验证码识别.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import tensorflow as tf\n",
    "from PIL import Image\n",
    "import numpy as np\n",
    "from nets  import nets_factory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1\n"
     ]
    }
   ],
   "source": [
    "#字符数量\n",
    "CHAR_NUM=10\n",
    "#图片高度\n",
    "IMAGE_HEIGHT=60\n",
    "IMAGE_WIDTH=160\n",
    "BATCH_SIZE=20\n",
    "TFRECORD_FILE=\"D:/Tensorflow/captcha/train.tfrecords\"\n",
    "\n",
    "x=tf.placeholder(tf.float32,[None,224,224])\n",
    "y0=tf.placeholder(tf.float32,[None])\n",
    "y1=tf.placeholder(tf.float32,[None])\n",
    "y2=tf.placeholder(tf.float32,[None])\n",
    "y3=tf.placeholder(tf.float32,[None])\n",
    "\n",
    "lr=tf.Variable(0.003,dtype=tf.float32)\n",
    "print('1')\n",
    "def  read_and_decode(filename):\n",
    "    #根据文件名生成队列\n",
    "    filename_queue=tf.train.string_input_producer([filename])\n",
    "    reader=tf.TFRecordReader()\n",
    "    #返回文件名和文件\n",
    "    _, serialized_example=reader.read(filename_queue)\n",
    "    features=tf.parse_single_example(serialized_example,features={'image':tf.FixedLenFeature([],tf.string),\n",
    "                                                                  'label0':tf.FixedLenFeature([],tf.int64),\n",
    "                                                                  'label1':tf.FixedLenFeature([],tf.int64),\n",
    "                                                                  'label2':tf.FixedLenFeature([],tf.int64),\n",
    "                                                                  'label3':tf.FixedLenFeature([],tf.int64)\n",
    "                                                                 })\n",
    "    image=tf.decode_raw(features['image'],tf.uint8)\n",
    "    image=tf.reshape(image,[224,224])\n",
    "    image=tf.cast(image,tf.float32)/255.0\n",
    "    image=tf.subtract(image,0.5)\n",
    "    image=tf.multiply(image,2.0)\n",
    "    \n",
    "    label0=tf.cast(features['label0'],tf.int32)\n",
    "    label1=tf.cast(features['label1'],tf.int32)\n",
    "    label2=tf.cast(features['label2'],tf.int32)\n",
    "    label3=tf.cast(features['label3'],tf.int32)\n",
    "    return image,label0,label1,label2,label3\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter:0 Loss:2171.862 Accuracy:0.25,0.25,0.20,0.30  Learning_rate:0.0010\n",
      "Iter:20 Loss:2.307 Accuracy:0.15,0.15,0.00,0.05  Learning_rate:0.0010\n",
      "Iter:40 Loss:2.285 Accuracy:0.10,0.25,0.25,0.20  Learning_rate:0.0010\n",
      "Iter:60 Loss:2.301 Accuracy:0.20,0.05,0.00,0.25  Learning_rate:0.0010\n",
      "Iter:80 Loss:2.304 Accuracy:0.10,0.05,0.05,0.00  Learning_rate:0.0010\n",
      "Iter:100 Loss:2.305 Accuracy:0.15,0.05,0.10,0.10  Learning_rate:0.0010\n",
      "Iter:120 Loss:2.304 Accuracy:0.15,0.15,0.15,0.05  Learning_rate:0.0010\n",
      "Iter:140 Loss:2.296 Accuracy:0.10,0.10,0.15,0.10  Learning_rate:0.0010\n",
      "Iter:160 Loss:2.309 Accuracy:0.20,0.15,0.10,0.15  Learning_rate:0.0010\n",
      "Iter:180 Loss:2.299 Accuracy:0.10,0.05,0.10,0.10  Learning_rate:0.0010\n",
      "Iter:200 Loss:2.302 Accuracy:0.10,0.15,0.25,0.00  Learning_rate:0.0010\n",
      "Iter:220 Loss:2.310 Accuracy:0.10,0.05,0.15,0.05  Learning_rate:0.0010\n",
      "Iter:240 Loss:2.295 Accuracy:0.00,0.05,0.15,0.05  Learning_rate:0.0010\n",
      "Iter:260 Loss:2.306 Accuracy:0.10,0.10,0.15,0.15  Learning_rate:0.0010\n",
      "Iter:280 Loss:2.293 Accuracy:0.10,0.05,0.10,0.05  Learning_rate:0.0010\n",
      "Iter:300 Loss:2.315 Accuracy:0.10,0.15,0.05,0.10  Learning_rate:0.0010\n",
      "Iter:320 Loss:2.295 Accuracy:0.15,0.10,0.15,0.10  Learning_rate:0.0010\n",
      "Iter:340 Loss:2.318 Accuracy:0.00,0.10,0.05,0.05  Learning_rate:0.0010\n",
      "Iter:360 Loss:2.311 Accuracy:0.05,0.15,0.10,0.15  Learning_rate:0.0010\n",
      "Iter:380 Loss:2.295 Accuracy:0.25,0.20,0.10,0.00  Learning_rate:0.0010\n",
      "Iter:400 Loss:2.313 Accuracy:0.05,0.10,0.10,0.00  Learning_rate:0.0010\n",
      "Iter:420 Loss:2.295 Accuracy:0.00,0.10,0.20,0.05  Learning_rate:0.0010\n",
      "Iter:440 Loss:2.308 Accuracy:0.00,0.00,0.10,0.35  Learning_rate:0.0010\n",
      "Iter:460 Loss:2.288 Accuracy:0.10,0.20,0.15,0.05  Learning_rate:0.0010\n",
      "Iter:480 Loss:2.318 Accuracy:0.05,0.10,0.15,0.05  Learning_rate:0.0010\n",
      "Iter:500 Loss:2.298 Accuracy:0.10,0.05,0.15,0.15  Learning_rate:0.0010\n",
      "Iter:520 Loss:2.300 Accuracy:0.10,0.10,0.10,0.00  Learning_rate:0.0010\n",
      "Iter:540 Loss:2.317 Accuracy:0.15,0.05,0.10,0.05  Learning_rate:0.0010\n",
      "Iter:560 Loss:2.288 Accuracy:0.15,0.10,0.05,0.05  Learning_rate:0.0010\n",
      "Iter:580 Loss:2.285 Accuracy:0.05,0.15,0.10,0.20  Learning_rate:0.0010\n",
      "Iter:600 Loss:2.310 Accuracy:0.10,0.15,0.10,0.05  Learning_rate:0.0010\n",
      "Iter:620 Loss:2.294 Accuracy:0.10,0.15,0.20,0.05  Learning_rate:0.0010\n",
      "Iter:640 Loss:2.303 Accuracy:0.15,0.10,0.00,0.25  Learning_rate:0.0010\n",
      "Iter:660 Loss:2.305 Accuracy:0.05,0.05,0.15,0.15  Learning_rate:0.0010\n",
      "Iter:680 Loss:2.292 Accuracy:0.15,0.10,0.10,0.10  Learning_rate:0.0010\n",
      "Iter:700 Loss:2.305 Accuracy:0.10,0.10,0.20,0.10  Learning_rate:0.0010\n",
      "Iter:720 Loss:2.301 Accuracy:0.20,0.10,0.15,0.20  Learning_rate:0.0010\n",
      "Iter:740 Loss:2.292 Accuracy:0.20,0.35,0.10,0.00  Learning_rate:0.0010\n",
      "Iter:760 Loss:2.268 Accuracy:0.05,0.10,0.35,0.15  Learning_rate:0.0010\n",
      "Iter:780 Loss:2.314 Accuracy:0.10,0.05,0.00,0.05  Learning_rate:0.0010\n",
      "Iter:800 Loss:2.304 Accuracy:0.25,0.25,0.00,0.10  Learning_rate:0.0010\n",
      "Iter:820 Loss:2.296 Accuracy:0.00,0.15,0.20,0.15  Learning_rate:0.0010\n",
      "Iter:840 Loss:2.299 Accuracy:0.10,0.05,0.00,0.10  Learning_rate:0.0010\n",
      "Iter:860 Loss:2.300 Accuracy:0.15,0.15,0.10,0.10  Learning_rate:0.0010\n",
      "Iter:880 Loss:2.312 Accuracy:0.20,0.05,0.05,0.05  Learning_rate:0.0010\n",
      "Iter:900 Loss:2.301 Accuracy:0.15,0.10,0.10,0.15  Learning_rate:0.0010\n",
      "Iter:920 Loss:2.298 Accuracy:0.05,0.10,0.05,0.10  Learning_rate:0.0010\n",
      "Iter:940 Loss:2.312 Accuracy:0.00,0.15,0.15,0.15  Learning_rate:0.0010\n",
      "Iter:960 Loss:2.286 Accuracy:0.25,0.00,0.10,0.15  Learning_rate:0.0010\n",
      "Iter:980 Loss:2.287 Accuracy:0.25,0.20,0.05,0.05  Learning_rate:0.0010\n",
      "Iter:1000 Loss:2.305 Accuracy:0.05,0.10,0.15,0.05  Learning_rate:0.0010\n",
      "Iter:1020 Loss:2.303 Accuracy:0.15,0.10,0.25,0.05  Learning_rate:0.0010\n",
      "Iter:1040 Loss:2.306 Accuracy:0.10,0.25,0.15,0.05  Learning_rate:0.0010\n",
      "Iter:1060 Loss:2.299 Accuracy:0.15,0.00,0.15,0.10  Learning_rate:0.0010\n",
      "Iter:1080 Loss:2.292 Accuracy:0.25,0.10,0.10,0.10  Learning_rate:0.0010\n",
      "Iter:1100 Loss:2.311 Accuracy:0.20,0.05,0.15,0.05  Learning_rate:0.0010\n",
      "Iter:1120 Loss:2.314 Accuracy:0.20,0.10,0.15,0.10  Learning_rate:0.0010\n",
      "Iter:1140 Loss:2.307 Accuracy:0.15,0.20,0.05,0.15  Learning_rate:0.0010\n",
      "Iter:1160 Loss:2.309 Accuracy:0.05,0.15,0.15,0.15  Learning_rate:0.0010\n",
      "Iter:1180 Loss:2.280 Accuracy:0.00,0.30,0.20,0.05  Learning_rate:0.0010\n",
      "Iter:1200 Loss:2.297 Accuracy:0.05,0.15,0.05,0.25  Learning_rate:0.0010\n",
      "Iter:1220 Loss:2.279 Accuracy:0.05,0.15,0.15,0.10  Learning_rate:0.0010\n",
      "Iter:1240 Loss:2.316 Accuracy:0.00,0.00,0.00,0.15  Learning_rate:0.0010\n",
      "Iter:1260 Loss:2.279 Accuracy:0.10,0.15,0.10,0.05  Learning_rate:0.0010\n",
      "Iter:1280 Loss:2.299 Accuracy:0.10,0.15,0.05,0.20  Learning_rate:0.0010\n",
      "Iter:1300 Loss:2.270 Accuracy:0.20,0.25,0.05,0.15  Learning_rate:0.0010\n",
      "Iter:1320 Loss:2.287 Accuracy:0.10,0.15,0.10,0.10  Learning_rate:0.0010\n",
      "Iter:1340 Loss:2.221 Accuracy:0.20,0.15,0.25,0.25  Learning_rate:0.0010\n",
      "Iter:1360 Loss:2.226 Accuracy:0.25,0.15,0.15,0.25  Learning_rate:0.0010\n",
      "Iter:1380 Loss:2.204 Accuracy:0.00,0.20,0.35,0.15  Learning_rate:0.0010\n",
      "Iter:1400 Loss:2.175 Accuracy:0.20,0.25,0.15,0.20  Learning_rate:0.0010\n",
      "Iter:1420 Loss:2.201 Accuracy:0.20,0.15,0.20,0.10  Learning_rate:0.0010\n",
      "Iter:1440 Loss:2.143 Accuracy:0.15,0.25,0.30,0.15  Learning_rate:0.0010\n",
      "Iter:1460 Loss:2.089 Accuracy:0.30,0.15,0.20,0.20  Learning_rate:0.0010\n",
      "Iter:1480 Loss:2.158 Accuracy:0.15,0.20,0.35,0.10  Learning_rate:0.0010\n",
      "Iter:1500 Loss:2.111 Accuracy:0.15,0.20,0.30,0.30  Learning_rate:0.0010\n",
      "Iter:1520 Loss:2.070 Accuracy:0.50,0.10,0.25,0.20  Learning_rate:0.0010\n",
      "Iter:1540 Loss:2.000 Accuracy:0.25,0.25,0.10,0.20  Learning_rate:0.0010\n",
      "Iter:1560 Loss:1.842 Accuracy:0.45,0.30,0.25,0.30  Learning_rate:0.0010\n",
      "Iter:1580 Loss:1.679 Accuracy:0.65,0.40,0.35,0.45  Learning_rate:0.0010\n",
      "Iter:1600 Loss:1.788 Accuracy:0.50,0.20,0.35,0.40  Learning_rate:0.0010\n",
      "Iter:1620 Loss:1.777 Accuracy:0.55,0.20,0.20,0.25  Learning_rate:0.0010\n",
      "Iter:1640 Loss:1.680 Accuracy:0.65,0.10,0.40,0.30  Learning_rate:0.0010\n",
      "Iter:1660 Loss:1.696 Accuracy:0.60,0.35,0.15,0.45  Learning_rate:0.0010\n",
      "Iter:1680 Loss:1.618 Accuracy:0.70,0.15,0.15,0.40  Learning_rate:0.0010\n",
      "Iter:1700 Loss:1.659 Accuracy:0.55,0.30,0.25,0.35  Learning_rate:0.0010\n",
      "Iter:1720 Loss:1.406 Accuracy:0.85,0.20,0.25,0.60  Learning_rate:0.0010\n",
      "Iter:1740 Loss:1.459 Accuracy:0.65,0.25,0.30,0.35  Learning_rate:0.0010\n",
      "Iter:1760 Loss:1.493 Accuracy:0.60,0.30,0.25,0.50  Learning_rate:0.0010\n",
      "Iter:1780 Loss:1.377 Accuracy:0.80,0.20,0.35,0.65  Learning_rate:0.0010\n",
      "Iter:1800 Loss:1.324 Accuracy:0.65,0.35,0.45,0.60  Learning_rate:0.0010\n",
      "Iter:1820 Loss:1.274 Accuracy:0.70,0.55,0.35,0.50  Learning_rate:0.0010\n",
      "Iter:1840 Loss:1.312 Accuracy:0.70,0.35,0.30,0.60  Learning_rate:0.0010\n",
      "Iter:1860 Loss:1.396 Accuracy:0.80,0.25,0.40,0.35  Learning_rate:0.0010\n",
      "Iter:1880 Loss:1.237 Accuracy:0.85,0.30,0.30,0.55  Learning_rate:0.0010\n",
      "Iter:1900 Loss:1.046 Accuracy:0.95,0.35,0.55,0.75  Learning_rate:0.0010\n",
      "Iter:1920 Loss:1.208 Accuracy:0.70,0.50,0.50,0.60  Learning_rate:0.0010\n",
      "Iter:1940 Loss:0.953 Accuracy:0.80,0.40,0.60,0.45  Learning_rate:0.0010\n",
      "Iter:1960 Loss:1.400 Accuracy:0.65,0.15,0.55,0.65  Learning_rate:0.0010\n",
      "Iter:1980 Loss:1.214 Accuracy:0.75,0.50,0.45,0.55  Learning_rate:0.0010\n",
      "Iter:2000 Loss:1.185 Accuracy:0.65,0.30,0.35,0.70  Learning_rate:0.0003\n",
      "Iter:2020 Loss:1.068 Accuracy:0.85,0.65,0.45,0.60  Learning_rate:0.0003\n",
      "Iter:2040 Loss:1.010 Accuracy:0.85,0.40,0.55,0.70  Learning_rate:0.0003\n",
      "Iter:2060 Loss:1.142 Accuracy:0.85,0.55,0.40,0.60  Learning_rate:0.0003\n",
      "Iter:2080 Loss:0.907 Accuracy:0.90,0.55,0.75,0.60  Learning_rate:0.0003\n",
      "Iter:2100 Loss:0.797 Accuracy:0.95,0.45,0.40,0.80  Learning_rate:0.0003\n",
      "Iter:2120 Loss:0.910 Accuracy:0.85,0.55,0.60,0.75  Learning_rate:0.0003\n",
      "Iter:2140 Loss:0.969 Accuracy:0.75,0.80,0.45,0.70  Learning_rate:0.0003\n",
      "Iter:2160 Loss:0.842 Accuracy:0.90,0.60,0.55,0.65  Learning_rate:0.0003\n",
      "Iter:2180 Loss:0.823 Accuracy:0.85,0.70,0.60,0.70  Learning_rate:0.0003\n",
      "Iter:2200 Loss:0.832 Accuracy:0.80,0.50,0.80,0.80  Learning_rate:0.0003\n",
      "Iter:2220 Loss:1.031 Accuracy:1.00,0.45,0.55,0.55  Learning_rate:0.0003\n",
      "Iter:2240 Loss:0.740 Accuracy:0.75,0.70,0.80,0.80  Learning_rate:0.0003\n",
      "Iter:2260 Loss:0.843 Accuracy:0.80,0.65,0.65,0.80  Learning_rate:0.0003\n",
      "Iter:2280 Loss:0.679 Accuracy:0.85,0.65,0.70,0.90  Learning_rate:0.0003\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter:2300 Loss:0.675 Accuracy:0.95,0.75,0.70,0.70  Learning_rate:0.0003\n",
      "Iter:2320 Loss:0.697 Accuracy:1.00,0.85,0.45,0.80  Learning_rate:0.0003\n",
      "Iter:2340 Loss:0.818 Accuracy:0.90,0.70,0.60,0.75  Learning_rate:0.0003\n",
      "Iter:2360 Loss:0.743 Accuracy:0.90,0.70,0.60,0.80  Learning_rate:0.0003\n",
      "Iter:2380 Loss:0.622 Accuracy:0.90,0.65,0.65,0.85  Learning_rate:0.0003\n",
      "Iter:2400 Loss:0.578 Accuracy:0.95,0.75,0.65,0.85  Learning_rate:0.0003\n",
      "Iter:2420 Loss:0.769 Accuracy:0.95,0.80,0.50,0.65  Learning_rate:0.0003\n",
      "Iter:2440 Loss:0.779 Accuracy:0.80,0.70,0.55,0.75  Learning_rate:0.0003\n",
      "Iter:2460 Loss:0.695 Accuracy:0.90,0.60,0.65,0.80  Learning_rate:0.0003\n",
      "Iter:2480 Loss:0.517 Accuracy:0.85,0.75,0.75,1.00  Learning_rate:0.0003\n",
      "Iter:2500 Loss:0.853 Accuracy:0.80,0.75,0.55,0.85  Learning_rate:0.0003\n",
      "Iter:2520 Loss:0.631 Accuracy:0.85,0.75,0.65,0.75  Learning_rate:0.0003\n",
      "Iter:2540 Loss:0.615 Accuracy:1.00,0.60,0.55,0.85  Learning_rate:0.0003\n",
      "Iter:2560 Loss:0.869 Accuracy:0.90,0.40,0.55,0.75  Learning_rate:0.0003\n",
      "Iter:2580 Loss:0.570 Accuracy:0.95,0.75,0.85,0.80  Learning_rate:0.0003\n",
      "Iter:2600 Loss:0.478 Accuracy:0.85,0.80,0.85,0.90  Learning_rate:0.0003\n",
      "Iter:2620 Loss:0.511 Accuracy:0.85,0.85,0.75,1.00  Learning_rate:0.0003\n",
      "Iter:2640 Loss:0.791 Accuracy:0.75,0.65,0.60,0.75  Learning_rate:0.0003\n",
      "Iter:2660 Loss:0.705 Accuracy:0.80,0.65,0.60,0.85  Learning_rate:0.0003\n",
      "Iter:2680 Loss:0.556 Accuracy:0.95,0.75,0.60,0.90  Learning_rate:0.0003\n",
      "Iter:2700 Loss:0.597 Accuracy:0.90,0.80,0.80,0.80  Learning_rate:0.0003\n",
      "Iter:2720 Loss:0.679 Accuracy:0.95,0.70,0.65,0.65  Learning_rate:0.0003\n",
      "Iter:2740 Loss:0.519 Accuracy:1.00,0.80,0.85,0.85  Learning_rate:0.0003\n",
      "Iter:2760 Loss:0.481 Accuracy:0.95,0.65,0.80,0.85  Learning_rate:0.0003\n",
      "Iter:2780 Loss:0.690 Accuracy:0.85,0.60,0.80,0.65  Learning_rate:0.0003\n",
      "Iter:2800 Loss:0.366 Accuracy:0.95,0.75,0.80,0.90  Learning_rate:0.0003\n",
      "Iter:2820 Loss:0.799 Accuracy:0.90,0.60,0.75,0.65  Learning_rate:0.0003\n",
      "Iter:2840 Loss:0.582 Accuracy:0.90,0.60,0.90,0.85  Learning_rate:0.0003\n",
      "Iter:2860 Loss:0.595 Accuracy:0.85,0.85,0.80,0.70  Learning_rate:0.0003\n",
      "Iter:2880 Loss:0.649 Accuracy:0.90,0.80,0.65,0.70  Learning_rate:0.0003\n",
      "Iter:2900 Loss:0.445 Accuracy:0.90,0.95,0.75,0.85  Learning_rate:0.0003\n",
      "Iter:2920 Loss:0.515 Accuracy:0.95,0.75,0.70,0.80  Learning_rate:0.0003\n",
      "Iter:2940 Loss:0.703 Accuracy:0.75,0.70,0.65,0.75  Learning_rate:0.0003\n",
      "Iter:2960 Loss:0.601 Accuracy:0.95,0.60,0.70,0.80  Learning_rate:0.0003\n",
      "Iter:2980 Loss:0.501 Accuracy:0.90,0.75,0.60,0.90  Learning_rate:0.0003\n",
      "Iter:3000 Loss:0.441 Accuracy:1.00,0.70,0.85,0.80  Learning_rate:0.0003\n",
      "Iter:3020 Loss:0.418 Accuracy:0.95,0.75,0.70,0.90  Learning_rate:0.0003\n",
      "Iter:3040 Loss:0.550 Accuracy:1.00,0.90,0.80,0.70  Learning_rate:0.0003\n",
      "Iter:3060 Loss:0.778 Accuracy:0.95,0.75,0.65,0.75  Learning_rate:0.0003\n",
      "Iter:3080 Loss:0.271 Accuracy:0.95,0.85,0.95,0.90  Learning_rate:0.0003\n",
      "Iter:3100 Loss:0.659 Accuracy:0.75,0.50,0.80,0.80  Learning_rate:0.0003\n",
      "Iter:3120 Loss:0.576 Accuracy:0.95,0.75,0.75,0.85  Learning_rate:0.0003\n",
      "Iter:3140 Loss:0.486 Accuracy:0.85,0.75,0.90,0.80  Learning_rate:0.0003\n",
      "Iter:3160 Loss:0.410 Accuracy:0.95,0.80,0.80,0.90  Learning_rate:0.0003\n",
      "Iter:3180 Loss:0.271 Accuracy:1.00,0.95,0.75,0.90  Learning_rate:0.0003\n",
      "Iter:3200 Loss:0.355 Accuracy:1.00,0.80,0.90,0.80  Learning_rate:0.0003\n",
      "Iter:3220 Loss:0.438 Accuracy:0.90,0.75,0.85,0.80  Learning_rate:0.0003\n",
      "Iter:3240 Loss:0.468 Accuracy:0.90,0.70,0.85,0.85  Learning_rate:0.0003\n",
      "Iter:3260 Loss:0.447 Accuracy:0.85,0.80,0.85,0.85  Learning_rate:0.0003\n",
      "Iter:3280 Loss:0.555 Accuracy:0.80,0.80,0.65,0.90  Learning_rate:0.0003\n",
      "Iter:3300 Loss:0.405 Accuracy:0.85,0.85,0.90,0.90  Learning_rate:0.0003\n",
      "Iter:3320 Loss:0.222 Accuracy:0.95,0.90,1.00,0.95  Learning_rate:0.0003\n",
      "Iter:3340 Loss:0.272 Accuracy:1.00,0.90,0.90,0.85  Learning_rate:0.0003\n",
      "Iter:3360 Loss:0.353 Accuracy:1.00,0.85,0.70,0.80  Learning_rate:0.0003\n",
      "Iter:3380 Loss:0.463 Accuracy:1.00,0.75,0.85,0.80  Learning_rate:0.0003\n",
      "Iter:3400 Loss:0.472 Accuracy:0.95,0.85,0.60,0.90  Learning_rate:0.0003\n",
      "Iter:3420 Loss:0.464 Accuracy:0.95,0.90,0.70,0.80  Learning_rate:0.0003\n",
      "Iter:3440 Loss:0.384 Accuracy:0.90,0.90,0.80,0.90  Learning_rate:0.0003\n",
      "Iter:3460 Loss:0.394 Accuracy:0.95,0.75,0.75,0.90  Learning_rate:0.0003\n",
      "Iter:3480 Loss:0.359 Accuracy:1.00,0.80,0.90,0.85  Learning_rate:0.0003\n",
      "Iter:3500 Loss:0.431 Accuracy:0.95,0.80,0.80,0.80  Learning_rate:0.0003\n",
      "Iter:3520 Loss:0.362 Accuracy:0.90,0.90,0.80,0.95  Learning_rate:0.0003\n",
      "Iter:3540 Loss:0.338 Accuracy:1.00,0.85,0.85,0.85  Learning_rate:0.0003\n",
      "Iter:3560 Loss:0.329 Accuracy:0.95,0.80,0.90,0.90  Learning_rate:0.0003\n",
      "Iter:3580 Loss:0.300 Accuracy:1.00,0.90,0.80,0.90  Learning_rate:0.0003\n",
      "Iter:3600 Loss:0.161 Accuracy:1.00,0.95,0.95,0.95  Learning_rate:0.0003\n",
      "Iter:3620 Loss:0.447 Accuracy:0.95,0.80,0.80,1.00  Learning_rate:0.0003\n",
      "Iter:3640 Loss:0.205 Accuracy:0.95,1.00,1.00,0.90  Learning_rate:0.0003\n",
      "Iter:3660 Loss:0.429 Accuracy:1.00,0.85,0.75,0.85  Learning_rate:0.0003\n",
      "Iter:3680 Loss:0.356 Accuracy:1.00,0.80,0.90,0.90  Learning_rate:0.0003\n",
      "Iter:3700 Loss:0.390 Accuracy:0.90,0.95,0.85,0.80  Learning_rate:0.0003\n",
      "Iter:3720 Loss:0.274 Accuracy:0.95,0.85,0.85,0.95  Learning_rate:0.0003\n",
      "Iter:3740 Loss:0.230 Accuracy:0.95,0.90,0.95,0.90  Learning_rate:0.0003\n",
      "Iter:3760 Loss:0.318 Accuracy:0.90,0.90,0.90,0.90  Learning_rate:0.0003\n",
      "Iter:3780 Loss:0.272 Accuracy:1.00,0.90,0.85,0.85  Learning_rate:0.0003\n",
      "Iter:3800 Loss:0.222 Accuracy:0.90,0.95,0.85,0.90  Learning_rate:0.0003\n",
      "Iter:3820 Loss:0.300 Accuracy:0.95,0.80,0.85,0.85  Learning_rate:0.0003\n",
      "Iter:3840 Loss:0.199 Accuracy:0.90,0.95,0.90,0.95  Learning_rate:0.0003\n",
      "Iter:3860 Loss:0.403 Accuracy:1.00,0.85,0.80,0.80  Learning_rate:0.0003\n",
      "Iter:3880 Loss:0.268 Accuracy:1.00,0.90,0.80,0.85  Learning_rate:0.0003\n",
      "Iter:3900 Loss:0.303 Accuracy:0.90,0.85,0.85,0.90  Learning_rate:0.0003\n",
      "Iter:3920 Loss:0.291 Accuracy:0.95,0.70,0.90,0.95  Learning_rate:0.0003\n",
      "Iter:3940 Loss:0.216 Accuracy:0.95,0.90,0.85,0.85  Learning_rate:0.0003\n",
      "Iter:3960 Loss:0.280 Accuracy:0.95,0.90,0.80,1.00  Learning_rate:0.0003\n",
      "Iter:3980 Loss:0.377 Accuracy:0.85,0.90,1.00,0.85  Learning_rate:0.0003\n",
      "Iter:4000 Loss:0.181 Accuracy:0.95,1.00,0.90,0.90  Learning_rate:0.0001\n",
      "Iter:4020 Loss:0.174 Accuracy:1.00,0.80,0.95,0.90  Learning_rate:0.0001\n",
      "Iter:4040 Loss:0.269 Accuracy:0.90,0.80,1.00,0.85  Learning_rate:0.0001\n",
      "Iter:4060 Loss:0.323 Accuracy:0.90,0.95,0.90,0.85  Learning_rate:0.0001\n",
      "Iter:4080 Loss:0.273 Accuracy:1.00,0.85,0.90,0.95  Learning_rate:0.0001\n",
      "Iter:4100 Loss:0.451 Accuracy:0.95,0.90,0.85,0.75  Learning_rate:0.0001\n",
      "Iter:4120 Loss:0.189 Accuracy:0.95,0.95,0.80,1.00  Learning_rate:0.0001\n",
      "Iter:4140 Loss:0.197 Accuracy:1.00,0.85,0.95,0.85  Learning_rate:0.0001\n",
      "Iter:4160 Loss:0.154 Accuracy:1.00,0.95,0.85,1.00  Learning_rate:0.0001\n",
      "Iter:4180 Loss:0.252 Accuracy:0.85,1.00,0.85,0.95  Learning_rate:0.0001\n",
      "Iter:4200 Loss:0.309 Accuracy:0.90,0.95,0.85,0.90  Learning_rate:0.0001\n",
      "Iter:4220 Loss:0.161 Accuracy:1.00,0.90,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:4240 Loss:0.095 Accuracy:0.95,1.00,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:4260 Loss:0.208 Accuracy:1.00,1.00,0.75,0.95  Learning_rate:0.0001\n",
      "Iter:4280 Loss:0.203 Accuracy:1.00,0.95,0.85,0.95  Learning_rate:0.0001\n",
      "Iter:4300 Loss:0.189 Accuracy:1.00,0.85,1.00,1.00  Learning_rate:0.0001\n",
      "Iter:4320 Loss:0.342 Accuracy:0.95,0.80,0.85,0.95  Learning_rate:0.0001\n",
      "Iter:4340 Loss:0.276 Accuracy:1.00,0.85,0.80,0.95  Learning_rate:0.0001\n",
      "Iter:4360 Loss:0.220 Accuracy:1.00,0.80,0.85,1.00  Learning_rate:0.0001\n",
      "Iter:4380 Loss:0.219 Accuracy:1.00,0.85,0.90,0.95  Learning_rate:0.0001\n",
      "Iter:4400 Loss:0.219 Accuracy:1.00,0.95,0.75,0.95  Learning_rate:0.0001\n",
      "Iter:4420 Loss:0.226 Accuracy:1.00,0.90,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:4440 Loss:0.149 Accuracy:1.00,0.90,0.85,1.00  Learning_rate:0.0001\n",
      "Iter:4460 Loss:0.168 Accuracy:1.00,0.85,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:4480 Loss:0.255 Accuracy:0.90,0.90,0.75,1.00  Learning_rate:0.0001\n",
      "Iter:4500 Loss:0.275 Accuracy:0.85,0.80,0.90,0.95  Learning_rate:0.0001\n",
      "Iter:4520 Loss:0.278 Accuracy:1.00,0.95,0.90,0.70  Learning_rate:0.0001\n",
      "Iter:4540 Loss:0.246 Accuracy:1.00,0.90,0.85,0.95  Learning_rate:0.0001\n",
      "Iter:4560 Loss:0.168 Accuracy:1.00,0.95,1.00,0.85  Learning_rate:0.0001\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter:4580 Loss:0.115 Accuracy:1.00,0.95,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:4600 Loss:0.292 Accuracy:0.95,0.80,0.80,0.95  Learning_rate:0.0001\n",
      "Iter:4620 Loss:0.183 Accuracy:1.00,0.95,0.90,0.95  Learning_rate:0.0001\n",
      "Iter:4640 Loss:0.133 Accuracy:0.95,0.95,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:4660 Loss:0.170 Accuracy:1.00,0.85,0.90,0.85  Learning_rate:0.0001\n",
      "Iter:4680 Loss:0.109 Accuracy:1.00,0.95,1.00,0.90  Learning_rate:0.0001\n",
      "Iter:4700 Loss:0.078 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:4720 Loss:0.156 Accuracy:0.95,0.90,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:4740 Loss:0.202 Accuracy:1.00,0.85,0.85,1.00  Learning_rate:0.0001\n",
      "Iter:4760 Loss:0.144 Accuracy:1.00,0.95,0.95,0.90  Learning_rate:0.0001\n",
      "Iter:4780 Loss:0.155 Accuracy:1.00,1.00,0.90,0.90  Learning_rate:0.0001\n",
      "Iter:4800 Loss:0.216 Accuracy:1.00,0.85,0.90,0.85  Learning_rate:0.0001\n",
      "Iter:4820 Loss:0.177 Accuracy:0.95,0.90,1.00,1.00  Learning_rate:0.0001\n",
      "Iter:4840 Loss:0.189 Accuracy:1.00,1.00,0.85,0.95  Learning_rate:0.0001\n",
      "Iter:4860 Loss:0.257 Accuracy:1.00,0.85,0.90,0.95  Learning_rate:0.0001\n",
      "Iter:4880 Loss:0.160 Accuracy:0.95,0.90,0.95,0.90  Learning_rate:0.0001\n",
      "Iter:4900 Loss:0.103 Accuracy:0.95,1.00,1.00,0.90  Learning_rate:0.0001\n",
      "Iter:4920 Loss:0.235 Accuracy:1.00,0.85,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:4940 Loss:0.118 Accuracy:0.95,0.95,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:4960 Loss:0.148 Accuracy:1.00,0.90,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:4980 Loss:0.221 Accuracy:1.00,0.80,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:5000 Loss:0.081 Accuracy:1.00,0.90,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:5020 Loss:0.246 Accuracy:0.90,1.00,0.90,0.90  Learning_rate:0.0001\n",
      "Iter:5040 Loss:0.113 Accuracy:1.00,1.00,0.95,0.85  Learning_rate:0.0001\n",
      "Iter:5060 Loss:0.118 Accuracy:0.95,0.90,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:5080 Loss:0.125 Accuracy:1.00,0.95,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:5100 Loss:0.134 Accuracy:1.00,0.90,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:5120 Loss:0.143 Accuracy:1.00,0.95,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:5140 Loss:0.142 Accuracy:0.95,0.90,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:5160 Loss:0.098 Accuracy:0.95,0.95,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:5180 Loss:0.137 Accuracy:1.00,1.00,0.90,0.95  Learning_rate:0.0001\n",
      "Iter:5200 Loss:0.121 Accuracy:0.95,0.95,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:5220 Loss:0.221 Accuracy:1.00,0.75,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:5240 Loss:0.114 Accuracy:1.00,1.00,0.90,0.95  Learning_rate:0.0001\n",
      "Iter:5260 Loss:0.207 Accuracy:0.95,0.90,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:5280 Loss:0.074 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:5300 Loss:0.061 Accuracy:0.95,0.95,1.00,1.00  Learning_rate:0.0001\n",
      "Iter:5320 Loss:0.224 Accuracy:1.00,0.80,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:5340 Loss:0.226 Accuracy:1.00,0.85,0.85,0.90  Learning_rate:0.0001\n",
      "Iter:5360 Loss:0.074 Accuracy:0.95,1.00,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:5380 Loss:0.073 Accuracy:0.95,0.95,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:5400 Loss:0.112 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:5420 Loss:0.167 Accuracy:0.95,0.85,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:5440 Loss:0.104 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:5460 Loss:0.206 Accuracy:1.00,0.90,0.85,0.95  Learning_rate:0.0001\n",
      "Iter:5480 Loss:0.198 Accuracy:0.95,0.85,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:5500 Loss:0.204 Accuracy:0.90,0.90,1.00,1.00  Learning_rate:0.0001\n",
      "Iter:5520 Loss:0.351 Accuracy:0.95,0.90,0.90,0.85  Learning_rate:0.0001\n",
      "Iter:5540 Loss:0.233 Accuracy:0.95,0.85,0.85,0.90  Learning_rate:0.0001\n",
      "Iter:5560 Loss:0.283 Accuracy:0.90,0.85,0.85,1.00  Learning_rate:0.0001\n",
      "Iter:5580 Loss:0.147 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:5600 Loss:0.163 Accuracy:1.00,0.90,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:5620 Loss:0.259 Accuracy:0.95,0.95,0.90,0.85  Learning_rate:0.0001\n",
      "Iter:5640 Loss:0.126 Accuracy:1.00,1.00,0.80,1.00  Learning_rate:0.0001\n",
      "Iter:5660 Loss:0.112 Accuracy:0.95,0.95,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:5680 Loss:0.060 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0001\n",
      "Iter:5700 Loss:0.060 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:5720 Loss:0.265 Accuracy:0.90,0.95,0.95,0.85  Learning_rate:0.0001\n",
      "Iter:5740 Loss:0.094 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:5760 Loss:0.139 Accuracy:0.95,0.95,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:5780 Loss:0.185 Accuracy:0.90,0.85,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:5800 Loss:0.167 Accuracy:1.00,0.95,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:5820 Loss:0.086 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:5840 Loss:0.100 Accuracy:1.00,1.00,0.90,0.95  Learning_rate:0.0001\n",
      "Iter:5860 Loss:0.083 Accuracy:1.00,0.85,1.00,0.95  Learning_rate:0.0001\n",
      "Iter:5880 Loss:0.140 Accuracy:0.90,1.00,1.00,0.90  Learning_rate:0.0001\n",
      "Iter:5900 Loss:0.180 Accuracy:0.90,0.95,0.90,0.90  Learning_rate:0.0001\n",
      "Iter:5920 Loss:0.136 Accuracy:1.00,0.90,0.90,0.90  Learning_rate:0.0001\n",
      "Iter:5940 Loss:0.167 Accuracy:0.95,0.90,0.95,0.95  Learning_rate:0.0001\n",
      "Iter:5960 Loss:0.129 Accuracy:0.95,0.95,0.95,1.00  Learning_rate:0.0001\n",
      "Iter:5980 Loss:0.126 Accuracy:0.90,1.00,0.90,1.00  Learning_rate:0.0001\n",
      "Iter:6000 Loss:0.079 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:6020 Loss:0.060 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6040 Loss:0.081 Accuracy:1.00,1.00,1.00,0.90  Learning_rate:0.0000\n",
      "Iter:6060 Loss:0.137 Accuracy:1.00,0.90,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:6080 Loss:0.039 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6100 Loss:0.135 Accuracy:1.00,1.00,0.90,0.95  Learning_rate:0.0000\n",
      "Iter:6120 Loss:0.033 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6140 Loss:0.080 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6160 Loss:0.105 Accuracy:0.90,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:6180 Loss:0.057 Accuracy:1.00,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:6200 Loss:0.293 Accuracy:0.90,0.85,0.85,0.85  Learning_rate:0.0000\n",
      "Iter:6220 Loss:0.107 Accuracy:1.00,0.95,0.90,0.95  Learning_rate:0.0000\n",
      "Iter:6240 Loss:0.066 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:6260 Loss:0.087 Accuracy:0.95,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:6280 Loss:0.170 Accuracy:1.00,0.85,1.00,0.90  Learning_rate:0.0000\n",
      "Iter:6300 Loss:0.096 Accuracy:1.00,1.00,1.00,0.90  Learning_rate:0.0000\n",
      "Iter:6320 Loss:0.158 Accuracy:0.90,0.90,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:6340 Loss:0.083 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6360 Loss:0.094 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:6380 Loss:0.141 Accuracy:0.95,1.00,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:6400 Loss:0.093 Accuracy:1.00,1.00,0.90,0.95  Learning_rate:0.0000\n",
      "Iter:6420 Loss:0.085 Accuracy:1.00,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:6440 Loss:0.130 Accuracy:0.95,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6460 Loss:0.078 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6480 Loss:0.110 Accuracy:0.95,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:6500 Loss:0.084 Accuracy:1.00,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6520 Loss:0.171 Accuracy:0.95,0.85,1.00,0.90  Learning_rate:0.0000\n",
      "Iter:6540 Loss:0.210 Accuracy:0.95,0.85,0.85,1.00  Learning_rate:0.0000\n",
      "Iter:6560 Loss:0.048 Accuracy:1.00,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:6580 Loss:0.110 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:6600 Loss:0.088 Accuracy:1.00,0.90,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:6620 Loss:0.144 Accuracy:1.00,1.00,0.90,0.85  Learning_rate:0.0000\n",
      "Iter:6640 Loss:0.220 Accuracy:1.00,0.95,0.90,0.90  Learning_rate:0.0000\n",
      "Iter:6660 Loss:0.110 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6680 Loss:0.147 Accuracy:1.00,0.90,0.85,0.90  Learning_rate:0.0000\n",
      "Iter:6700 Loss:0.105 Accuracy:1.00,1.00,0.95,0.85  Learning_rate:0.0000\n",
      "Iter:6720 Loss:0.111 Accuracy:1.00,0.95,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:6740 Loss:0.079 Accuracy:1.00,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6760 Loss:0.129 Accuracy:1.00,1.00,0.95,0.90  Learning_rate:0.0000\n",
      "Iter:6780 Loss:0.096 Accuracy:0.95,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:6800 Loss:0.092 Accuracy:0.95,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:6820 Loss:0.028 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6840 Loss:0.132 Accuracy:0.95,1.00,0.95,0.90  Learning_rate:0.0000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter:6860 Loss:0.126 Accuracy:0.95,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:6880 Loss:0.155 Accuracy:1.00,0.85,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:6900 Loss:0.136 Accuracy:1.00,0.90,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:6920 Loss:0.102 Accuracy:1.00,1.00,0.85,1.00  Learning_rate:0.0000\n",
      "Iter:6940 Loss:0.179 Accuracy:0.90,0.90,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:6960 Loss:0.185 Accuracy:1.00,0.90,0.95,0.90  Learning_rate:0.0000\n",
      "Iter:6980 Loss:0.107 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:7000 Loss:0.216 Accuracy:1.00,0.95,0.85,0.95  Learning_rate:0.0000\n",
      "Iter:7020 Loss:0.138 Accuracy:0.95,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:7040 Loss:0.135 Accuracy:0.95,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7060 Loss:0.143 Accuracy:1.00,0.95,0.90,0.90  Learning_rate:0.0000\n",
      "Iter:7080 Loss:0.076 Accuracy:0.95,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7100 Loss:0.088 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7120 Loss:0.066 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7140 Loss:0.061 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7160 Loss:0.101 Accuracy:1.00,1.00,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:7180 Loss:0.094 Accuracy:1.00,0.95,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:7200 Loss:0.084 Accuracy:1.00,0.90,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:7220 Loss:0.036 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7240 Loss:0.055 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7260 Loss:0.174 Accuracy:0.95,0.85,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7280 Loss:0.191 Accuracy:0.95,0.95,0.85,1.00  Learning_rate:0.0000\n",
      "Iter:7300 Loss:0.131 Accuracy:1.00,0.90,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:7320 Loss:0.026 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7340 Loss:0.008 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7360 Loss:0.047 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7380 Loss:0.035 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7400 Loss:0.087 Accuracy:0.95,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:7420 Loss:0.135 Accuracy:0.95,1.00,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:7440 Loss:0.075 Accuracy:1.00,1.00,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:7460 Loss:0.070 Accuracy:1.00,1.00,1.00,0.90  Learning_rate:0.0000\n",
      "Iter:7480 Loss:0.066 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:7500 Loss:0.061 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7520 Loss:0.172 Accuracy:0.95,0.95,0.85,0.95  Learning_rate:0.0000\n",
      "Iter:7540 Loss:0.143 Accuracy:0.95,0.95,0.85,1.00  Learning_rate:0.0000\n",
      "Iter:7560 Loss:0.074 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7580 Loss:0.064 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7600 Loss:0.149 Accuracy:0.95,1.00,0.90,0.95  Learning_rate:0.0000\n",
      "Iter:7620 Loss:0.120 Accuracy:1.00,0.85,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7640 Loss:0.066 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7660 Loss:0.078 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:7680 Loss:0.128 Accuracy:0.90,0.90,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7700 Loss:0.108 Accuracy:1.00,1.00,0.95,0.90  Learning_rate:0.0000\n",
      "Iter:7720 Loss:0.072 Accuracy:1.00,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:7740 Loss:0.028 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7760 Loss:0.047 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7780 Loss:0.090 Accuracy:1.00,0.90,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7800 Loss:0.084 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7820 Loss:0.088 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:7840 Loss:0.039 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:7860 Loss:0.070 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:7880 Loss:0.093 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:7900 Loss:0.062 Accuracy:0.95,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:7920 Loss:0.072 Accuracy:1.00,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:7940 Loss:0.072 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:7960 Loss:0.058 Accuracy:1.00,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:7980 Loss:0.172 Accuracy:1.00,1.00,0.90,0.85  Learning_rate:0.0000\n",
      "Iter:8000 Loss:0.067 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8020 Loss:0.089 Accuracy:1.00,1.00,1.00,0.85  Learning_rate:0.0000\n",
      "Iter:8040 Loss:0.075 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:8060 Loss:0.047 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8080 Loss:0.061 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8100 Loss:0.051 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8120 Loss:0.070 Accuracy:0.95,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:8140 Loss:0.067 Accuracy:0.95,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:8160 Loss:0.051 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:8180 Loss:0.078 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:8200 Loss:0.148 Accuracy:1.00,0.90,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:8220 Loss:0.050 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8240 Loss:0.015 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8260 Loss:0.090 Accuracy:1.00,0.95,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:8280 Loss:0.211 Accuracy:1.00,0.90,0.90,0.95  Learning_rate:0.0000\n",
      "Iter:8300 Loss:0.079 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:8320 Loss:0.064 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8340 Loss:0.162 Accuracy:0.90,0.95,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:8360 Loss:0.074 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:8380 Loss:0.052 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:8400 Loss:0.043 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8420 Loss:0.054 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8440 Loss:0.029 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8460 Loss:0.144 Accuracy:1.00,0.95,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:8480 Loss:0.144 Accuracy:0.95,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:8500 Loss:0.032 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8520 Loss:0.142 Accuracy:0.95,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8540 Loss:0.145 Accuracy:1.00,0.95,0.90,0.95  Learning_rate:0.0000\n",
      "Iter:8560 Loss:0.065 Accuracy:0.95,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:8580 Loss:0.151 Accuracy:0.95,1.00,1.00,0.85  Learning_rate:0.0000\n",
      "Iter:8600 Loss:0.080 Accuracy:1.00,0.95,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:8620 Loss:0.077 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8640 Loss:0.089 Accuracy:0.95,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:8660 Loss:0.100 Accuracy:1.00,0.95,1.00,0.90  Learning_rate:0.0000\n",
      "Iter:8680 Loss:0.103 Accuracy:0.95,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:8700 Loss:0.046 Accuracy:1.00,1.00,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:8720 Loss:0.025 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8740 Loss:0.089 Accuracy:0.90,1.00,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:8760 Loss:0.077 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:8780 Loss:0.077 Accuracy:0.95,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8800 Loss:0.049 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:8820 Loss:0.028 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8840 Loss:0.132 Accuracy:1.00,0.95,1.00,0.85  Learning_rate:0.0000\n",
      "Iter:8860 Loss:0.110 Accuracy:0.95,1.00,1.00,0.85  Learning_rate:0.0000\n",
      "Iter:8880 Loss:0.099 Accuracy:0.95,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:8900 Loss:0.057 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:8920 Loss:0.104 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:8940 Loss:0.084 Accuracy:1.00,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:8960 Loss:0.094 Accuracy:0.95,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:8980 Loss:0.079 Accuracy:1.00,1.00,1.00,0.90  Learning_rate:0.0000\n",
      "Iter:9000 Loss:0.059 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9020 Loss:0.048 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:9040 Loss:0.075 Accuracy:1.00,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:9060 Loss:0.023 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9080 Loss:0.113 Accuracy:0.95,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:9100 Loss:0.083 Accuracy:1.00,1.00,0.90,0.95  Learning_rate:0.0000\n",
      "Iter:9120 Loss:0.107 Accuracy:1.00,0.90,0.90,1.00  Learning_rate:0.0000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Iter:9140 Loss:0.045 Accuracy:0.95,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:9160 Loss:0.048 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9180 Loss:0.031 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:9200 Loss:0.041 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:9220 Loss:0.074 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:9240 Loss:0.108 Accuracy:1.00,1.00,0.95,0.90  Learning_rate:0.0000\n",
      "Iter:9260 Loss:0.037 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:9280 Loss:0.074 Accuracy:1.00,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:9300 Loss:0.093 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:9320 Loss:0.079 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:9340 Loss:0.020 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9360 Loss:0.076 Accuracy:1.00,1.00,1.00,0.90  Learning_rate:0.0000\n",
      "Iter:9380 Loss:0.145 Accuracy:1.00,0.95,0.95,0.90  Learning_rate:0.0000\n",
      "Iter:9400 Loss:0.084 Accuracy:0.95,0.90,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:9420 Loss:0.070 Accuracy:1.00,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9440 Loss:0.117 Accuracy:1.00,0.85,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9460 Loss:0.057 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9480 Loss:0.035 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9500 Loss:0.153 Accuracy:1.00,0.95,0.80,0.95  Learning_rate:0.0000\n",
      "Iter:9520 Loss:0.024 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9540 Loss:0.044 Accuracy:0.90,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9560 Loss:0.081 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:9580 Loss:0.110 Accuracy:0.90,1.00,0.90,1.00  Learning_rate:0.0000\n",
      "Iter:9600 Loss:0.044 Accuracy:1.00,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9620 Loss:0.136 Accuracy:1.00,0.95,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:9640 Loss:0.050 Accuracy:1.00,1.00,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:9660 Loss:0.133 Accuracy:0.95,0.95,0.90,0.95  Learning_rate:0.0000\n",
      "Iter:9680 Loss:0.171 Accuracy:1.00,0.90,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:9700 Loss:0.164 Accuracy:1.00,0.95,0.95,0.90  Learning_rate:0.0000\n",
      "Iter:9720 Loss:0.053 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:9740 Loss:0.043 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9760 Loss:0.055 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9780 Loss:0.083 Accuracy:1.00,0.95,0.95,0.90  Learning_rate:0.0000\n",
      "Iter:9800 Loss:0.068 Accuracy:0.95,0.90,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9820 Loss:0.053 Accuracy:0.95,1.00,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9840 Loss:0.090 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9860 Loss:0.059 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9880 Loss:0.124 Accuracy:0.95,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9900 Loss:0.027 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:9920 Loss:0.044 Accuracy:1.00,1.00,0.95,0.95  Learning_rate:0.0000\n",
      "Iter:9940 Loss:0.052 Accuracy:1.00,0.95,1.00,0.95  Learning_rate:0.0000\n",
      "Iter:9960 Loss:0.054 Accuracy:1.00,0.95,1.00,1.00  Learning_rate:0.0000\n",
      "Iter:9980 Loss:0.067 Accuracy:1.00,0.95,0.95,1.00  Learning_rate:0.0000\n",
      "Iter:10000 Loss:0.037 Accuracy:1.00,1.00,0.95,1.00  Learning_rate:0.0000\n"
     ]
    }
   ],
   "source": [
    "image,label0,label1,label2,label3=read_and_decode(TFRECORD_FILE)\n",
    "image_batch,label_batch0,label_batch1,label_batch2,label_batch3=tf.train.shuffle_batch(\n",
    "    [image,label0,label1,label2,label3],batch_size=BATCH_SIZE,capacity=1075,min_after_dequeue=1000,num_threads=1)\n",
    "train_network_fn=nets_factory.get_network_fn(\n",
    "  'alexnet_v2',\n",
    "  num_classes=CHAR_NUM,\n",
    "  weight_decay=0.0005,\n",
    "  is_training=True)\n",
    "gpu_options = tf.GPUOptions(allow_growth=True)\n",
    "with tf.Session(config=tf.ConfigProto(log_device_placement=False,allow_soft_placement=True,gpu_options=gpu_options)) as sess:\n",
    "#     gpu_options = tf.GPUOptions(allow_growth=True)\n",
    "#     tf.Session(config=tf.ConfigProto(log_device_placement=False,allow_soft_placement=True,gpu_options=gpu_options))\n",
    "    X=tf.reshape(x,[BATCH_SIZE,224,224,1])\n",
    "    logits0,logits1,logits2,logits3,end_pintos=train_network_fn(X)\n",
    "    one_hot_labels0=tf.one_hot(indices=tf.cast(y0,tf.int32),depth=CHAR_NUM)\n",
    "    one_hot_labels1=tf.one_hot(indices=tf.cast(y1,tf.int32),depth=CHAR_NUM)\n",
    "    one_hot_labels2=tf.one_hot(indices=tf.cast(y2,tf.int32),depth=CHAR_NUM)\n",
    "    one_hot_labels3=tf.one_hot(indices=tf.cast(y3,tf.int32),depth=CHAR_NUM)\n",
    "    \n",
    "    loss0=tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits0,labels=one_hot_labels0))\n",
    "    loss1=tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits1,labels=one_hot_labels1))\n",
    "    loss2=tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits2,labels=one_hot_labels2))\n",
    "    loss3=tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits3,labels=one_hot_labels3))\n",
    "    total_loss=(loss0+loss1+loss2+loss3)/4.0\n",
    "    optimizer=tf.train.AdamOptimizer(learning_rate=lr).minimize(total_loss)\n",
    "    \n",
    "    correct_prediction0=tf.equal(tf.argmax(one_hot_labels0,1),tf.argmax(logits0,1))\n",
    "    accuracy0=tf.reduce_mean(tf.cast(correct_prediction0,tf.float32))\n",
    "    \n",
    "    correct_prediction1=tf.equal(tf.argmax(one_hot_labels1,1),tf.argmax(logits1,1))\n",
    "    accuracy1=tf.reduce_mean(tf.cast(correct_prediction1,tf.float32))\n",
    "    \n",
    "    correct_prediction2=tf.equal(tf.argmax(one_hot_labels2,1),tf.argmax(logits2,1))\n",
    "    accuracy2=tf.reduce_mean(tf.cast(correct_prediction2,tf.float32))\n",
    "    \n",
    "    correct_prediction3=tf.equal(tf.argmax(one_hot_labels3,1),tf.argmax(logits3,1))\n",
    "    accuracy3=tf.reduce_mean(tf.cast(correct_prediction3,tf.float32))\n",
    "    \n",
    "    saver=tf.train.Saver()\n",
    "    sess.run(tf.global_variables_initializer())\n",
    "    coord=tf.train.Coordinator()  \n",
    "    threads=tf.train.start_queue_runners(sess=sess,coord=coord)\n",
    "    for i in range(10001):\n",
    "        b_image,b_label0,b_label1,b_label2,b_label3=sess.run([image_batch,label_batch0,label_batch1,label_batch2,label_batch3])\n",
    "        sess.run(optimizer,feed_dict={x:b_image,y0:b_label0,y1:b_label1,y2:b_label2,y3:b_label3})       \n",
    "        if i%20==0:\n",
    "            if i%2000==0:\n",
    "                sess.run(tf.assign(lr,lr/3))\n",
    "            acc0,acc1,acc2,acc3,loss_=sess.run([accuracy0,accuracy1,accuracy2,accuracy3,total_loss],feed_dict={x:b_image,                                                                                                y0:b_label0,\n",
    "                                                                                                          y1:b_label1,\n",
    "                                                                                                          y2:b_label2,\n",
    "                                                                                                          y3:b_label3})\n",
    "            learning_rate=sess.run(lr)\n",
    "            print(\"Iter:%d Loss:%.3f Accuracy:%.2f,%.2f,%.2f,%.2f  Learning_rate:%.4f\"%(i,loss_,acc0,acc1,acc2,acc3,learning_rate))\n",
    "            if i==10000:  \n",
    "                saver.save(sess,'./captcha/models/crack_captcha.model',global_step=1)\n",
    "                break\n",
    "    coord.request_stop()\n",
    "    coord.join(threads)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}