqiKIqiKIqiKL7jN7Xdvffee6WGjpkoZjtNmOXizR/JnTu3uMi6b+PYMsndJcGkSpUqkkVx14W6nfcBOyv+/fff49KlS769+ktISLAee+wxGS/TAyJPnjwAHGdfE6o1mE1m/evx48fNjiEAnLGkX8Xx48clg8JWTUOHDgVgq36YAWOdO2EGOiUlRbJ0RvcNX9dnMhbNullm9VkfS+8Ofh9waqN5m8p6wISEhCDlAcePLdJmzJgh2St3vH788cd44oknADiuz/Qo4E1y8+bNA+q2AfUtMP+dmJgY1GKaWcVatWph6dKlABxHdDqkk5SUFMlGUx3k8TtFbcWsKF3gzY5RzJQzS8O5Yf369aJeYeZt6dKlvo7FnDlzWk2bNpX28EDareEaNmz4i22wO3fujLfeeguA3YYTcBRd9O7xUu6xq1uxYsUkdqka4/xJIiIiJCs3f/587Nq1CxcuXNBY/B9eygJmxizLCmprTJi5zJMnj4yd+3vMJHvtbTi/XrlyRXwoqMYzXqu8Dqrx6DPy3nvv+ToWOY7FixcHAJQqVUpigV5Y7IYwbNgw6W7I93rw4MEAHA+n/z1nwO+gdxJVIOa86QX9CZgd9Yp/rt27d+/2vW9BtmzZrAcffFDUiVu2bMEPP/wAwHmf6Dfxj3/8QxSS7owz2bNnj3TKo4KcCgW3ohxwfEFSUlIA2GcKdqnhWHLfRGWPCRXtly5d8nUsRkdHWwUKFBAPnjx58ogCg3tMnvOKFy8uyjkqJbmHoeJ87Nix0k6V/nQcIyo+3nrrLRkTeuqZymV6j9CLhG12+XmqVauWKGenTJmCPXv24OLFi76NxbCwMCsmJkbmzm+++UbmVsYiY4ExCjjnOyqweN4AnLmW9wLc4/L82KNHD1F0uBV35pnfDT8/9evXlzMH12z8nm13+R9hazEgeONnwpKIzp07A3Ckp926dZMPdloXIAMGDBAjIy/ckhhKnyij8TIu00OYc9nw0UcfSakMpWVmT26aMfJzwnaQfJ/Tgz9TsGBBuZjyMmvkh52/iwc9MxhYZsWyK/jcPI7jSMnfuXPn5EDESYhx8/jjj8tlhylNTAvK6vnzzZo1A2Bv0nkY5jiam0VOgjxoT5gwAYBjyGu2+iR+jsW8efNa/fv3l4uj1atXyyLDSwvKEc1NNw9BXOh5QZI1a1aR87slh506dQJgX37wgMWWZDSmW7dunZiq8qLUfcnZrFkzOZxzkRsxYoSvY5HGYyxBMUuPCA9hoaGhsqDTtJhr53PPPQcA2Lhxo1xCcvPAclGadiYlJclmzk2tWrXMeRKAU87I11ikSJGgMg0/x2KpUqWsJUuWyEHYvLiqVq0agMAL5rTgBr1y5cpSKsFYzpcvHwDIxfFdd90l0mImGhhnVatWlfWQMX2dbQ19HYv58uWzXn75ZYkNSvIBZ97kxv3SpUv48ccfAQSXzHB/5LV/ZDKIh6i9e/fK1xiflH+vWrVKnpsXIvyTm/Zy5cqJ/Jt7az/HIvc2pikqSyXYspxJtGvXrklL8vvuuw8AkCtXLgCOaWOLFi3kIqNUqVIAnFJhjluPHj1kPmTSzeSOO+4A4LSJdZ9T+vTpI3Mry0KmTJni61h0XzLHxMTIXEYTVK5rkyZNklIjvtfcd9IwetWqVUjrvMp5c+XKldIUgvHFy2IzOe6GRrg8owL2+eObb77B+fPnfR+LaZm7Xy8ct3Llysl8yH0Qk25cZ7ds2SJlOu7fm5SUJOVRJ06cCPgdv9D+V01VFUVRFEVRFEVRFEVRvLghhUjmzJmt0qVLS3aYt3wmlFa72w6aUC6VVvtHwDGpYjszk/79+wOwFSbMUtJ4h3In3jo9/fTTkm1bsmQJKlasiI0bN/r+xo/kzZtXbtyY/ecNa8GCBeV2tkuXLgAQpNYZN26clGEwo8U2W6bpGLOfLL3h5y42NlYyNqYy5TrQ23cElC0EPYY37TNnzpT4cLemoylru3btxDTODTMnR44cESNHtuo0YekFFSaUJVISt27dOsnIET9nwsLDw61MmTIFSA0J1TWU7cfGxspcxywXs15UfJQqVQqNGjUC4MQrM5VsPff222/L+LC8kZmYN998U9R7zJRQrcW5ul27dmLaSxkxNBYDVHd169bF8uXLAUCyl8wmDxgwQOS9zBRzraP67uDBgwFmfSb33nsvAGDTpk3yNUqSueZ9//33olKZOHEiAKfEhhns3r17y1ycmJiI48eP4+eff/ZtLEZERFjZsmWTTNTJkyel1IJwjn3mmWdkrWNJErOR1atXB2C/v+445VzN9fHHH3+UjDM/JzQCNPdWNMilCpYlAtmzZxcll1GGobFo0KxZM1FFcs2iqpKG4ECwEs6EY8S1iyogfgbM52SpNtu0WpYl8zP3UvzMUBXEPStgK1UuXLiAq1ev+jYW2To5KSkJAILKSIFAZTiVxFQ4M5Y4zxYsWFBUyJxzqUwwx9v9GaCZ7tWrV0XBwzmBWWiWYJw+fVrGV2PRxh2LJnyvqWbevHmzjDfLRVmaRjuEggULypz66quvAoAoh1heU7x4cVGf04hz3bp1AGxjcap3OM+zTJ+lqDVq1BClNZ/Hz3tUjiHPCy1bthQlD2PAnEcJx4cKZ7P1sruqw30Hce3aNVy7dg0AxJD3hRdeAGDvkRjXkZGRAJzzKvc/Xm26oQoRRVEURVEURVEURVEUb36TqaoXzEqvWbNGPEaoPGAGjRmT3bt3y8+xPQ8zLGXKlAFg1wPy9o/Px9vaF198UZ7bXTOdHnrjF2BOKnWUzEDTjGjkyJEyLsRtalWuXDmp92ILQbdxpukh4mbo0KGiHnHDOs/GjRuLysfA17fvxYoVsyZPnoy7775bvsZsPj0lvGBWil4evLn99ttv5eaXfi3MlLDl3cWLFwPM5kz69u0rZlZU+LA9N42NBgwYIFk6KhH8HIv58+e3Bg4cKD5KXtBn4sKFC3KTTs8fmo3RGK5YsWIyVjRXNW/2fwmv7CizNpyXO3ToIJ4IhsePr2MxMjLSypUrl7x/bKMLODFJEz/TXIykl51mBo1qDsZto0aNJBPKLBeVRn379g1ob2fC9fK2226T+XX27Nk4c+YMLl++7NtYZFaaGcRNmzbJGke1lZeCjnHBebRmzZoA7DGgLxrHlY+lwnXRokWi9uBj6H9w9OhRhIba+Sq2liRUEvXo0UNMQVknD5+vi+Hh4VZsbKwo5caNGxcUXwMHDgRgr5Pci3JdpEKHGUkzjuhpQbUVH/vJJ5+IGpK/q0mTJgBspQg/N+59Dr1kihYtKvXwnNP9vC4WLFjQGjlypHj2mC0zOWcxTkzFMlWVVARRMeIFPwtUeEydOlX2L9yjUO3z1ltviSqeY8jMNbPdzz//vKgq2U4dPo9FnjVM/z+qUhmfpHXr1nKGY+zQW4tjVaNGDWllzlikgS1VW6+88orEYHR0NABHOVClShXZF3GvS880nl2GDx8uj69bty6WLVuG06dP+zYWvc78jAF6BqYHlSJnz54F4KgbAWDhwoUAHOUdx71Lly7i0eNWe6xevVpaYHOcuB5yzk4DVYgoiqIoiqIoiqIoiqJ4cUMKkfDwcCsuLg4nT55M8zFmu1u3XwFvdekQDQTXExHemDds2FBujlg3TUfgcePGYfr06QAcVYMXrEHctWsXOnTogJ07d/r2xo+dLagQmTVrlmTF3JQuXVpq2U3HbwCSvWKtF+BdCwbYt6503GeGlDe7pUqVkpthehIwC8AawoMHD0odoIHevsNpp/nyyy+LlwBrnqnM2LJli2cdLuAofpo2bSqZM/pO8IbddFT3ymK7YcvWtWvXBn2PdX8LFixAly5dsGvXLt/GIseQdbBDhgyRTCMzV+nBeKMSZ+bMmeJ0361bNwDOZ4HKgiFDhkhcsVsMnd9r164t3WiIu8NGnTp1JFvGOtzZs2f7OhYzZMhgJScnS0vxBQsWSMaJMUnvDgCS4eCaxfmPGefbb78dGzduBOA447ODAuvTLcuS30FPLnP+TStOqfhhpg2w544JEybg0KFDvo3F5ORkq1u3buLXATj+ZXPnzgVg16ADtp8PWw8yC00PEa/3ne12mVUmjz32mPhmsSMG+fzzzyX2CBUn/HPBggWizmSmc/Lkyb6Oxdy5c1s9e/bEM888A8Dej9K7hfsLjtH58+dlX8FMJD16qH7zaoXOGGSL5MceewxNmzYF4HQUoqpo8eLF4rHGvZN7jgacPc/cuXNx8uRJX6u1uC7y/DBjxgzxDSD0MpszZ46sZ4xTwgw0FQqAM3ZUq7ONa4kSJeRzQbUXPWK6d++e5jnF7I7JuYOfsx9//NHXschxNLtFMi7NFvWArTZgB6Ft27YBgKyBL774IgB7v0HPO44/O5Vwjl2+fLm0X6ZXDz3PQkND5UySnnK2YcOGAOyuOAsXLsSJEyd8H4ukfPnyooRiVxeyYMGCgFgDnD0J1RwZMmSQjncVKlTg7wDgxOSBAwfk7DBp0iQAjo/hW2+9Jfsmei+5fz45OVmU0kbHGVWIKIqiKIqiKIqiKIqieHFDCpFixYpZ7733nmSA4+Li5CaH2WjWNT/xxBP47rvvAEAc95nJZB3zoEGD5ObHfZvH7FeuXLnEvZneIcyImbe6afH222/La6MPhZ/rM9PzgRkxYgQLV/fiAAAgAElEQVQAu54SgGenA35e6PEycOBAyTCn9Vlq3769dDsg/B3p+Sf8Ar6+fc+ZM6fVrFmzgIwFa2iffPJJAE5N3n//+19ReVChQd8BZlXq1asn9e+MN/p80JH98uXLePfddz1fz6JFi8Spm7Wfq1evBuBkzQDnM9azZ08AGouAo77KkSMHdu3aBcCpueT4Tpw4UTKSnCuPHTsGANi/fz8Au56acyTjk5lnZjCzZMkiz825lh1oNm7cKGN1PUogA1/HIv0nqAbJly+fzJ30f+A8d/nyZVHJcYyo8jLf87x58wJIv1sb4bzLmH7qqafkd7Bm28unhOqEo0eP4uzZs7hy5YrvY5FYlhWkWCR58+ZNc1yYeTx9+nSAAgBwMtWsX586dapkKNkthvG6bNkymWvpbbBq1SoAztxgwrGPjIz0dSxyHNmNYt++feLjQ2UGmTt3riiMuWYlJib+4u/guDIjec899wR1e6NCpHbt2uI5Yqqn3VCN+fHHH6ty0rUuRkREyHvm3j9ER0dLFzvud6hwzZcvHwD73NG3b18AjkcLu+7R765UqVIyN3INpd8E1eom9Blh3Hbs2FH2WMa+2dexeMcdd1gffvihvNe9evUK6KgEON2Y3njjDZlnOZ7sEkPvpRUrVkhHTDdu1YEJPw9ff/21fI3rIX8nlZP9+vUTBcSUKVNw8eJFX3d8cq+LEydOlPWI7yHHq1KlSuIrQv8t+pqVLl2azxf0O9zd2AAnljlm7Pjz73//WzoeUilSuXLloOfk3Ex13+rVq68rFn+VqSrlomxP5MXgwYPF5ISbQU4+ZntVSvYpc0qvHIfSYrbDCgkJkQ3Es88+G/BYyroHDRokB8WmTZti2LBhOHDggG8/4NHR0VbevHnTbW2bI0cOAM6BCwiUBgIQI9YlS5aIrIntO3kRRhPcsmXLyiWWW3pYp04d2bybsnLA+Zxkz55d5P1sUXno0CFfLzY0cuQFxyuvvCKTCGFZ0qVLlwKMrQBngqIBGS8rveAckZiYKJeZ6cGNO0sDTPi1mJgYrFmzBmfPnvVtLHI+5Sbh888/F5kvN3A0oapSpYpcNhMuRNxM58iRQ2KWC49pwOiG8cV2ydOnT5cYTovmzZvLZaaxuPk6FjmOLCsySx3cFxH33HMP6tSpA8C57OA48gBsGk27TaxNzFIp83c9+uijcqnphvLyefPmSZlFTEwMPvvsMzWPM5g5c6asa1yXaCh9+PDhAFNywDEJ5AWUF0wK8GcB5+KLhzd+hpKTk5E1a1YAzqaQ7QW5FhcoUEAMOykzh8aiBTjv1cqVK9O83B0xYoTEFzflefLkAeBI+l944QUp13CXPLEExrIsKR/kYZiXzElJSSLnp2z8+PHjAc9TsWJFSR4QTRQ4DB06VA68PIwR83KSBuSMV+41wsLC5GKaF2AcSzMhx5JhxiAP36NGjZK2yry85nizTGbMmDFyHuLrmDdvnq9jMTEx0fqfRQEAe27kGsXLZpoW/+c//5EkONtUM4a5n+zXr5/sh1heynJj04aB8CLL3T4dcM42vHRh0m/YsGEyps899xzee+89HD582LexGBsba1WsWFHWt6FDh8oYco9qxpD7PoHjxPPf+PHjpdSGVg3ch3jB5+O++Ny5c7LvoZE4L7z5vGXKlJE9GNfjPXv2aMmMoiiKoiiKoiiKoiiKFzekEClRooT18ccfi/TP6+Y9W7ZsAIC7775bbtYp8aS6wEthwmw2b4so8TYNPykLpXQ8JCQkXRkiYGdwzIwM4O/b9/j4eKt27dqm2YxkNaga4E2b+Rhmp2nuR0no4MGDxWyOKp/XXnsNQLBRHBBsKNaiRQsptaDxKg2TTDMmws+cZVm+vn1PTk62evbsKYobE7fCp0qVKiLpZctH3qoym3Lx4kUZP5qK8VaX8VqrVi2RojFTwzHu3bu3qLJo2MjHUlZKA0ATP8ciM2E0Dxs+fLhkKKnaYAs/s+Un52zKCJs3bw7Abr/rzqARykYbNGggWTJmpdm6OSkpSeKLZpHMiDGjNn36dNSrVw9AQGtJX8eiO6NZqVIlyWp5meS65dYcc2ZaBg8eHPQ7mMl6++235Ws0hGTZBVV/M2fODFJ8cS5ltu7HH38MUp/4ORbdZpwmXAdZ+vvUU09JzHkpd9LCvdc6fPhwUIkG19JChQrJvEC1D0sayahRo7xer8YivPcOhDH06quvirqAY/Pll18CcObEChUqiBEjx4NzJOfGqlWrSkaSRpwsEyhVqpRIwd3G9NzzbtmyBYULFw54jX6OxdDQUCtDhgySAf7oo4/Qpk0bAIF7UsL3laWjhLG5efNmMcokbHnMz0lqaqqssVOmTAHgjHOdOnVEQcBzBhUj6Sna4fNYDA0NtcLDw4MMcU0Yd2fPnpWYYXxQ+cizx9tvvx3UmMNU7ZGEhAQAwUqsBx54QPa2HFt3S/WJEyfKeZWqWj/HIufTTJkyAbBL7mlYSmWyua/n+sVSmZiYGACOKnLNmjW44447ADj7FZ7/YmNjAdhzMBVdLIcx107Ow1Rn8txoYqrj+WOqEFEURVEURVEURVEURfHghhQiCQkJVv369cWTA3BMimiOSvr16yc3fVQeMJvNete+ffuKQcrKlSsBOHV/NCf78MMPpbZMXrSHMoXZNipM6Dmxd+9e8Z1gGz298QusJacCgJlKvpe33367ZKEJ699pdNSpU6egrAhb8vKWLyQkRG52aQjJVlqAM2b8vVSo0LC3ZMmSclts4Ovbd3dWunr16uJF4G5BnZKSIp99t7EcmTx5srzvHA/6T9Brwow7GhUz4/zee+9J1oW4/RNKly4txpOMST/HYkxMjFWkSBFP82IzPgHgzJkzcoNOA0Defrv9kwAnc8X3m6Smpko2k0oucw2gbwIzclSRsF60bdu2Mp5sUbls2TJfx2JYWJiVIUMG8UJKj8KFC0s9Lk1VqbgxYQwy48JaWL7nH3zwgagLGLecIx999FGZO5kBpUEgM6zTpk0Tv5gmTZpgyZIlOHnypG9jkfMpDfWKFi0qa5WbmJgYGeuxY8cCcLx+6C3QoUMH2cOwnSTXTq6XjRs3Nr0/ANhtBfk6mFVjRo5ZULffhAtfxyINjklISIgo2aj0YAyxHa4JjaYZLwkJCRJ7HCvGFufGypUry1hzLh42bBgAe+yozvRSqwC2wouvjT57fl4X3XubAgUKiDKDaxUV5NOnT5f1iGsXzxsc37i4OMn6sw091SdeijB6iCxbtgyA4y0C2Mo6AGjXrh0AZ16eOnWqKPyqVKkCAFixYoWvY5HjyPdozZo14s3DceRcNmjQIHz66acAHI9IKuO8VJbmPAnYe1zAVibQw5BQSdm4cWMZP55R6NXFs6rZDviBBx7A2rVrfe1zFx8fb9WsWVPO+TwjpoX7PoFKdKquvHBXH/Tr10/ik0o9Uq9ePdkTuz3wvO4FGO/X2wJbFSKKoiiKoiiKoiiKoviOX9VlxoQ36cxi8FZv6dKl0kmENdG8IecNz4svvigeBLxFp2swb1nXrFkjbX5ZT0/VQu/evXHu3DkA3nVEgHedrZ9v3yMiIqz4+HipwXvjjTcwe/ZsAHZbq1+CDtB0x+/UqZN0neBYuLEsS8aVWWkv3FnxX8DXt++hoaFWZGSk+IW8/fbbKFasGABnbEj//v3le7xJp5qHVKtWTVzw6cvDbBVvXqdOnSqdo9y38CZUitCHgp1tvPBzLNLPhzWvEyZMkNpLNy+//LJkN+gDQ/UAOw1dvnxZbuhZp8nnpv+BCX0mqPjLkiWLeFXQZ4S38KzbLF68uFd3MV/HItdFdpZ45513pH0xM8WsaS1fvryMEbMu9PGhssur6wQVdaa/CH2YOnXqBMDxDPJqQ0c4/yYlJWHatGkAnPbafo5F+qOxhXx6HfSKFi0q9elUSs6aNQuAU2sdHx8vyoQtW7YACMxwAva8mlZ3mipVqsh6TIUs909UnLRq1Urqr6kEK1++vK9jMVeuXNYTTzwh+8qRI0d6dtj632NlH8S5kG1zzVarXi2rAUcV9Pzzz8sax88POXDgAHr16gXA+Yy48Wrj7OdY5HxKv4djx45JFphdSegzd//998v8RzheZoted8cRKiC559y2bZt03GObZHY56d+/vyjX6Q1FqFBYvHixqFi0+5pNiRIlrLlz5wb54wBOi1Uq6sqWLSuKnpdeegmA8/4ztipXrizvLVuXs3si90LZsmUL8pIhc+fO9fI/A+D4V3jtv/wciwkJCVadOnWkQxrgtDimp5XZjZTrIb0/eT6hyqdDhw7iJcqYopcZzyY1atQQLyfuWzmm48aNM30kATjqEyplDxw4IL5CxtytChFFURRFURRFURRFURQvbkghkj17dqtmzZqiCOjbt6/c2LoZM2aM3NqxjogqAd6sA049J7Mgw4cPD3ge8/V51QgxM+J2e3b7UgD2LfDjjz+OrVu3+vbGz63yad68uSgBroeuXbsCcNyZCxUqJD/PjCezZOxSAji3d8yIsdOC2ffbfQvPejXzxrdq1aoAgOXLl/v69j0hIcGqW7eu9PkGHN8dut+z68uYMWOkrtYNPQU4PgAkW8WaaXd3Ay/atWsnnUx2794NwKnH9qJDhw6YO3cujh8/rrHogVdXEc6FHEuO89ChQwE4XbwAiL/EoUOHAp63Ro0aklX2eD0Sl6xpp3qBmdMHHnjAK2Pq61ikhwjVOPv375dY+Prrr4MezzplKjz4PtKHq1KlSjIOfK/p1M917ciRI5L5ZraTa+nnn38uWWlmX9x+CZGRkdK1jfg5ExYeHm7FxsaKqg1w1h/Oh+Y4ca3jvoV7G45FYmKiuOhzDPfu3QvA8TcrVqyYPCZDhgwAICo9wMm2MQa99j9UOMydOxcA0KlTJ1/HYs6cOa1mzZqJOnnSpEnyftM7KT3oH0HlwLRp00Qdx/h0dzDo06ePdFbjGuxFWvX0DRo0kO5gLVu2xKJFi3DixAnfxiJ9YLJmzQrAfn+YoWYsFSpUCEBglycqW6k6oKLyxIkTogyhWpxealSfr1mzJqgLkFeXGUJlAtULGTNmRMeOHQE43TMWLlzo61hMSkqyOnfuLIplelYBEO9IxlmbNm2COm4RUxHgVvqkB9dHzgV9+/YVnxLO85wTuF43atQISUlJ8vh3330Xhw4d8m0sxsbGWqmpqbIWlStXTrw+WLHhjpfrhSo+3geYay/nYY4Pz40pKSmibKVKiMos+o6kwXXF4m8qmWnZsqUsALzI4Af9vvvuEykuN3du8uTJEyRv4uuhZLV48eJSjsPJzGyLlZbUiTKtPXv2yOY+JSUFe/bswcWLF337AXePYe7cueUDTZk3J4TLly/LQkAjR7fRXKtWrUTCS/jh5WXUV199JRtz4tEWSUhLourC14sNS2a83j/D7DLoezSY4kbAK/4pieNhjhPWJ598goULFwJwzKhYrmEerviclJ5yQ2+yevVqtGnTBtu3b/dtLObKlctq0aIFwsPDAaRfTvb000/LBRMl+JTUU7abHixJ3Llzp8gX2TqQZnB79+6V0kMetObMmQPAbh8JBF66GPg6FiMiIqy4uLgAmT3hQZkmgGXLlpXDL02QOafSkDEsLEwuV3gYZwkTS9a2bt0q7e4oS+XBr0+fPlIGSeNHys8ZizVq1JDXWLx4cezdu9fX62JUVJSVmJgoxqWXLl2Si33+yZgwYTtqGi+aBoyExm4se+L8ytgy4Wdgy5YtMmY8GHIN5uWWuefhYWzcuHG+jkXub3j5sGHDBjGxJTzEJiUlyX41LYoWLSoHXMK1z5yvaQTJ9ZGb85UrV0opMZNJHEcSHx8vc0fHjh3x0Ucf4dixY76NxcyZM1ulS5eWpNtnn30max7XH65nV65ckcOTG+5ne/ToISU23FMyicCxXLlypRywmIzgnjUkJEQO6zTz5EHexOMS3NexGB0dbeXNm1fixzTHdV8o9erVS5KdtGFgyYxJxowZAQS2egWcpKlp+smEPMuH27RpI2PL79Fsl2eV+vXrB5mj+zlREBkZaeXMmVOSq0WLFpXzt1drc15yeBn9A3ZCh6XB3PeaFyGAnRRgm2zOp6Rfv35yccnEHi+TyfPPPy+lUdzjQtvuKoqiKIqiKIqiKIqiePObFCJxcXEip6eMhZg367yNo4SJt7PFihULKKsAHLMk89aH2TUqRWiEs2HDBsnQUIXizlh369ZNjEBbtGiB5557Dt9++61vb/w4hryZfe2116REhVJ8rzagzG55lUjx1p2SRUqpOG4dOnSQ1kq8WedtMG/lAQQZ2tH4Mw2TVV/fvnMcmW3OlStXkDGbCdse86aXbXgZiyY0I2M5jin9dqt2zHjj2PLml1lqL1atWoU2bdpgx44dvo1FSoPTU0LRaKpUqVJSmkYVB1VWzEbedtttQeoqzs+MyXHjxolKyLg9F3LmzAnAMcqiMoQmaICjVmBczp4929exmD9/fuvVV1+V9e3w4cNB48BsoynxTouYmBjJalPhQfh8gwcPFrUI5eOM8QcffFDaJLs/W5Qap6SkSGa1QYMG+Oyzz3Dq1CnfxmJ4eLgVFxcnpbd9+vQRZRxL05hFDAsLCzK9dWccH3zwQVHIMoNWoUKFgJ9Zvny5rIMsXaTx6qVLlyQGT58+DcBRmjAWzYxrz549AQAjRozwdSxyXYyOjgZgq7CoiGO5GRU2Dz74oKirWAZOvMq7aZDM2DLLmxhnXmW+hGoErrl8Pc8884yoKatVq4YNGzbgxx9/9G0spldK6oV7rmUGOVu2bABslR3nX5rLc7y7d+8OwC6vcatQjNcjf+c+mC20r169CsBuZ8/Y5ecNPt+jUl3AJhivvvqqlD6xfNAkLWU41VPZs2cXM2sqdlhKyp85deqUKA5oQE41tFk2SlNVquNZFtewYcOgucDPCpHExETrySeflDWwSJEiePPNNwE4+w2zQQPLCllS2rx5cwCOSqdPnz5BqnS3Sar5NcK9SsmSJUU5lJ7xOasUaM564MABVYgoiqIoiqIoiqIoiqJ4EX4jD46IiECuXLmkvnnNmjWSIaZChJmOrFmzSmaZWRP6HbAW/vz585Ih4S3Trl27ADhZkeLFi4uxIBUEvLV//fXX5WaP0GSVtcCjR49GgwYNANjtCfm8fiUpKQmdO3eWGr6DBw+KZ8imTZvkMYB9C8vsVnqGRzS+4S0eFTn8d44cOaQemnXPVIY8/fTTUkt/9913A3Bu6M1aPhozpaeC8CO8Ie/Ro4fUKHvB+nf+yXGlp0CzZs0C2mcBTssrxkyePHmCjG+pKPjuu+8kI8PsHPnss88AOKoUPiY01N/3sRs3bgy4CV+yZIlkCgmVAlOnTpW5krFIpZz5HJzrmOXgDT2NxQBvZQifh212+VlgNpo+I1OnTpUWocx0pqcE8gP//e9/JeNEWMNK/wcqQyzLktpb99pFfvrpJ/EDIfQoYBalffv2ouijiTE/B1988YVkatx+Qsy2mXX3s2fPltfkV65evRpgzD5kyBDJajHjTE8CM5YIVVozZ84E4LSOBCBZUcPnA0Bgm3sqPPgaQkJCZJ912223AXBi0SuTSnNXxYZZ+pkzZ4qKmBlitlfNly8fnnzySc+fp8+LCQ2q3f4FISEhptk7AGf+PX78OIoUKQLA+UxQXcC5wZw78ubNK0oFvxIWFoZMmTLJ/uDUqVNpqihHjhwp36NvoJepP/c7XMc4vozlokWLSrxzXqb64+eff5Zzzs6dOwE4ai/TU89Qhiiw9yn9+/eX94yKRhNzfUprjE0DVSqvuO/lv6m2ioiIEKX6pEmTAp7nySeflDMFY47jSrWX6YVXp06dNP1p/MLRo0cxYsQI2dvv3LlTDFa5Dpl7QyqRCQ1YOT9SeQc4nqDuc8s777wTdN7jfurUqVMyJjx7UPVlwj2qe839Jfx9IlEURVEURVEURVEUxZfckIdIzpw5rSZNmkjGd+jQoVKjPnbsWAC2CoOwlSdv6lhzztueYcOGyU0fb47oUcEMV3qYngZ8HnadcTsFA/Yt0ZAhQ7B//37f1oTdaH0moXM2x8l0/+XYcwzYTpIMGzZM/A/4meDtYHrwZw4fPiyP540ufF6f6R7HmJgYiT12fGIG+c477xRHfN70Et6+Z8iQAd99913A96hOMGsEmalmJowZtg0bNkiGhD9HJQIVQF9++aXc3BI/12fShZ1+PuXKlRM1HW+2H3/8cQB23FWuXBmA4+Tt5sSJE3Jrz+wyoS/QnDlzpL6TGTRml2vVqiWvhQoFqvPo6p8GGosGkydPliw0u7kwY/zRRx+JisfNokWLAHi3uXa78g8fPlxUBVSIMEMKIChj7WbkyJEyX9etWxf16tXDli1bfBuLZcqUsZYvXy77BxMq25ixptrRhOpV0xOLmSyOBddFdhlJTU0V3xnG+bvvvgsAaNu2rfj30PeJay4zYjt27JDWoo0bNwYALFq0yNexmDt3buvZZ5+VzH1KSoq871Qvc2/Yo0cP6ZDHOKFPC30I+vTpI2sn/SbYPpTZywYNGogymTHFedyEay+fm5+na9euiSo2NTUVgL/XRc6nZut4znH0SyInTpyQeY/vr9unacqUKdJ1KC0iIiJk7AjX0r59+8p5hJ5tVBJwP+TVSQ8+XxezZMlilS9fXtSJGTNmlPmVaqvrgSqSIUOGiCqACgLCsZowYYIowRjb9KXMnDmzfEaoeqbKlV1Rzp8/j5iYGAAB3ha+jcXY2FirUqVK4p/lRVhYGAB7PmMMMV545ueZYOjQofJes9KD6yn9l9xtyU1SU1ODPH4Y/6wuAZzOi+wKhVvRdveuu+7C2rVrAx5jtkVyt7ZKD/flCd8Uft2E5o07d+4UYxVe0lCuT3lr48aNRcZK9ANeSWTAEyZMkO/R+JQyp/Lly8vmjTI0mpBxsqlYsaJ8kHlZQQlVREQEAHtMeaHC1mRmuyseAFiyw+ehGVIaUjpfLzZeF1tuKTU3chMnTpRNAichXmxx012yZElpoW38joB/L168WKSJhBciXbp0kdhPS+5vXqJ17doVM2fOxNGjR30bi3FxcdY999wjG23z0oGL+MqVKwEgoJzJPWdTEh4eHi6SexqncrxYrjhkyBAxyGILOpZRfPrpp/JcfIz7YnrDhg1iamWYCmosAqhSpQoAe8w4RpRvcyP30ksvoWXLlp7PQ+O/EydOiIF4u3btADhlFzQP//7774Ok9e6SQ/PnONZsQT9mzBiRqiYkJOD06dO4fPmyb2MxNDTUCg8PDzoUmZhlRWarQcC57GCbeq/STsq0eUGcLVs2OaQTfm5KlCghm0iuubwU46XJww8/LBcp3Gvt2LFDY9EgNDRUWjh6XMbLJTPb73K/w037d999Jz/HC4ylS5cCAAYMGADAlvszGcC45QVLr1695GKMnxk+D0synnnmGYwePdr92nwbi157G14oc7/KM0WtWrWkdJRm8Lw8oXzfbLvsbtXZu3dvAPYehWsuLzd5CBwwYAC2b98OwC7hB5yLTz7WhGeWyZMnaywaLF++HJ9++ikAZ75i4ufdd9+VfT/bVHPMeX6LiYmRNZKXZbz4NOdN97pIC4AyZcpI8serPXpa+DkWExMTrQ4dOshcZ4oQ0itH8TJKBWwz3bSM4r3g3O1Vls19KNda2mQkJyd7rb9qqqooiqIoiqIoiqIoiuLFDSlE4uLirKpVqwbIQrdt2wbAaatjwqwU1Rpu2dr06dPlNrB9+/b2C/rfbVFcXBwA+8adN71uSX/NmjVFZkwzQBq58qaez2Pi5xu/kJAQKywsTFpSDRo0KMgkjMybN0/kpZSJUvJGc9QPP/wQ48ePB+AoTI4ePQrAyYoCjgEks+A0sxoyZIhkyXiz7iU/9kBv3+GUvDDL/0tQocNMMQ375s6dK5kwZqUJZbwPPfSQKAduBGbJGeuAbSL4wQcf4MiRI76NxbCwMCtDhgwSYx07dkzX/IkteCkXpWqEGZB3331XbvJHjRoFAGjTpg0ASAYyS5Ys8nh3NtxUoVB2ythmqUb79u0lS0Op5PLlyzUWfwEq8saPHy/KLaorqbShdJRlNkCwmTTjfcCAAVIiR4M/ZqDvvfdeMbbmz7nVY927dw8yB/XzupgzZ06rWbNmAe+Ju02qibu9oBvLsoJat7JtK79+9epVUeV5qYYY78xwc+3l67IsS9qAGmgsXiepqalS1kllCNXH3Lv885//DIohQlXy+vXrpUTuiSeeAOCoIe+66y5RGrCMkU0FWOb03XffyR63cOHC2L9/P3766SffxmJ0dLSVP39+GYNPPvlE2lpzfCjFnzdvnnzNDfeczZo1k/0O9yJdunQB4KhoExISRAFGhTT3pdu2bZN5k+NLhbSJu7kANBaDYpEKG3fL1CJFioj5KuHcxrPK1KlTRf3IkiWW9M6dOxeAbTzNOZX7G54r3n//fYl3KoSIWWLBz9PJkyfx1Vdf+boFdoYMGay8efPKeHHPlxZUL3Jd5NzJCgGz/J7tmDn3jRw5EoB9L8C9jPvs0K1btyA1nddnysN4XBUiiqIoiqIoiqIoiqIoXvwmD5HFixdLdpP1d6ayg/VabGPFTMnq1asB2LdxvB1ilovqAlMNwhp71tIyw/Lwww+na/ZCeKO0ePFirF69Gj/88INvb/w4hqx5Hj16tLTmpImpF14+EITjQu8IZrDMm3tmlWmw5G7PBDh+Bxx7ju3o0aOlnrBXr14AgN69e/v69p3mccwymSaauXLlAuC851u3bhXFFNu6spaWPPDAA5Kh9vIiIKzb++KLLwA4N7Hly5eXOD948CBfIwCnVvrYsWNicpQlSxa8/vrr+O6773wfizTxunDhgijtqLzjGO7ZsyfIWNOrhpMeBtOnTwdg12wCjnqgdu3aYsTIGHR7xwDerT3dME6Tk5N9HYu/1qia0PyYxreA4+3DVnp96zIAACAASURBVKtu+vTpI/49VHBRtUePCsDxHPnnP/8JwDYIBOxMJxWbbKHtZ4VIrly5rJYtW4qKA3DUOMxScW+TNWtWiVnGK1UkjJeUlBRROLI1JM0BqZxldjMtqDpgFvs68XUshoaGWhkyZBAj3AwZMgRlg7lmLl++XOrOqTTm2HAN89ofc52kIuCJJ56Q+ZaeFDTCHTFihMyTjFN6HFCZ62UU6OdYLFy4sPXWW2/J+1O3bl3xFqTi0W1QCzj7T/5Jr5ccOXLIz1Fp4tUClrj9nyIjIwNMdgEnJqkqcXsK/Q9fx6LbkDMqKkq8zNxUqVJF1AQHDhwAAGlXXaZMGQC2ooB/p8qKcc596fDhw+Vzw3Ofu8kD4IwxzZB5VvUyzPZ7LL755pvo378/AFs1Q+UV95Hcl+bMmVM8mdwKKnMedZuLpwfjjvE+a9YsHDlyBICzrvLugIquggULimeicc5RhYiiKIqiKIqiKIqiKIoXv0khYuLOKNaoUUNaUbHrCLP8Zl206Q4MOLfnZmaSt+90BqaL9NWrV6WTSVru8MePH5eMwNatW3Hy5Elfu+nHx8dbNWvWDMgK80ad2WQ6QT/22GNS11W9enUAjlcLs8v58+dPs7XjSy+9BMCu7fJyCQZsxQJv/+lt4KZ8+fKSKTc6Bvn69t3d0iw9zBtTdwtJtqwy2lMJ7IrAG2HA8SChmoeZ0YsXL4oiiNloxj0/V9mzZ5cbeGax/Xz77p5PCxcuLK3D2GaaXiB33XWX+MS0bdsWQHCWKyoqSjxemMlKDyrvKlWqBMDOerkzXe4sS6NGjSQGb7Q+8++Kexx79+4tSoP0PCLcmD5Ybm8uxhazZ1RjeZE3b17xIaEKgWNGxcmFCxdkbB988EF8+eWXvlZOxsbGWqmpqQHtM8PDwwE4XZzY5rhkyZLiJcB5kFky+roAzp6IiljGGUlJSQna79A/q1q1atKSNy210OTJk6UbH19b7dq1NRYNWrRoIfXn9J1gnMXHx4vS0d26kWMXFhYmnXy4h+HngWq88PBw8W7iPpRjZXZl4GeF+1d6ZFChBdiZ6u3bt+P8+fO+jcXk5GSrW7duorYBnI6D7q6Ww4YNC3jcLzF8+HAAjqKDalqus4CzR2JWvGrVquI5QmUCofJk7dq1ooBnxyHouhgQi2YsuL1g9u/fL90qSXpdTNyY3X/canQq8yZOnBjkj+cmLi5O/Cy5hvt5j1q+fHlrw4YN4i9n+kK6/T3y5s0rKh+OK1Xi7OL1xhtvSGcgt/qYKpAKFSpInPJeYNasWQCc9vSAo9KikpIKznXr1sl+i4qVYcOGqUJEURRFURRFURRFURTFixtSiGTKlMkqVaqUZDyaNGkiN6x0Zze7Xrgd8jdt2gTAqQlL73fz9mjw4MFeTuppwi4mrB/0ws83fomJiVa7du3EBXjmzJmioDH9CtKCLsLMbJk36+5OP3y+/fv3y3jSHdqrBjQtUlJSAnrJ/w9f377z5pYO2lOmTJG/s5aOKoFZs2ZJrFGtwa5BfP8TExNlbNPD7fxNOnfuLPW5HCtmW1k3WqRIEXmNJ0+exPvvv+/rLjPuDEqdOnWkrpLqqvXr1wOA1+ffEyq3mMFmnDdt2hSAXSvPzDezW5wrN2/eLF1l2PmESpUCBQoAAL799tsgRRl8HovR0dFWwYIFZYyKFi0qcyBjiqqeMWPGoGvXrgCCMxzPPvssAFtJSW+DtNQBlmVJXGXIkCHN12Z6egGOj8GaNWuwYsUKAHb99v+e0/exaGYl3esZVTf/+te/grqtEfeex4ReXRyvokWLim8X/WO4ZpYoUUKybOnBtZrxGRIS4utYjIiIsLJmzYrjx48DsNUbVHR44VY2M6NJJdakSZNkniReXfAiIyMBOKpKs4aesZ/e+sr1eNy4cahatSq++uor38cix+CHH36QTLF7vO68806sW7cu4Oejo6MBAIUKFQJgz3lUdlDFQSUlx437UsDxf6H3S5YsWcQvgZ0sOHdTAThw4EBRlBj4OhZDQ0Ot8PBw6V5Yr169oMdkzpwZgO1HQa8f/sn5j+uc2WGE502qXNktcdu2baJ2piqBaqDPP/9cVCNUa9HfhF4wM2bMkL1t9+7d0aRJE2zbts33sXg9JCcnSzcntwcTY/LixYvil0SvwsTERADAoUOHruv3UMHzzjvveH6/WbNm8pmjwvY///nPdcXibyqZWbVqlchA3QZgnTt3lg8mJ/tp06YBcEwBH330USm3oGxw5cqVAICzZ88CsOWmnPDYMskLvqlZs2YF4ASaaVa1fv16tGzZEt98841vP+BZs2a1qlWrJnJPwBkPyn5Jzpw55fDEcaaRGDeCffr0EZkpZadu80fTTMndSnDQoEHSwpBlHAwq0r59eyndoUngqlWrfL3YUOLNODE/5ywjY/u3SpUq4eTJkwAcEyku+mx9tXbtWjRq1AiAs5nn4Ykyuc2bN2PEiBEAgAYNGgBwFiSzXSUNqrhpMcsw3BsaPx/C3K0+ExMTZVGgNJiH1ZEjR4oc3y01NE05WdLEseTYmVLRtFqSr1y5UiSG7g2HFyxh7N+/v69jkeVrbNfYvHlzOUQ98sgjAOAp6+alF2PYjCH3ZQlhyVm/fv0k5nm5zOTE7t27g0pIjbFK8//h51i8no1fVFQUAHszba6fv4S79MaE7a23b98OAAHlpzTA5meBZas87Ldq1Uouzmh63rVrV1/HYrly5awvv/xSLna9DBVNE3n3esQx4pglJSVJq0ke0PkzjLsRI0ZIGQwP3O6SAC84txcuXBgTJ04EYM/lO3fuxIULFzQWPeBBlnPdpEmT5L3jGsc9KmOpdOnSsv/hJQfL9/lZCAkJkYsrzsOcu0+fPi0laSwn/8c//gHAKTcvVaqUXFrztR05csTXsegeR/McwIQN18wnnnhCyhx4yeROBqSmpkpS3W3OyvLsTZs2SZKPeyk2HEhJSZFL7bRKikNDQ2XPNG/ePLRq1Qrbt2/3fSzy3MA4ApzLfybPypUrJ0030jIbPnnypIgmWOZUrFgxAI6xdevWraVMiskDlsyNGzdOfo5/snyYFyyTJ08W0QX3XVoyoyiKoiiKoiiKoiiKkga/2VSVEkPe3E6YMCHNn+fNLTPXOXLkkDa7fB7e4PKWbsyYMZLtdFOnTh3J1BgtIAE4krh58+bJDWO3bt0wY8YMHD161Lc3frfddpvVunVryWS8++67or5o2LAhAOf2PGPGjNLm1m0KSAkhlTgmNB/j8wFOFpsGPGZZDjNvzJjSKI7KE/M18SYYPpcjMhYZU1WqVJH2U1To8GacGUUgWCro1V7VfVNvkpaKZ9q0aaIEYztPtuSmIRLgxGexYsWwZs0aXxs5cgzN7D3VNVTn0BTz2LFjkvGgDJCw5M0wHA6C4z1q1ChREHm10Ha3gKUqj5J802DXQGPRoHXr1pLtoBkZx/PLL79Ms5VxWrFlQnO/bdu2SZaMJaxeMIvDjLWpQmGJW6ZMmXD8+HH8/PPPvo9FYhpRU3WaMWNGAI6q7pdwtzhnnFMZUK1aNZl/qSiiQja9slVmtXfv3i1/p1EdfB6LYWFhVkxMjOwzAGdOo2k7Y3Pv3r2S4XdDRQAVkelhlm1wL0VDXABiFklVLJUI5rrozsD6Wa2VI0cOq379+rIvHD16tJgp8j3jGNasWVN+ziwHBNJvieulEEmL2rVrS3kU45VxyrX30KFD8jljxnzHjh2+jkXOqeYek2Pibs/qVTrvjkHTlJVnBaqDKlasCACezRvM38n1j2VvLFnkHP/QQw/h22+/BQC8+eab6NSpE3bu3OnbWHSvi0WKFJESUpr68yxpWZacu1mCxu+ZUKXK8kIqH7mupqamSsxSsed1TmGJMdvwnj59GoBTmeBCFSKKoiiKoiiKoiiKoihe3KhC5DiA/b/fy7kl5LEsK+GPfhF/FH+TMQR0HP8O46hj+NcfQ0DH8e8wjjqGf/0xBHQc/w7jqGP41x9DQMfx7zCOOoZ//TEErnMcb+hCRFEURVEURVEURVEU5e+AlswoiqIoiqIoiqIoiuI79EJEURRFURRFURRFURTfoRciiqIoiqIoiqIoiqL4Dr0QURRFURRFURRFURTFd+iFiKIoiqIoiqIoiqIovkMvRBRFURRFURRFURRF8R16IaIoiqIoiqIoiqIoiu/QCxFFURRFURRFURRFUXxH+I08OCQkxPq9XsitxLKskD/6NfxR/F3GEMAJy7IS/ugX8UfxdxlHjcW/BRqLfwM0Fv8WaCz+DdBY/Fugsfg3QGPxb8F1xaIqRJS/Kvv/6BegKAoAjUVF+bOgsagofw40FhXlz8F1xaJeiCiKoiiKoiiKoiiK4jtuqGRGUUJCbPVYeLj90bEsC5Zlq6quXbsmX1MU5c9HSEiIxDDjVflj4DgQnTcVRVEURVFuPXoholwXoaG2mKhEiRIAgJ49ewIAChQogFOnTgEAfv75ZwDAzp07AQArVqwAAKxevRo//vgjAN30/5UJCwsLOkxzPHVc/5xER0cDsOMUAJKSknDw4EEATpxevnz5j3lxPoexxD/NWNJ4UhRFURRFuTVoyYyiKIqiKIqiKIqiKL5DFSLKdZEjRw4AwOTJkwEAKSkpAAJl3+5M5zPPPAMA2LhxI5566ikAwK5duwAAV65cCXis8ueD4xkREQEAKF68OAoXLgwAWLduHQDg+++/BwBcvXpVx/JPRJ48eQAAY8eOBQDkzp0bgD1uQ4YMAWCPmXLrcStDzPJDwFZfcWyuJ6b4PFTxhYWFiepHY/LPS2RkpPzdvR7quP0xMJY4NhEREbh06RKAX79nyZAhAwBHrUe1LJ9PuXVwjuSfOgZ/TRinUVFRAID4+HgAwB133IFs2bIBAM6cOQMAOH36NADgwIEDsq4yvn/44QcAwIULFwDo58HvqEJEURRFURRFURRFURTfoQoRJV14k/7www8DAIoWLRrw/XPnzknG4/jx4wCATJkyAQBy5swJAEhOThZlwYEDB+TnlD83vIXneD711FMoVqwYAGDv3r0AgMOHDwMIVBu4lULKrSEkJARZs2YFAIwbNw4AULlyZQBObC5cuFBiUE1V/1g4tzJblTlzZgD2uJw8eRKAE1fpjRXjjWOfJ08efPPNNwAg2W1VHvx5yJgxIwCgRYsWMmazZs0CABw5cgQAcPHiRY3PWwgzzfRIq169OgB7PFauXCl/B4Dz588DuL5Yypo1K5588kkAjlJk1KhRAJwMtvL7EhoaiixZsgBwlM5870+cOAFA18I/I1zXwsLCANjxExcXBwCoWrUqAKBx48YAgEKFCgEAsmXLJnFJZTO9DS3LEjUmzx/799vdWKma/eyzz1Ql4mNUIaIoiqIoiqIoiqIoiu/4XRUizICl1ZlC+fPDsbv99tsBODfq27dvB2Bnonfs2AHAqcd76KGHAADDhw8HAGTJkkVu6N0ZS+XPC8eI2bNKlSpJluvs2bMAoF4FfwIYo7GxsejWrRsAIDU1NeAxn3/+OQBg1apVmgH5k8D1MTY2FgBQunRp+fry5csBAD/99BOA9OOM458rVy4AQP369aVumh4/2kno1pCeOo7jzdh8+eWXg7wM3nvvPQC2QkT5/WHscd5s0qQJAMfXZ+zYsbJvvREPEWa177//frRp0wYAsHbtWgAai7cKerY0atQIXbp0AQDpiPjWW28BAJYuXQpAFVl/Jjgnct9JP7S6devikUceAeB4GFK9bI4d45NzMc8eoaGh8jV6jvBc06FDBwCBHTGVWwPnWvJH+hHe9AsRfpjj4uJExsRyCf4njx07BsCenLhx40TlZS7GwHBvNi5cuCCPvxETOuX64fv5ySefAACWLVsGANi0aRMAe8PulqhxkomJiQFgbwAOHToEQE2L/kow3hITEwHY4/rtt98CcEpldBPxx+Een969e6Nt27YAnA05L654sE5ISJANOb+n5qq3Fo4bN+x33HEHAKBChQoAgG+++SbAYNXEvdkzv8b5tkyZMsiXLx8AR+avJta3lpCQkKCxYqlMjx49AADZs2cXOXfBggUBOHGr/P6Eh4ejefPmACCXFhyradOmAQCmTJkiB6TrWes4fpyTBw4ciLx58wIANm/efN3Po/x6OA+++uqrAOxSXx66eM547LHHADj7mF27dsk463r4x8DYYQnpbbfdBsAuLQTsCxF+jWsnY8k8/7mTCKZ5Nc+nPKvwMUzm8vvK7wdjkeWiLH9isvXAgQNyR+Aey/Pnz8uayeT6zRRa6OgriqIoiqIoiqIoiuI7bppChLc+lDINGTIEZcqUAeBkRngD5C6lAYJveczbHn6Pf1JVMmvWLEycOBEAsGfPHgCBBjrKb4fvOU36iJeJZkJCAgCgadOmAJyb2TNnzki7XR2Xvw6M6UcffRSAnXnZsGEDAKdNmRc6xrcGSkEHDhwIAGjQoIHMtZwHOYYPPvggAFuuz3mYqq/x48cDAA4ePKgKrluAW7ZbrVo1AE6GbNu2bWLcmFb2w/w319Ny5coBsNUGLJ/RMtVfz281h3a3V6bJOI07Q0JCZJwXLVoEwDF7VBXB7wfHI1u2bGjVqhUAZ2/6+uuvAwDeeOMNAN6lS+a+lf9mxpnKLKoTChcuLHulxYsXA1CV7O8FleTt2rUDADGzjYqKCjoXVKpUCYBjrjl27FgtlfgDCQ0NlfIXzo9UTJYtWxaAfZ7geHGe5D7ULGuj4f+KFSsAAF9//TUAWw3En2drXpaprl69GoA2e7jZcG9C1VbZsmXlfMh9T3JyMoDA0hm3Uofr5P79+7Fw4UIAjoqPqvWbcfZXhYiiKIqiKIqiKIqiKL7jpilEqAagb8i5c+fEePPo0aMAnFsi/pkjRw6ptWRNGLNk/PPnn3+W2z9mP1lH9uijj2L27NkA1EPk9yatrEZISIiMC2/m8+fPL98DgA0bNshnQMfnz4/bjKpBgwYA7FvbDz74AIBmMP8I3G2Q6UVQv359APa8ynmQt+XMcHJ+LliwoMy1/PmGDRsCAGrWrImdO3f+7v8Pv8O17a677gLgZCvnz58PANi3b9911bHz88DsdKlSpQDYmRaOoyomfzs3SylCnxAqg0JDQ7F7924AEOUda6aV3w9mIitUqCB7FZrFc33jvGmOOcfRbfqYMWNGFChQAADQuXNnAMC9994LwN6XLlmyBADw4YcfAlBT1ZsNx4Vnj969ewNwxvn06dNi/M95lbHIcVq6dKmozJVbB8cuJiZGFAM9e/YE4Ow/6T+4YMEC8TDct28fAMeHgsqOK1eu4ODBgwCcSgJTJcnfR48KPg+VJuofc3PgHHnPPfcAAN58800AtoLObaJqeoICdtxSsef2p8yVKxfq1asHwNkHs3kHx10VIoqiKIqiKIqiKIqiKDfATVOI0PH1X//6FwDgiy++kFv2tOqhQ0JC5LaIN0DMZJqtenlTOGLECACOp8GRI0ckw6J1mX8MYWFhUuPXunVrAM5YcvxfeeUV+Xwof34Ye8x68c8zZ85IPaZmnG89vDVv1KgRACcbmTlzZgB2doOZzvXr1wNw5uOTJ08CsLNoderUAQAUK1YMAJA7d24AQK9evaTuWhVAvx/MLLP9ao4cOQA4tc7p+fOYMAtDb4qSJUsCsFvtcl3Ucfx1hISE/CplSHqPvfPOOwE443/58mVMmjQJgJOx1Hn11pGcnCwxxAwzYy+9caDCrnr16gCAWrVqiRcB51I+5vDhw3jhhRcAaCvl3wuOIZWsPC/Qt+7FF18U9cf9998vXwMcNXPbtm1lrdSzxK2DZ4Vy5cqhT58+AJyupNu2bQMAjBkzBoB9pvyl+Pyl+ZPfpxKEca/r5M2B53nuUUePHg3A7joL2PcEVK9SOUf/LHbEy5Qpk+xpihQpAsDpxFeoUCHxaapYsSIAu1sbYO97fiuqEFEURVEURVEURVEUxXfcNIUIb9zo1GxmuXjj6nV7x59LT0HA59y4cSMA4JFHHgFg3+7RbVi5tTB7lj9/fkyZMgWA4xbMMWU97saNGzXz9ReCY8vOJMxofvjhh6L2Um4tYWFhcjPesWNHAE79LMfr7NmzEnO8mad3j1kbO3LkSABA165dAQDPPvssANv1m/5MrNvVuL25hISESPajRo0aAJy1b926dQCuL1sVEhIiiiHW1HL+nTNnjmajfyPm5/63xgDjk9lpZrRPnz4tjvmalb51ML727dsnf7/99tsBONlperpcvXpVxo/qZSomGXeVK1eWrk78rJw6dQqA3W2GGVGdS38fmJWmHwjj6+OPPwYA/Pvf/xbfFo4hvXroQ5A/f34kJSUBgHQp0fH6/WBMMW66du0qsUcvCO5z6EV5M/091Cvk5hMSEoLixYsDAF577TUAjlqLCsjx48djzpw5ABxFiJfqh54/VM1SyVOxYkXx4CL0ZLsZ3LQLEWKWw7jbk/1a3OaOfAPmzJmjG4k/CG7GR4wYITJRjv2aNWsAQKSiOkZ/LXgB8vDDDwNwFo/58+ertPAPIiIiQsxTuXHgxo+H3+nTp2PQoEEAkG4LQV5qvfPOOwCAZ555BgCQmJiIli1bAnCMqtQA8OYSFhaGhx56CICzrk2dOhVA+mPmJiQkRC7IqlatCsA5hM2YMUPH7SZwsw5EnE9TUlICvr5t2zYZM+XWwfVs/fr1cvFLQ85Ro0YBAF566SUAwNatW2Wepez7vvvuC/h3hgwZZIw5nmzbO336dI3F3xmWXfAyn9Aw89KlSzLmhw8fBuBceLHsqUCBAnj55ZcBOIkCJlv1YuTmw5ji+1+pUiW5pOrfvz+A3+ciRPn9CAsLQ9u2bQEA/9/eucf6XP9x/HHO6eBQP5KOMVvYUJlUSEuMVkm5dWFJStbWVZZSNlo1pTXrotusobms1dGYbGXSRJGzQ5IiUi4hS+5x2DnO9/fHd8/355zvQQff7/l+j/N8/PPt6Nz2fZ335fN6PV/P16WXXgpEdxrtq/Pnz+fQoUNAlBTT83x54YT+TXcctXK3bt06/O0ocbl9+/bwdeeKW2aMMcYYY4wxxhhT60iZQgSiDNC5jK278MILef/99wGCcZUkiLNnz3b2tppRLAcPHgzAzTffHP7fmjVrgGiMp6tfNRNld2W6KUVBYWGh11uaaNiwYahMqhqpkaoyp3r55ZfPSGWgjLzkiI0aNQrtiB9++CEQrWHHPTnk5eUFMzBJRj/77DOg6q0yEFeXSDEkU92lS5cCsG7dOscrg9B+KkWQWLBggdWTaWTfvn2hojlz5kwgMtlURXPz5s2sX78eiKTdanHT+q1fv374nj/++CMAM2bMAHCLaTWgfVNqH7WSNmjQAIgreLTOFCtJ8nWm1q9fP5x9hYWFQKSg9ECA5KMza8iQIUC87enLL78E4i1OkBqj03MdoW5OTU5ODj169AAiBZBUPx06dADiKq6DBw8CUfuaDOD1Wq9evbAuH3zwQSAyTi0rKwt77PPPPw+QVNsMK0SMMcYYY4wxxhhT60i6QkSUlZUFs6OzUYqoh2jYsGFBjaCsk3relX0y1YdGRL7yyitAPJsnD4Px48cDHiFYU9H6lNmjKppLliwBYM+ePen5xQx5eXmVjPs0Zuzpp58Gqq7IUpwTx77m5OSEyo1UKCa5NGrUKLzfqoCpp/Z0KGbql2/evDndu3ev8G9z584FXNHMFBSz/v37A5H5o+Izb948n5FpJBaLBSPjrl27AtHZd+211wLxfVfKEI2HlA+FTHIHDhwY7qtTp04FIjNrxzf1SCm5fPlyIPLqGTFiBBCvLutZQX538huRyicnJycYrA4fPhyAOXPmAFEsTfKQMaYMinNzc8OdQ2oAxexkQzn0LCh0lv6XqsTrMXXUrVs3qI31qrtJz549gbgqRDFKzA+U/z7yqFSctQevXr06DAWQd0ji38K5YIWIMcYYY4wxxhhjah0pVYgkZvbOJDun6vTo0aNDBkgVMPWYmepDFZCxY8cChBFlAF9//TUA33//PZCa3j+TehRjVTTlji+PA7t9p4969epV8iDYtGkTEFW5qrq/qhI2ZswYIOqrjsVioY/Tve/JpXwVRFUTTam44447gKgXdt++feHs1NcpRlKX3HDDDcGB/ffffweiyrWrYJmBKp4DBgwAonPxl19+AaJ1a9KP+to1ElLjkHNzc8NaVMValU1VuRs0aBBiuXDhQsBnZXWi93rx4sUADB06FICWLVsCcUW51p4qzdu2bQOiUeddunQJ+6lUCxpjboVI8tF60x2zQYMG9OvXD4iUq2vXrgWi9z8vLy+sRcVWa1Fn39SpU4OX4T///AN4LVYXxcXFvPDCC0Ck0pJ/lnx9mjRpwpYtWwBo2rQpEHmDav3VrVs33Ht0Dy0qKgLi8ZX/j+5R8iLR39K5PH9aIWKMMcYYY4wxxphaR8oUIrFYLGRszqRipayP+v9atmzJ3r17ARg3bhzgjF86ULVLWVxl8A4fPsyECROAqG/M1Ey09tq1awdE2Vl5iFR1HUvRdTbKMHNqtL6kFNH0gzOZVFG/fn0ee+wxADp16gREa/nff/8NKjxV0hy75KD3ce/evSxbtgyI9tKBAwcCUaXkyJEj4cxTzOVxoH75/Pz8oBqRd4wUJo5Z+snKyqJFixZA5LCve8vnn38OnNm6NdWDqovyRdMrVPbxGTZsGBCvUmvfrIofkEku2u9WrlwJwOuvvw5ECsimTZuGz5FqQAqgr776CoC+ffvy1FNPAZGC8tlnnwXiihM/cySX/fv3A5Evz4gRI8L7LmWO9s/y3pOKY6Ivpejfvz9//vknAM888wwAixYtAuIKL6vXU0dpaSnfffcdAN9++y0Qqc61Z5aWlobnA/n5aN3puQMqK/Y09evvv/8OSj19b8U0GfeelCVE4Mx+Qf1ht2/fHoCRI0cC8T9iGVXJrFPjtEpLS4NBmS+BqeW6664Dok1KeljFegAAC/9JREFUB0RBQQHr1q0D3CpT05FUVDGWaVFVWzKSMWbbnJzi4uKQoNLak8mqDobTPWBpzxw8eHDYW2WgKnbv3s0PP/xQ4WeY5HLs2DHeeustIBrRefnllwPRHtu7d+8gAxW6UCjpUVxcHJLUkqUm01zMnBvZ2dncd999QJTAVJLrk08+AXxe1jR0rimRfOWVVwLx5PH06dMB75vpRMkrjYxXC3efPn3CGblq1SogGrurB/OsrCzuueceANq2bQtAr169gPj61eeZ5KC7yksvvQTE2wiVgFI7vs688uabia2kia8XXHBBeNB+5JFHANiwYQMAW7ZsOasivakasVis0h1U+6EKOVC58HrXXXcBUVvw0aNHmTJlCgCvvvoqUHGAimKt+8/JTHfPFt+gjDHGGGOMMcYYU+tIqUIkUc4kymdy9DkyWJk2bVqFj3/66SdmzZpV4euUwW3cuHFQJ6j6YhlqclGlUvIzfaz3uaCg4JyzrqfK9pZvuXBGN7VkZ2fTp08fgDDySpn18rLh02GFSOpQlQSiNajRkDLa3LlzZ3jPpRZo2LAhQKhWjxo1iubNm1f4HKnsFi1aFAzMXL1ODWVlZaHaOH/+fCCSDSuOgwYNonPnzkBULZOaR0ZzsViM22+/HYjMWa0QyRwuuuii0Aql/VDVaZup1kxUkZSxvD4uKioKxsYm/agqLQWenhEg2iMTZfbr1q1j5syZALz44otA/PkC4vuxVOo+F5OLWiM++ugjFixYAESto3oGVGtpp06dwl6qOEo1q/tRs2bNwrqUkkutU++9915Yp2oJtqKresnOzg5KdLXB6N6jtbVw4cKgDFGcTkZVn0vO6PdL+nc0xhhjjDHGGGOMyXBSaqqaqBBJrCDn5OSELOwbb7wBRON6duzYAcS9RDZv3gxEGSSNm8zPzw8VUGWLDh8+HH6+OXeUbVWc9L6qqtylS5dK4z8TVTrZ2dkVYg6RcVV+fn7ICF9yySVAVNVWZn/NmjXs3LkTiEYrmeRSt27dUHFWrFS5rqrqSll6Z92Tz5EjR8Lepvf5sssuA6K989NPPw0Gm7feeisAPXv2BOCKK64A4utO1RWtJRnLTZo0yaaA1YD2UK0rvcoMbvLkyUGlJe8X9d0qZjfddFPoeZeHiM5An33pQ3vnVVddRZs2bYDozNN4+lRUtkzqUEw7duwIwI033ghE59zYsWMr9MibzKL8feRUd5Pi4mLmzZsHwMMPPwxEpo+jRo3iiy++AGDXrl2AlSLJ5sSJE8EjUiao2jdXrFgBxFXqUk7qDqS9VM8Oubm5lcZiDxo0CIifpZMnTwYIXmnJNOQ0/03Dhg3DOtPoZN1/CgsLgfgzf1XOyFTEzAoRY4wxxhhjjDHG1DrS4iFSfnyZxgnecsstQOQm++STTwKwevXqSlldVcmOHj0a+un1Oc70JY+srKygENG4MmXz9L6PGTOGJ554AoimAGk6ibLpF198cVCYtG7dGogyunXq1AnVT8VOWVt9v+nTp/Pmm28CVogkG63FVq1aBdd8qX/k0n6yakiiT0h2dnal+JnksXfvXr755hsArr/+egDy8vIAuPvuuwG48847K6nwTuYroTWkjLx64nfs2OHYpZHyyhGN29VrosKutLQ0qOyk6tG6NelD6+3qq68OZ6TiokpnMl3xTerR/WTo0KEVPi4qKgLi3jCJ40CFY1wzKCsrY+vWrQDMmDEDiMbuNm/enOHDhwPwzjvvAFaip4LE9zJRObl27dpw99HzhNai7kLZ2dnhDqNnQvmUrF27lt9++w3A02aqGZ2Ljz76aFBOiuXLlwPxCYgAe/bsqd5frhwpTYic6nKtN6dOnToMGDAAiP6gdenXHOPTye/Lysp8CUwxev8XL14MQIcOHYB4kgPirRYy/NMYUH1O+fif6qJQWloajHN0sddDgEyxCgoKLDNOEXrA6tevX4ij2tW2bNkCnP7QKJ8YcVIydZSUlPDBBx8AhNYmtZpJInqy5Ed5Y2KIJ5F1kX/88ccB2LhxY4XPMZlHYrKxdevW4d9kFGfZfvrRftqkSZNK60kJklMVikz181+xyMrKCi28uqP+8ccfALz22mtA/A6jBzO1uuksLCkpcRGnhqD98+OPPwaicaBt2rShd+/eAMyZMweI7qg+M1OP1o/uKRC1kmq/Ffv37w/n4bJly4CoJbioqMhxSxMyhx89enSImdq7R44cCURF93TilhljjDHGGGOMMcbUOlKqEPkvWrRoQbdu3YAoC1hQUABU3XjMmb7UEYvFwlir2bNnA5ExnCRrrVq1YsiQIUA0BlSVFFVNIBqPpSyg1B/z5s0L/y2plNqmVA0tKSmxlD9FyHjq3nvvDfGSgVhVKlvl15/XYmrRSFy1Fz700EMAPPDAA0A8llp7kvSqWrJ+/Xogbry6evVqIJLyO241B6mA2rVrF/bEn3/+GXCrWiagPbROnTohVokmul5vmUNii2Eiubm5oZVUbb5qVevVqxcQb2e87bbbgKhNWGeozOBN5qN1qRjKgHPixIlhBGyPHj0qfI7OWZM6dK6tWLEiqAnUviZjTo2yLywsDCoe/ZuUPzb7r36kXpatQuPGjUM83333XSCzFMpWiBhjjDHGGGOMMabWUa0KEWXhZdTZvXt38vPzgUhBIO8QV7syA2VX9aqsq2K5fPny0HOp6vT//ve/Cq+xWIwDBw4AUe+lqtMnTpzIiMxgbUPxUy9maWlpGHM8ZcoU4PQZdccsfWgtvf322wDBWyQvLy9UL1WNVgVLqitXSWom5Y3IIV6l1pmp/ddnZvpRnIqKikKPtF63bdsGeO/MJBQv+buooqm7TPv27cPYTo3blX/a/fffD8Btt90W1JTTp08HIiNH77c1B61L3XV1H9q+fTstWrQAokr3mjVrgMrjW03qOH78eDDgXLVqFRCtRa3fAwcOhGcMj9RNH9pX27VrB0SjrLOzs8P9ddasWUBmDcqwQsQYY4wxxhhjjDG1jrR4iEgh0rlz51Dxkn+EvEMSx3qazKJ8XFQFSRwVqT5Lk3kofvKlmDhxYojXpk2bAFc9Mh3FUGqr48ePh+y7OT/Reblr1y5WrlwJEEYJ+qxMP7q/LFq0iHHjxgGRWktr03HKHHR3kYJOvi9SCWzcuJFp06YBke9Zx44dgciXaePGjWFkqya02Z+p5qJ7j7wNJkyYQN++fYEo9l27dgXio1zLf41JLVpPWq9//fUXUHHCXuJkPVP9aNrW+PHjAWjUqBEQ32819ScTnw+tEDHGGGOMMcYYY0yto1oVIsriyTekW7duoVdT2Tx9rF7OkpISZ/qMSRHKtC9ZsoSjR48C7ns2JlPRubh06VIKCwuByNfJpJ/yaklNZlPMtNeazEN3TMVPr7t27QqVTPlFJKqXfT89v1A8dR9asWJF8OKS34GmJTr2mYEVOpmB9sa2bdsCcM011wDRutm6dSvPPfccEKnoMolqTYho8zh06BAQNx5TkkTjdiUr1YHkDceY1KF1dvDgQa81YzIUrU21Iy5btixcMjLJlMzEicViFVrZTM3HCZDahR6yi4uL+fXXXwGYO3cuAJs3bwZcPDKmPHqeV2vvhg0bgMhYfNKkSezevRvIzH3ULTPGGGOMMcYYY4ypdWSdSZYmKysrKSkdZY+aNWsWMkqSoEmmlkoJVCwWy0rZN89wkhXDDGB1LBbrnO5fIl2cL3H0Wjwv8Fo8D/BaPC/wWjwP8FrMLNQKoOcVPTf9x3OK1+J5gNfimaM2UZmrap0cO3YsXe1NVVqLVogYY4wxxhhjjDGm1nGmHiL/ANvO9Yeq53n79u3n+q3OhsvS8UMziKTEMANwHGt+HB3Dmh9DcBzPhzg6hjU/huA4ng9xdAwzLIaJprtVxHHMsDieBY7hWcRQo+bleZYBVCmOZ9QyY4wxxhhjjDHGGHM+4JYZY4wxxhhjjDHG1DqcEDHGGGOMMcYYY0ytwwkRY4wxxhhjjDHG1DqcEDHGGGOMMcYYY0ytwwkRY4wxxhhjjDHG1DqcEDHGGGOMMcYYY0ytwwkRY4wxxhhjjDHG1DqcEDHGGGOMMcYYY0ytwwkRY4wxxhhjjDHG1Dr+DyRpAZRr64yxAAAAAElFTkSuQmCC\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "import warnings \r\n", + "warnings.filterwarnings('ignore')\r\n", + "\r\n", + "import matplotlib.pyplot as plt\r\n", + "%matplotlib inline\r\n", + "\r\n", + "n = 10 \r\n", + "\r\n", + "x_data = x_data.numpy()\r\n", + "predicts = predicts.numpy()\r\n", + "\r\n", + "plt.figure(figsize=(20,2)) \r\n", + "for i in range(n): \r\n", + " ax = plt.subplot(2,n,i+1) \r\n", + " plt.imshow(x_data[i][0].reshape(28,28)) \r\n", + " plt.gray() \r\n", + " ax.get_xaxis().set_visible(False) \r\n", + " ax.get_yaxis().set_visible(False)\r\n", + "\r\n", + "for i in range(n): \r\n", + " ax = plt.subplot(2,n,10+i+1) \r\n", + " plt.imshow(predicts[i].reshape(28,28)) \r\n", + " plt.gray() \r\n", + " ax.get_xaxis().set_visible(False) \r\n", + " ax.get_yaxis().set_visible(False)\r\n", + "\r\n", + "plt.show()\r\n", + "\r\n" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "请点击[此处](https://ai.baidu.com/docs#/AIStudio_Project_Notebook/a38e5576)查看本环境基本用法.
\n", + "Please click [here ](https://ai.baidu.com/docs#/AIStudio_Project_Notebook/a38e5576) for more detailed instructions. " + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "py35-paddle1.2.0" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.7.4" + } + }, + "nbformat": 4, + "nbformat_minor": 4 +} diff --git a/docs/practices/jit/Semantic segmentation-jit.ipynb b/docs/practices/jit/Semantic segmentation-jit.ipynb new file mode 100644 index 00000000000..2617087bfd3 --- /dev/null +++ b/docs/practices/jit/Semantic segmentation-jit.ipynb @@ -0,0 +1,1040 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Paddle 运行UNet语义分割-动态图转静态图" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T06:45:14.265722Z", + "iopub.status.busy": "2022-05-04T06:45:14.265452Z", + "iopub.status.idle": "2022-05-04T06:45:16.643908Z", + "shell.execute_reply": "2022-05-04T06:45:16.642804Z", + "shell.execute_reply.started": "2022-05-04T06:45:14.265696Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "# 安装paddleseg\n", + "! pip install -q paddleseg\n", + "# 解压数据集\n", + "#! mkdir /home/aistudio/DataSet\n", + "#! unzip -q /home/aistudio/data/data101908/IAILDdataset.zip -d DataSet" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 一、数据集预处理，及读取数据集" + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T07:01:02.515007Z", + "iopub.status.busy": "2022-05-04T07:01:02.513931Z", + "iopub.status.idle": "2022-05-04T07:01:04.369405Z", + "shell.execute_reply": "2022-05-04T07:01:04.368138Z", + "shell.execute_reply.started": "2022-05-04T07:01:02.514952Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "数据列表生成完成\n", + "(array([[[-0.04313726, 0.03529412, 0.01176471, ..., -0.04313726,\n", + " -0.05098039, -0.07450981],\n", + " [-0.02745098, 0.01960784, -0.01960784, ..., -0.01960784,\n", + " -0.01960784, -0.03529412],\n", + " [-0.00392157, 0.03529412, -0.04313726, ..., -0.01960784,\n", + " -0.02745098, -0.05882353],\n", + " ...,\n", + " [-0.6 , -0.6 , -0.5921569 , ..., -0.01960784,\n", + " -0.12941177, -0.22352941],\n", + " [-0.6 , -0.6 , -0.6 , ..., -0.21568628,\n", + " -0.23921569, -0.3647059 ],\n", + " [-0.6 , -0.6 , -0.6 , ..., -0.42745098,\n", + " -0.4745098 , -0.5058824 ]],\n", + "\n", + " [[-0.04313726, 0.03529412, 0.01176471, ..., -0.01176471,\n", + " -0.01960784, -0.04313726],\n", + " [-0.02745098, 0.01960784, -0.01960784, ..., 0.01176471,\n", + " 0.01176471, -0.00392157],\n", + " [-0.00392157, 0.03529412, -0.04313726, ..., 0.01176471,\n", + " 0.00392157, -0.02745098],\n", + " ...,\n", + " [-0.5137255 , -0.5137255 , -0.5058824 , ..., -0.07450981,\n", + " -0.18431373, -0.2784314 ],\n", + " [-0.5137255 , -0.5137255 , -0.5137255 , ..., -0.22352941,\n", + " -0.24705882, -0.37254903],\n", + " [-0.5137255 , -0.5137255 , -0.5137255 , ..., -0.4117647 ,\n", + " -0.45882353, -0.49019608]],\n", + "\n", + " [[-0.04313726, 0.03529412, 0.01176471, ..., -0.01960784,\n", + " -0.02745098, -0.05098039],\n", + " [-0.02745098, 0.01960784, -0.01960784, ..., 0.00392157,\n", + " 0.00392157, -0.01176471],\n", + " [-0.00392157, 0.03529412, -0.04313726, ..., 0.00392157,\n", + " -0.00392157, -0.03529412],\n", + " ...,\n", + " [-0.49803922, -0.49803922, -0.49019608, ..., -0.14509805,\n", + " -0.2627451 , -0.34901962],\n", + " [-0.49803922, -0.49803922, -0.49803922, ..., -0.28627452,\n", + " -0.30980393, -0.43529412],\n", + " [-0.49803922, -0.49803922, -0.49803922, ..., -0.46666667,\n", + " -0.5137255 , -0.54509807]]], dtype=float32), array([[0, 0, 0, ..., 0, 0, 0],\n", + " [0, 0, 0, ..., 0, 0, 0],\n", + " [0, 0, 0, ..., 0, 0, 0],\n", + " ...,\n", + " [0, 0, 0, ..., 0, 0, 0],\n", + " [0, 0, 0, ..., 0, 0, 0],\n", + " [0, 0, 0, ..., 0, 0, 0]]))\n" + ] + } + ], + "source": [ + "import os\n", + "import io\n", + "import numpy as np\n", + "import matplotlib.pyplot as plt\n", + "from PIL import Image\n", + "import paddle\n", + "from paddle.nn import functional as F\n", + "import random\n", + "from paddle.io import Dataset\n", + "from visualdl import LogWriter\n", + "from paddle.vision.transforms import transforms as T\n", + "import warnings\n", + "warnings.filterwarnings(\"ignore\")\n", + "\n", + "import os\n", + "import random\n", + "from PIL import Image\n", + "import matplotlib.pyplot as plt\n", + "\n", + "def create_list(data_path):\n", + " image_path = os.path.join(data_path, 'image')\n", + " label_path = os.path.join(data_path, 'label')\n", + " data_names = os.listdir(image_path)\n", + " random.shuffle(data_names) # 打乱数据\n", + " with open(os.path.join(data_path, 'train_list.txt'), 'w') as tf:\n", + " with open(os.path.join(data_path, 'val_list.txt'), 'w') as vf:\n", + " for idx, data_name in enumerate(data_names):\n", + " img = os.path.join('DataSet/image', data_name)\n", + " lab = os.path.join('DataSet/label', data_name.replace('jpg', 'png'))\n", + " if idx % 9 == 0: # 90%的作为训练集\n", + " vf.write(img + ' ' + lab + '\\n')\n", + " else:\n", + " tf.write(img + ' ' + lab + '\\n')\n", + " print('数据列表生成完成')\n", + "\n", + "data_path = 'DataSet'\n", + "create_list(data_path) # 生成数据列表\n", + "\n", + "import random\n", + "import os\n", + "import io\n", + "from paddle.io import Dataset\n", + "from paddle.vision.transforms import transforms as T\n", + "from PIL import Image as PilImage\n", + "import cv2\n", + "import numpy as np\n", + "import paddle\n", + "class dataset(paddle.io.Dataset):\n", + " def __init__(self,mode):\n", + " self.train_path = 'DataSet/train_list.txt'\n", + " self.val_path='DataSet/val_list.txt'\n", + " self.file_list = list()\n", + " self.mode = mode\n", + " self.image_size=(128,128)\n", + " if mode == 'train':\n", + " file_path = self.train_path\n", + " elif mode == 'val':\n", + " file_path = self.val_path\n", + " with open(file_path, 'r') as f:\n", + " for line in f:\n", + " items = line.strip().split(' ')\n", + " image_path = items[0]\n", + " label_path = items[1]\n", + " self.file_list.append([image_path, label_path])\n", + " \n", + " def _load_img(self, path, color_mode='rgb', transforms=[]):\n", + " \n", + " with open(path, 'rb') as f:\n", + " img = PilImage.open(io.BytesIO(f.read()))\n", + " if color_mode == 'grayscale':\n", + " img = cv2.imread(f.name, cv2.IMREAD_GRAYSCALE)\n", + " img = img.clip(max=1)\n", + " img=Image.fromarray(cv2.cvtColor(img,cv2.IMREAD_GRAYSCALE))\n", + " # w=img.size[0]\n", + " # h=img.size[1]\n", + " # for x in range(w):\n", + " # for y in range(h):\n", + " # L=img.getpixel((x,y))\n", + " # if(L==255):\n", + " # img.putpixel((x,y),(1)) \n", + " elif color_mode == 'rgba':\n", + " if img.mode != 'RGBA':\n", + " img = img.convert('RGBA')\n", + " elif color_mode == 'rgb':\n", + " if img.mode != 'RGB':\n", + " img = img.convert('RGB')\n", + " else:\n", + " raise ValueError('color_mode must be \"grayscale\", \"rgb\", or \"rgba\"')\n", + "\n", + " return T.Compose([\n", + " T.Resize(self.image_size)\n", + " ] + transforms)(img)\n", + " \n", + " def __getitem__(self, idx):\n", + " image_path, label_path = self.file_list[idx]\n", + " train_image = self._load_img(image_path,\n", + " transforms=[\n", + " T.Transpose(),\n", + " T.Normalize(mean=127.5, std=127.5)\n", + " ])\n", + " label_image = self._load_img(label_path,\n", + " color_mode='grayscale',\n", + " transforms=[T.Grayscale()]) # 加载Label图像\n", + " \n", + " \n", + " \n", + "\n", + " train_image = np.array(train_image, dtype='float32')\n", + " label_image = np.array(label_image, dtype='int64') \n", + " return train_image, label_image\n", + "\n", + " def __len__(self):\n", + " return len(self.file_list)\n", + "\n", + "train_dataset = dataset(mode='train')\n", + "val_dataset = dataset(mode='val')\n", + "\n", + "print(val_dataset[50])" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 定义模型" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### 动态图模型（与前面的静态图模型选一运行即可）" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T07:01:06.812574Z", + "iopub.status.busy": "2022-05-04T07:01:06.811871Z", + "iopub.status.idle": "2022-05-04T07:01:06.840305Z", + "shell.execute_reply": "2022-05-04T07:01:06.839497Z", + "shell.execute_reply.started": "2022-05-04T07:01:06.812523Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "from paddle.nn import functional as F\n", + "from paddle.jit import to_static\n", + "class SeparableConv2D(paddle.nn.Layer):\n", + " def __init__(self,\n", + " in_channels,\n", + " out_channels,\n", + " kernel_size,\n", + " stride=1,\n", + " padding=0,\n", + " dilation=1,\n", + " groups=None,\n", + " weight_attr=None,\n", + " bias_attr=None,\n", + " data_format=\"NCHW\"):\n", + " super(SeparableConv2D, self).__init__()\n", + "\n", + " self._padding = padding\n", + " self._stride = stride\n", + " self._dilation = dilation\n", + " self._in_channels = in_channels\n", + " self._data_format = data_format\n", + "\n", + " # 第一次卷积参数，没有偏置参数\n", + " filter_shape = [in_channels, 1] + self.convert_to_list(kernel_size, 2, 'kernel_size')\n", + " self.weight_conv = self.create_parameter(shape=filter_shape, attr=weight_attr)\n", + "\n", + " # 第二次卷积参数\n", + " filter_shape = [out_channels, in_channels] + self.convert_to_list(1, 2, 'kernel_size')\n", + " self.weight_pointwise = self.create_parameter(shape=filter_shape, attr=weight_attr)\n", + " self.bias_pointwise = self.create_parameter(shape=[out_channels],\n", + " attr=bias_attr,\n", + " is_bias=True)\n", + "\n", + " def convert_to_list(self, value, n, name, dtype=np.int):\n", + " if isinstance(value, dtype):\n", + " return [value, ] * n\n", + " else:\n", + " try:\n", + " value_list = list(value)\n", + " except TypeError:\n", + " raise ValueError(\"The \" + name +\n", + " \"'s type must be list or tuple. Received: \" + str(\n", + " value))\n", + " if len(value_list) != n:\n", + " raise ValueError(\"The \" + name + \"'s length must be \" + str(n) +\n", + " \". Received: \" + str(value))\n", + " for single_value in value_list:\n", + " try:\n", + " dtype(single_value)\n", + " except (ValueError, TypeError):\n", + " raise ValueError(\n", + " \"The \" + name + \"'s type must be a list or tuple of \" + str(\n", + " n) + \" \" + str(dtype) + \" . Received: \" + str(\n", + " value) + \" \"\n", + " \"including element \" + str(single_value) + \" of type\" + \" \"\n", + " + str(type(single_value)))\n", + " return value_list\n", + "\n", + " def forward(self, inputs):\n", + " conv_out = F.conv2d(inputs,\n", + " self.weight_conv,\n", + " padding=self._padding,\n", + " stride=self._stride,\n", + " dilation=self._dilation,\n", + " groups=self._in_channels,\n", + " data_format=self._data_format)\n", + "\n", + " out = F.conv2d(conv_out,\n", + " self.weight_pointwise,\n", + " bias=self.bias_pointwise,\n", + " padding=0,\n", + " stride=1,\n", + " dilation=1,\n", + " groups=1,\n", + " data_format=self._data_format)\n", + "\n", + " return out\n", + "\n", + "class Encoder(paddle.nn.Layer):\n", + " def __init__(self, in_channels, out_channels):\n", + " super(Encoder, self).__init__()\n", + "\n", + " self.relus = paddle.nn.LayerList(\n", + " [paddle.nn.ReLU() for i in range(2)])\n", + " self.separable_conv_01 = SeparableConv2D(in_channels,\n", + " out_channels,\n", + " kernel_size=3,\n", + " padding='same')\n", + " self.bns = paddle.nn.LayerList(\n", + " [paddle.nn.BatchNorm2D(out_channels) for i in range(2)])\n", + "\n", + " self.separable_conv_02 = SeparableConv2D(out_channels,\n", + " out_channels,\n", + " kernel_size=3,\n", + " padding='same')\n", + " self.pool = paddle.nn.MaxPool2D(kernel_size=3, stride=2, padding=1)\n", + " self.residual_conv = paddle.nn.Conv2D(in_channels,\n", + " out_channels,\n", + " kernel_size=1,\n", + " stride=2,\n", + " padding='same')\n", + "\n", + " def forward(self, inputs):\n", + " previous_block_activation = inputs\n", + "\n", + " y = self.relus[0](inputs)\n", + " y = self.separable_conv_01(y)\n", + " y = self.bns[0](y)\n", + " y = self.relus[1](y)\n", + " y = self.separable_conv_02(y)\n", + " y = self.bns[1](y)\n", + " y = self.pool(y)\n", + "\n", + " residual = self.residual_conv(previous_block_activation)\n", + " y = paddle.add(y, residual)\n", + "\n", + " return y\n", + "\n", + "\n", + "\n", + "class Decoder(paddle.nn.Layer):\n", + " def __init__(self, in_channels, out_channels):\n", + " super(Decoder, self).__init__()\n", + "\n", + " self.relus = paddle.nn.LayerList(\n", + " [paddle.nn.ReLU() for i in range(2)])\n", + " self.conv_transpose_01 = paddle.nn.Conv2DTranspose(in_channels,\n", + " out_channels,\n", + " kernel_size=3,\n", + " padding=1)\n", + " self.conv_transpose_02 = paddle.nn.Conv2DTranspose(out_channels,\n", + " out_channels,\n", + " kernel_size=3,\n", + " padding=1)\n", + " self.bns = paddle.nn.LayerList(\n", + " [paddle.nn.BatchNorm2D(out_channels) for i in range(2)]\n", + " )\n", + " self.upsamples = paddle.nn.LayerList(\n", + " [paddle.nn.Upsample(scale_factor=2.0) for i in range(2)]\n", + " )\n", + " self.residual_conv = paddle.nn.Conv2D(in_channels,\n", + " out_channels,\n", + " kernel_size=1,\n", + " padding='same')\n", + "\n", + " def forward(self, inputs):\n", + " previous_block_activation = inputs\n", + "\n", + " y = self.relus[0](inputs)\n", + " y = self.conv_transpose_01(y)\n", + " y = self.bns[0](y)\n", + " y = self.relus[1](y)\n", + " y = self.conv_transpose_02(y)\n", + " y = self.bns[1](y)\n", + " y = self.upsamples[0](y)\n", + "\n", + " residual = self.upsamples[1](previous_block_activation)\n", + " residual = self.residual_conv(residual)\n", + "\n", + " y = paddle.add(y, residual)\n", + "\n", + " return y\n", + "\n", + "\n", + "class UNet(paddle.nn.Layer):\n", + " def __init__(self, num_classes):\n", + " super(UNet, self).__init__()\n", + "\n", + " self.conv_1 = paddle.nn.Conv2D(3, 32,\n", + " kernel_size=3,\n", + " stride=2,\n", + " padding='same')\n", + " self.bn = paddle.nn.BatchNorm2D(32)\n", + " self.relu = paddle.nn.ReLU()\n", + "\n", + " in_channels = 32\n", + " self.encoders = []\n", + " self.encoder_list = [64, 128, 256]\n", + " self.decoder_list = [256, 128, 64, 32]\n", + "\n", + " for out_channels in self.encoder_list:\n", + " block = self.add_sublayer('encoder_{}'.format(out_channels),\n", + " Encoder(in_channels, out_channels))\n", + " self.encoders.append(block)\n", + " in_channels = out_channels\n", + "\n", + " self.decoders = []\n", + "\n", + " for out_channels in self.decoder_list:\n", + " block = self.add_sublayer('decoder_{}'.format(out_channels),\n", + " Decoder(in_channels, out_channels))\n", + " self.decoders.append(block)\n", + " in_channels = out_channels\n", + "\n", + " self.output_conv = paddle.nn.Conv2D(in_channels,\n", + " num_classes,\n", + " kernel_size=3,\n", + " padding='same')\n", + " \n", + " def forward(self, inputs):\n", + " y = self.conv_1(inputs)\n", + " y = self.bn(y)\n", + " y = self.relu(y)\n", + "\n", + " for encoder in self.encoders:\n", + " y = encoder(y)\n", + "\n", + " for decoder in self.decoders:\n", + " y = decoder(y)\n", + "\n", + " y = self.output_conv(y)\n", + " return y" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "### 静态图模型（与前面的动态图模型选一运行即可）" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "execution": { + "iopub.execute_input": "2022-05-04T07:01:14.295209Z", + "iopub.status.busy": "2022-05-04T07:01:14.294303Z", + "iopub.status.idle": "2022-05-04T07:01:14.322839Z", + "shell.execute_reply": "2022-05-04T07:01:14.322256Z", + "shell.execute_reply.started": "2022-05-04T07:01:14.295168Z" + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "from paddle.nn import functional as F\n", + "from paddle.jit import to_static\n", + "class SeparableConv2D(paddle.nn.Layer):\n", + " def __init__(self,\n", + " in_channels,\n", + " out_channels,\n", + " kernel_size,\n", + " stride=1,\n", + " padding=0,\n", + " dilation=1,\n", + " groups=None,\n", + " weight_attr=None,\n", + " bias_attr=None,\n", + " data_format=\"NCHW\"):\n", + " super(SeparableConv2D, self).__init__()\n", + "\n", + " self._padding = padding\n", + " self._stride = stride\n", + " self._dilation = dilation\n", + " self._in_channels = in_channels\n", + " self._data_format = data_format\n", + "\n", + " # 第一次卷积参数，没有偏置参数\n", + " filter_shape = [in_channels, 1] + self.convert_to_list(kernel_size, 2, 'kernel_size')\n", + " self.weight_conv = self.create_parameter(shape=filter_shape, attr=weight_attr)\n", + "\n", + " # 第二次卷积参数\n", + " filter_shape = [out_channels, in_channels] + self.convert_to_list(1, 2, 'kernel_size')\n", + " self.weight_pointwise = self.create_parameter(shape=filter_shape, attr=weight_attr)\n", + " self.bias_pointwise = self.create_parameter(shape=[out_channels],\n", + " attr=bias_attr,\n", + " is_bias=True)\n", + "\n", + " def convert_to_list(self, value, n, name, dtype=np.int):\n", + " if isinstance(value, dtype):\n", + " return [value, ] * n\n", + " else:\n", + " try:\n", + " value_list = list(value)\n", + " except TypeError:\n", + " raise ValueError(\"The \" + name +\n", + " \"'s type must be list or tuple. Received: \" + str(\n", + " value))\n", + " if len(value_list) != n:\n", + " raise ValueError(\"The \" + name + \"'s length must be \" + str(n) +\n", + " \". Received: \" + str(value))\n", + " for single_value in value_list:\n", + " try:\n", + " dtype(single_value)\n", + " except (ValueError, TypeError):\n", + " raise ValueError(\n", + " \"The \" + name + \"'s type must be a list or tuple of \" + str(\n", + " n) + \" \" + str(dtype) + \" . Received: \" + str(\n", + " value) + \" \"\n", + " \"including element \" + str(single_value) + \" of type\" + \" \"\n", + " + str(type(single_value)))\n", + " return value_list\n", + "\n", + " def forward(self, inputs):\n", + " conv_out = F.conv2d(inputs,\n", + " self.weight_conv,\n", + " padding=self._padding,\n", + " stride=self._stride,\n", + " dilation=self._dilation,\n", + " groups=self._in_channels,\n", + " data_format=self._data_format)\n", + "\n", + " out = F.conv2d(conv_out,\n", + " self.weight_pointwise,\n", + " bias=self.bias_pointwise,\n", + " padding=0,\n", + " stride=1,\n", + " dilation=1,\n", + " groups=1,\n", + " data_format=self._data_format)\n", + "\n", + " return out\n", + "\n", + "class Encoder(paddle.nn.Layer):\n", + " def __init__(self, in_channels, out_channels):\n", + " super(Encoder, self).__init__()\n", + "\n", + " self.relus = paddle.nn.LayerList(\n", + " [paddle.nn.ReLU() for i in range(2)])\n", + " self.separable_conv_01 = SeparableConv2D(in_channels,\n", + " out_channels,\n", + " kernel_size=3,\n", + " padding='same')\n", + " self.bns = paddle.nn.LayerList(\n", + " [paddle.nn.BatchNorm2D(out_channels) for i in range(2)])\n", + "\n", + " self.separable_conv_02 = SeparableConv2D(out_channels,\n", + " out_channels,\n", + " kernel_size=3,\n", + " padding='same')\n", + " self.pool = paddle.nn.MaxPool2D(kernel_size=3, stride=2, padding=1)\n", + " self.residual_conv = paddle.nn.Conv2D(in_channels,\n", + " out_channels,\n", + " kernel_size=1,\n", + " stride=2,\n", + " padding='same')\n", + "\n", + " def forward(self, inputs):\n", + " previous_block_activation = inputs\n", + "\n", + " y = self.relus[0](inputs)\n", + " y = self.separable_conv_01(y)\n", + " y = self.bns[0](y)\n", + " y = self.relus[1](y)\n", + " y = self.separable_conv_02(y)\n", + " y = self.bns[1](y)\n", + " y = self.pool(y)\n", + "\n", + " residual = self.residual_conv(previous_block_activation)\n", + " y = paddle.add(y, residual)\n", + "\n", + " return y\n", + "\n", + "\n", + "\n", + "class Decoder(paddle.nn.Layer):\n", + " def __init__(self, in_channels, out_channels):\n", + " super(Decoder, self).__init__()\n", + "\n", + " self.relus = paddle.nn.LayerList(\n", + " [paddle.nn.ReLU() for i in range(2)])\n", + " self.conv_transpose_01 = paddle.nn.Conv2DTranspose(in_channels,\n", + " out_channels,\n", + " kernel_size=3,\n", + " padding=1)\n", + " self.conv_transpose_02 = paddle.nn.Conv2DTranspose(out_channels,\n", + " out_channels,\n", + " kernel_size=3,\n", + " padding=1)\n", + " self.bns = paddle.nn.LayerList(\n", + " [paddle.nn.BatchNorm2D(out_channels) for i in range(2)]\n", + " )\n", + " self.upsamples = paddle.nn.LayerList(\n", + " [paddle.nn.Upsample(scale_factor=2.0) for i in range(2)]\n", + " )\n", + " self.residual_conv = paddle.nn.Conv2D(in_channels,\n", + " out_channels,\n", + " kernel_size=1,\n", + " padding='same')\n", + "\n", + " def forward(self, inputs):\n", + " previous_block_activation = inputs\n", + "\n", + " y = self.relus[0](inputs)\n", + " y = self.conv_transpose_01(y)\n", + " y = self.bns[0](y)\n", + " y = self.relus[1](y)\n", + " y = self.conv_transpose_02(y)\n", + " y = self.bns[1](y)\n", + " y = self.upsamples[0](y)\n", + "\n", + " residual = self.upsamples[1](previous_block_activation)\n", + " residual = self.residual_conv(residual)\n", + "\n", + " y = paddle.add(y, residual)\n", + "\n", + " return y\n", + "\n", + "\n", + "class UNet(paddle.nn.Layer):\n", + " def __init__(self, num_classes):\n", + " super(UNet, self).__init__()\n", + "\n", + " self.conv_1 = paddle.nn.Conv2D(3, 32,\n", + " kernel_size=3,\n", + " stride=2,\n", + " padding='same')\n", + " self.bn = paddle.nn.BatchNorm2D(32)\n", + " self.relu = paddle.nn.ReLU()\n", + "\n", + " in_channels = 32\n", + " self.encoders = []\n", + " self.encoder_list = [64, 128, 256]\n", + " self.decoder_list = [256, 128, 64, 32]\n", + "\n", + " for out_channels in self.encoder_list:\n", + " block = self.add_sublayer('encoder_{}'.format(out_channels),\n", + " Encoder(in_channels, out_channels))\n", + " self.encoders.append(block)\n", + " in_channels = out_channels\n", + "\n", + " self.decoders = []\n", + "\n", + " for out_channels in self.decoder_list:\n", + " block = self.add_sublayer('decoder_{}'.format(out_channels),\n", + " Decoder(in_channels, out_channels))\n", + " self.decoders.append(block)\n", + " in_channels = out_channels\n", + "\n", + " self.output_conv = paddle.nn.Conv2D(in_channels,\n", + " num_classes,\n", + " kernel_size=3,\n", + " padding='same')\n", + " @to_static #动静转换\n", + " def forward(self, inputs):\n", + " y = self.conv_1(inputs)\n", + " y = self.bn(y)\n", + " y = self.relu(y)\n", + "\n", + " for encoder in self.encoders:\n", + " y = encoder(y)\n", + "\n", + " for decoder in self.decoders:\n", + " y = decoder(y)\n", + "\n", + " y = self.output_conv(y)\n", + " return y" + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T07:01:19.894617Z", + "iopub.status.busy": "2022-05-04T07:01:19.894069Z", + "iopub.status.idle": "2022-05-04T07:01:25.419713Z", + "shell.execute_reply": "2022-05-04T07:01:25.418806Z", + "shell.execute_reply.started": "2022-05-04T07:01:19.894575Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "W0504 15:01:19.897287 14131 device_context.cc:404] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 10.1, Runtime API Version: 10.1\n", + "W0504 15:01:19.902781 14131 device_context.cc:422] device: 0, cuDNN Version: 7.6.\n" + ] + } + ], + "source": [ + "IMAGE_SIZE = (128, 128)\n", + "num_classes = 2\n", + "network = UNet(num_classes)\n", + "\n", + "#model = paddle.Model(network)\n", + "#model.summary((-1, 3,) + IMAGE_SIZE)" + ] + }, + { + "cell_type": "code", + "execution_count": 5, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T07:01:27.771135Z", + "iopub.status.busy": "2022-05-04T07:01:27.770618Z", + "iopub.status.idle": "2022-05-04T07:01:27.790562Z", + "shell.execute_reply": "2022-05-04T07:01:27.789682Z", + "shell.execute_reply.started": "2022-05-04T07:01:27.771090Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [], + "source": [ + "def get_color_map_list(num_classes):\n", + " color_map = num_classes * [0, 0, 0]\n", + " for i in range(0, num_classes):\n", + " j = 0\n", + " lab = i\n", + " while lab:\n", + " color_map[i * 3] |= (((lab >> 0) & 1) << (7 - j))\n", + " color_map[i * 3 + 1] |= (((lab >> 1) & 1) << (7 - j))\n", + " color_map[i * 3 + 2] |= (((lab >> 2) & 1) << (7 - j))\n", + " j += 1\n", + " lab >>= 3\n", + "\n", + " return color_map\n", + "\n", + "color_map = get_color_map_list(2)\n", + "\n", + "\n", + "train_dataset = dataset(mode='train')\n", + "val_dataset = dataset(mode='val')\n", + "\n", + "train_loader = paddle.io.DataLoader(train_dataset, batch_size=16, shuffle=True)\n", + "val_loader = paddle.io.DataLoader(val_dataset, batch_size=16, shuffle=True)\n", + "\n", + "\n", + "def _fast_hist(label_true, label_pred, n_class):\n", + " mask = (label_true >= 0) & (label_true < n_class)\n", + " hist = np.bincount(\n", + " n_class * label_true[mask].astype(int) +\n", + " label_pred[mask], minlength=n_class ** 2).reshape(n_class, n_class)\n", + " return hist\n", + "\n", + "\n", + "def label_accuracy_score(label_trues, label_preds, n_class):\n", + " \"\"\"Returns accuracy score evaluation result.\n", + " - overall accuracy\n", + " - mean accuracy\n", + " - mean IU\n", + " - fwavacc\n", + " \"\"\"\n", + " hist = np.zeros((n_class, n_class))\n", + " for lt, lp in zip(label_trues, label_preds):\n", + " hist += _fast_hist(lt.flatten(), lp.flatten(), n_class)\n", + " acc = np.diag(hist).sum() / hist.sum()\n", + " with np.errstate(divide='ignore', invalid='ignore'):\n", + " acc_cls = np.diag(hist) / hist.sum(axis=1)\n", + " acc_cls = np.nanmean(acc_cls)\n", + " with np.errstate(divide='ignore', invalid='ignore'):\n", + " iu = np.diag(hist) / (\n", + " hist.sum(axis=1) + hist.sum(axis=0) - np.diag(hist)\n", + " )\n", + " mean_iu = np.nanmean(iu)\n", + " freq = hist.sum(axis=1) / hist.sum()\n", + " fwavacc = (freq[freq > 0] * iu[freq > 0]).sum()\n", + " return acc, acc_cls, mean_iu, fwavacc" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "collapsed": false, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true + }, + "outputs": [], + "source": [] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 开始训练（等待时间较长）" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T07:01:31.724813Z", + "iopub.status.busy": "2022-05-04T07:01:31.724307Z", + "iopub.status.idle": "2022-05-04T07:03:39.476888Z", + "shell.execute_reply": "2022-05-04T07:03:39.475948Z", + "shell.execute_reply.started": "2022-05-04T07:01:31.724769Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "start training ... \n", + "INFO:1,Train Loss:1.52951, Val Loss:0.43075\n", + "INFO:2,Train Loss:0.43300, Val Loss:0.44719\n" + ] + } + ], + "source": [ + "print('start training ... ')\n", + "\n", + "EPOCH_NUM = 2\n", + "BATCH_SIZE = 1\n", + "train_num = 0\n", + "\n", + "visual_IMAGE = Image.open('DataSet/image/1.jpg').resize((128, 128)).convert('RGBA')\n", + "visual_img = np.array(Image.open('DataSet/image/1.jpg').resize((128, 128)))\n", + "optimizer = paddle.optimizer.RMSProp(learning_rate=0.001, parameters=network.parameters())\n", + "normlize = (visual_img-127.5)/127.5\n", + "input = paddle.to_tensor([normlize.transpose(2, 0, 1)])\n", + "\n", + "with LogWriter(logdir=\"./log\") as writer:\n", + " i = 0\n", + " j = 0\n", + " for epoch_id in range(EPOCH_NUM):\n", + " loss_list_each_epoch = []\n", + " network.train()\n", + " for data, label in train_loader:\n", + " i += 1\n", + " y_pred = network(data)\n", + " loss = F.cross_entropy(y_pred, label=label, axis=1)\n", + " train_loss = loss.numpy()[0]\n", + " loss.backward()\n", + " optimizer.step()\n", + " optimizer.clear_grad()\n", + " predict_label = paddle.argmax(network(input)[0], axis=0).numpy().astype('uint64')\n", + " pred_mask = Image.fromarray(predict_label, mode='P')\n", + " pred_mask.putpalette(color_map)\n", + " image = Image.blend(visual_IMAGE, pred_mask.convert('RGBA'), 0.3)\n", + " acc, acc_cls, mean_iu, fwavacc =label_accuracy_score(label.numpy(), np.argmax(y_pred.numpy(), axis=1), 2)\n", + " writer.add_image(tag=\"training Image\", img=np.array(image), step=i)\n", + " writer.add_scalar(tag=\"train/loss\", step=i, value=train_loss)\n", + " writer.add_scalar(tag=\"train/acc\", step=i, value=acc)\n", + " writer.add_scalar(tag=\"train/acc_cls\", step=i, value=acc_cls)\n", + " writer.add_scalar(tag=\"train/mean_iu\", step=i, value=mean_iu)\n", + " writer.add_scalar(tag=\"train/fwavacc\", step=i, value=fwavacc)\n", + " loss_list_each_epoch.append(train_loss)\n", + " network.eval()\n", + " val_loss_list_each_epoch = []\n", + " for data, label in val_loader:\n", + " j += 1\n", + " y_pred = network(data)\n", + " loss = F.cross_entropy(y_pred, label=label, axis=1)\n", + " val_loss = loss.numpy()[0]\n", + " acc, acc_cls, mean_iu, fwavacc =label_accuracy_score(label.numpy(), np.argmax(y_pred.numpy(), axis=1), 2)\n", + " val_loss_list_each_epoch.append(val_loss)\n", + " writer.add_scalar(tag=\"val/loss\", step=j, value=val_loss)\n", + " writer.add_scalar(tag=\"val/acc\", step=j, value=acc)\n", + " writer.add_scalar(tag=\"val/acc_cls\", step=j, value=acc_cls)\n", + " writer.add_scalar(tag=\"val/mean_iu\", step=j, value=mean_iu)\n", + " writer.add_scalar(tag=\"val/fwavacc\", step=j, value=fwavacc)\n", + " print(\"INFO:%d,Train Loss:%0.5f, Val Loss:%0.5f\"%(epoch_id +1, np.average(loss_list_each_epoch), np.average(val_loss_list_each_epoch)))" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "## 可视化训练结果" + ] + }, + { + "cell_type": "code", + "execution_count": 8, + "metadata": { + "collapsed": false, + "execution": { + "iopub.execute_input": "2022-05-04T07:07:08.137764Z", + "iopub.status.busy": "2022-05-04T07:07:08.137196Z", + "iopub.status.idle": "2022-05-04T07:07:08.729041Z", + "shell.execute_reply": "2022-05-04T07:07:08.728396Z", + "shell.execute_reply.started": "2022-05-04T07:07:08.137721Z" + }, + "jupyter": { + "outputs_hidden": false + }, + "scrolled": true, + "tags": [] + }, + "outputs": [ + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "plt.figure(figsize=(10, 10))\n", + "IMAGE_SIZE = (512, 512)\n", + "i = 0\n", + "mask_idx = 0\n", + "\n", + "with open('DataSet/val_list.txt', 'r') as f:\n", + " for line in f.readlines():\n", + " image_path, label_path = line.strip().split(' ')\n", + " resize_t = T.Compose([\n", + " T.Resize(IMAGE_SIZE)\n", + " ])\n", + " image = resize_t(Image.open(image_path))\n", + " label = resize_t(Image.open(label_path))\n", + "\n", + " image = np.array(image).astype('uint8')\n", + " label = np.array(label).astype('uint8')\n", + "\n", + " if i > 8:\n", + " break\n", + " plt.subplot(3, 3, i + 1)\n", + " plt.imshow(image)\n", + " plt.title('Input Image')\n", + " plt.axis(\"off\")\n", + "\n", + " plt.subplot(3, 3, i + 2)\n", + " plt.imshow(label, cmap='gray')\n", + " plt.title('Label')\n", + " plt.axis(\"off\")\n", + "\n", + " # 模型只有一个输出，所以我们通过predict_results[0]来取出1000个预测的结果\n", + " # 映射原始图片的index来取出预测结果，提取mask进行展示\n", + " mask = paddle.argmax(network(paddle.to_tensor([((image - 127.5) / 127.5).transpose(2, 0, 1)]))[0], axis=0).numpy()\n", + "\n", + " plt.subplot(3,3, i + 3)\n", + " plt.imshow(mask.astype('uint8'), cmap='gray')\n", + " plt.title('Predict')\n", + " plt.axis(\"off\")\n", + " i += 3\n", + " mask_idx += 1\n", + "\n", + "plt.show()" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "py35-paddle1.2.0" + }, + "language_info": { + "codemirror_mode": { + "name": "ipython", + "version": 3 + }, + "file_extension": ".py", + "mimetype": "text/x-python", + "name": "python", + "nbconvert_exporter": "python", + "pygments_lexer": "ipython3", + "version": "3.7.4" + }, + "toc-showtags": false + }, + "nbformat": 4, + "nbformat_minor": 4 +}