3d_ALL-IDB classification using pre-trained networks-Part4.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ALL-IDB classification using ResNet18"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import os\n",
    "import struct\n",
    "import torch\n",
    "from PIL import Image\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import torchvision\n",
    "from torch.autograd import Variable\n",
    "from torch.utils.data import TensorDataset,DataLoader\n",
    "from torchvision import models\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "import copy\n",
    "import time"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load Data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Datapath = 'ALL_IDB2/img/'\n",
    "listing = os.listdir(Datapath)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# ALL_IDB2 dataset has 260 images in total\n",
    "TrainImages = torch.FloatTensor(200,3,224,224)\n",
    "TrainLabels = torch.LongTensor(200)\n",
    "TestImages = torch.FloatTensor(60,3,224,224)\n",
    "TestLabels = torch.LongTensor(60)\n",
    "\n",
    "# First 200 images are used for training and the remaining 60 for testing\n",
    "img_no = 0\n",
    "for file in listing:\n",
    "    im=Image.open(Datapath + file)\n",
    "    im = im.resize((224,224))\n",
    "    im = np.array(im)   \n",
    "    if img_no < 200:\n",
    "        TrainImages[img_no] = torch.from_numpy(im).transpose(0,2).unsqueeze(0)\n",
    "        TrainLabels[img_no] = int(listing[img_no][6:7])\n",
    "    else:\n",
    "        TestImages[img_no - 200] = torch.from_numpy(im).transpose(0,2).unsqueeze(0)\n",
    "        TestLabels[img_no - 200] = int(listing[img_no][6:7])\n",
    "    img_no = img_no + 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([200, 3, 224, 224])\n",
      "torch.Size([200])\n",
      "torch.Size([60, 3, 224, 224])\n",
      "torch.Size([60])\n"
     ]
    }
   ],
   "source": [
    "print(TrainImages.size())\n",
    "print(TrainLabels.size())\n",
    "print(TestImages.size())\n",
    "print(TestLabels.size())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Creating pytorch dataset\n",
    "trainDataset = TensorDataset(TrainImages, TrainLabels)\n",
    "testDataset = TensorDataset(TestImages, TestLabels)\n",
    "# Creating dataloader\n",
    "BatchSize = 32\n",
    "trainLoader = DataLoader(trainDataset, batch_size=BatchSize, shuffle=True,num_workers=4, pin_memory=True)\n",
    "testLoader = DataLoader(testDataset, batch_size=BatchSize, shuffle=True,num_workers=4, pin_memory=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GPU is available!\n"
     ]
    }
   ],
   "source": [
    "# Check availability of GPU\n",
    "use_gpu = torch.cuda.is_available()\n",
    "if use_gpu:\n",
    "    print('GPU is available!')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Initialize the network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ResNet (\n",
      "  (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
      "  (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
      "  (relu): ReLU (inplace)\n",
      "  (maxpool): MaxPool2d (size=(3, 3), stride=(2, 2), padding=(1, 1), dilation=(1, 1))\n",
      "  (layer1): Sequential (\n",
      "    (0): BasicBlock (\n",
      "      (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (relu): ReLU (inplace)\n",
      "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
      "    )\n",
      "    (1): BasicBlock (\n",
      "      (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (relu): ReLU (inplace)\n",
      "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
      "    )\n",
      "  )\n",
      "  (layer2): Sequential (\n",
      "    (0): BasicBlock (\n",
      "      (conv1): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (relu): ReLU (inplace)\n",
      "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (downsample): Sequential (\n",
      "        (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "        (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
      "      )\n",
      "    )\n",
      "    (1): BasicBlock (\n",
      "      (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (relu): ReLU (inplace)\n",
      "      (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
      "    )\n",
      "  )\n",
      "  (layer3): Sequential (\n",
      "    (0): BasicBlock (\n",
      "      (conv1): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (relu): ReLU (inplace)\n",
      "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (downsample): Sequential (\n",
      "        (0): Conv2d(128, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "        (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
      "      )\n",
      "    )\n",
      "    (1): BasicBlock (\n",
      "      (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (relu): ReLU (inplace)\n",
      "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
      "    )\n",
      "  )\n",
      "  (layer4): Sequential (\n",
      "    (0): BasicBlock (\n",
      "      (conv1): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (relu): ReLU (inplace)\n",
      "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (downsample): Sequential (\n",
      "        (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "        (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
      "      )\n",
      "    )\n",
      "    (1): BasicBlock (\n",
      "      (conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
      "      (relu): ReLU (inplace)\n",
      "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
      "    )\n",
      "  )\n",
      "  (avgpool): AvgPool2d (size=7, stride=7, padding=0, ceil_mode=False, count_include_pad=True)\n",
      "  (fc): Linear (512 -> 2)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "# ResNet18\n",
    "net = models.resnet18(pretrained=True)\n",
    "num_ftrs = net.fc.in_features\n",
    "net.fc = nn.Linear(num_ftrs, 2)\n",
    "print(net)\n",
    "if use_gpu:\n",
    "    net = net.cuda()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define loss function and optimizer: "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "criterion = nn.NLLLoss() # Negative Log-Likelihood\n",
    "optimizer = optim.SGD(net.parameters(), lr=1e-4, momentum=0.9) # Stochastic gradient descent"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train the network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "At Iteration: 1 /20  ;  Training Loss: -0.029142 ; Testing Acc: 56.667 ; Time consumed: 0m 2s \n",
      "At Iteration: 2 /20  ;  Training Loss: -0.057836 ; Testing Acc: 55.000 ; Time consumed: 0m 1s \n",
      "At Iteration: 3 /20  ;  Training Loss: -0.098308 ; Testing Acc: 56.667 ; Time consumed: 0m 1s \n",
      "At Iteration: 4 /20  ;  Training Loss: -0.146625 ; Testing Acc: 56.667 ; Time consumed: 0m 1s \n",
      "At Iteration: 5 /20  ;  Training Loss: -0.199592 ; Testing Acc: 60.000 ; Time consumed: 0m 1s \n",
      "At Iteration: 6 /20  ;  Training Loss: -0.252919 ; Testing Acc: 63.333 ; Time consumed: 0m 1s \n",
      "At Iteration: 7 /20  ;  Training Loss: -0.306753 ; Testing Acc: 75.000 ; Time consumed: 0m 1s \n",
      "At Iteration: 8 /20  ;  Training Loss: -0.362130 ; Testing Acc: 80.000 ; Time consumed: 0m 1s \n",
      "At Iteration: 9 /20  ;  Training Loss: -0.421076 ; Testing Acc: 90.000 ; Time consumed: 0m 1s \n",
      "At Iteration: 10 /20  ;  Training Loss: -0.475971 ; Testing Acc: 88.333 ; Time consumed: 0m 1s \n",
      "At Iteration: 11 /20  ;  Training Loss: -0.535875 ; Testing Acc: 88.333 ; Time consumed: 0m 1s \n",
      "At Iteration: 12 /20  ;  Training Loss: -0.591472 ; Testing Acc: 93.333 ; Time consumed: 0m 1s \n",
      "At Iteration: 13 /20  ;  Training Loss: -0.655974 ; Testing Acc: 91.667 ; Time consumed: 0m 1s \n",
      "At Iteration: 14 /20  ;  Training Loss: -0.715094 ; Testing Acc: 90.000 ; Time consumed: 0m 1s \n",
      "At Iteration: 15 /20  ;  Training Loss: -0.776597 ; Testing Acc: 93.333 ; Time consumed: 0m 1s \n",
      "At Iteration: 16 /20  ;  Training Loss: -0.839747 ; Testing Acc: 95.000 ; Time consumed: 0m 1s \n",
      "At Iteration: 17 /20  ;  Training Loss: -0.902541 ; Testing Acc: 93.333 ; Time consumed: 0m 1s \n",
      "At Iteration: 18 /20  ;  Training Loss: -0.969837 ; Testing Acc: 90.000 ; Time consumed: 0m 1s \n",
      "At Iteration: 19 /20  ;  Training Loss: -1.034559 ; Testing Acc: 93.333 ; Time consumed: 0m 1s \n",
      "At Iteration: 20 /20  ;  Training Loss: -1.086820 ; Testing Acc: 91.667 ; Time consumed: 0m 1s \n",
      "Training completed in 0m 18s\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEPCAYAAABV6CMBAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBYeIiCSi4BARkUQUHCIikoiCQ0REEvn/84xwTxP6dTYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f56f4830890>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEPCAYAAABRHfM8AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3Xl8VPX1//HXCQkiIPvWL6v7CiIoWtQaxSotWrup2Epd\n+kNEqWsVvm2VRKsVd+taKnX5VktdaqVqFQGjQkVAAQFZVBYBQcq+KUtyfn/MTRhilplkbu7M5P18\nPPJg7n4yjnNyP+eee83dERERqU5O1AGIiEhmUMIQEZGEKGGIiEhClDBERCQhShgiIpIQJQwREUlI\n6AnDzPqb2QIzW2RmwytY3sLM/mFms81sqpkdEbdsaTB/pplNCztWERGpnIXZh2FmOcAioB/wBTAd\nGOjuC+LWuRPY4u63mtmhwMPufnqwbDHQ2903hBakiIgkJOwzjD7AJ+6+zN13AWOBc8qtcwQwCcDd\nFwLdzKxtsMzqIEYREUlA2F/GHYHlcdMrgnnxZgM/BjCzPkAXoFOwzIE3zWy6mQ0OOVYREalCbtQB\nAHcAD5jZh8AcYCZQHCw70d1XBWccb5rZfHefHFWgIiL1WdgJYyWxM4ZSnYJ5Zdx9C3Bp6bSZLQEW\nB8tWBf/+18xeIjbE9Y2EYWa6IZaISJLc3ZJZP+whqenAQWbW1cwaAgOBcfErmFlzM8sLXg8G3nb3\nrWbW2MyaBvObAGcAcys7kLvrJwU/I0eOjDyGbPrR+1nD923iSKzAoADa3dmOjds2lr2fQ18ZWrbs\nO3/5Drt27Yo83kz8qYlQE4a7FwPDgPHAPGCsu883syFmdlmw2uHAXDObD5wJXB3Mbw9MNrOZwFTg\nX+4+Psx4RSRary58lbxb8yh8t5B9cvdh2v+bxpc3fEnzxs3L1nlkwCPs/O1OTu5yMu98/g4Nb2vI\n1f++uoq9SqqEXsNw99eBQ8vN+1Pc66nllwfzlwA9w45PRKL3+abP6f5Idzbv3EyO5TD67NEM7lX5\ndS65ubm8c8k7bPx6I0c/ejR/nPZHHpnxCH8++89c3PPiugu8ntElq7KX/Pz8qEPIKno/q7Zr1y4O\n+uNBdL2/K5t3buaSoy+h+ObiSpNF+fezRaMWLLt2GdP+3zQa5TbikpcvodkfmvHhFx/WQfT1T6iN\ne3XFzDwbfg+R+mTAMwN47dPXAOj9rd7MuGxGrfc55sMxXP7q5ewu2U3X5l2ZdfksWjRqUev9ZiMz\nw5Msemd1wujWrRvLli2LIKL00rVrV5YuXRp1GCIAFEwq4JZ3b8Fx2jVux6IrF+1Vo0iFK169gsdm\nPIbjnNL1FCZcOIHc3HToIkgfShjfnF/jqwGyid4HSQevLnyVHz73Q3aX7KZRbiPeufgdjut4XGjH\n2717N6f932m8+/m7GMav+vyKB773QGjHyzRKGN+cry9K9D5kspzCHNo1bsfqG1ZHHUqttLmzDeu+\nWkeO5fDYWY9VWdBOtdLC+OebPycvJ4/V162mVZNWdXb88gaPG0zn5p256Ts3YZbU93VKKWF8c76+\nKNH7kKlKv2RLnb7/6bz5izcjjKhmutzXheWbl/O9g77Haz9/LbI4pq+cTt+/9GXf3H3ZcOMGGjRo\nUOcxXP/G9dz//v0c3uZw9ttnP+454x76du5b53FAzRKGrpISSUOfrv+UdV+tI8/yePei2JDKhCUT\nsELjlqJbog4vYcf/+XiWb15O3059I00WAMd1PI5Fwxaxbdc2uj7Qtc6Pf//U+7lv6n08++Nn+Wjo\nR1xx7BUMfGEg5z5/Lp+t/6zO46kJJQyRNHTwgwcDsOmGTZzU7SRKRpZw08k3ATDy7ZHkFOYweWl6\n31btghcuYNoX0+jWvBtTfjkl6nAA2L/l/ky5ZAort6yk52N11+b1/Lznue6N67j7jLs5/6jzybEc\nBh09iAXDFnBMh2M4/vHjufb1a1n/1fo6i6kmlDAisv/++zNp0qRa7eOpp57i5JNPTlFEki6G/GsI\nAL079Gbfffctm3/LabfgI53Tup2G45z81Mk0+n0jNn69MapQK1VQVMDYeWNp1rAZS65ZEnU4ezmh\n8wk899PnmP3lbAY8MyD0403+fDIDXxzI1cdfzXXfvm6vZY3zGvObk3/DvCvmsaN4B4c+dCj3vncv\nO3bvCD2umlDCyGDuHmnRTMIx+sPRAMwYUnFfwsSLJuIjnQ5NO7CjeActR7Wk4z3lnxoQnefnPU/h\n24Xk5eSx6X83RR1Ohc498lxGnT6K1z59jStevSK04yxcu5B+T/XjR4f9iPv631fpeu2btueRAY/w\nzsXvULS0iMMfPpzn5j2XfrXHqG+AlaKbaHlFKpsftUGDBnlOTo43btzY99tvP7/rrrt86tSp3rdv\nX2/RooX37NnTi4qKytZ/4okn/IADDvD99tvPDzjgAH/22Wd9/vz53qhRI8/NzfWmTZt6y5YtKz1e\nur4P8k2tR7V2CvCnZz6d0Pobvtrg+9y6j1OAU4Cf/tTpIUdYtdmrZjsFuBWY79ixI9JYEnHlq1c6\nBfgd796R8n1/ueVLb3JbEz/hzyckve3ExRP9mMeO8RMeP8EnL5uc8tjcy74XkvuuTXaDdPzJtITh\n7t6tWzefNGmSu7uvXLnSW7du7a+//rq7u0+YMMFbt27ta9eu9W3btnmzZs38k08+cXf31atX+8cf\nf+zu7k8++aSffPLJ1R4rnd8H2eOTdZ84BXheYV7S27675F23AitLHIVvFYYQYdXWbVlXdvwtO7bU\n+fFrasAzA5wC/Lm5z6Vsn9t2bPO2d7b1Ax840IuLi2u0j+KSYn961tPe+d7O/pO//8Q/WfdJyuJz\nr1nCqNetj1aYmuEcH1mz08bYfzP461//yoABAzjzzDMB6NevH8ceeyyvvfYaP/nJT2jQoAFz5syh\nU6dOtG/fnvbt26ckbkkv8YXuZJUWxm+edDO3vnsrI98eScHbBbxz0Tuc1O2kVIf6DTt37qT1Pa0B\nmHf5PJo2bBr6MVPllZ+9wjGPHcN5L5zHlGZTan2Za0lJCT3/FCuof3T5R+Tk1Gzkv7Qw/tMjfsr9\nU+/nhMdP4MIeF3LTd26idePWtYqxxpLNMOn4Q4aeYUycONHd3a+44gpv1KiRt2zZ0lu2bOktWrTw\npk2b+qhRo9zdffz48f7d737XW7Ro4WeddZYvWLDA3XWGkU0uG3eZU4D3fqx3SvZ32pOnlf21v8+t\n+/iGrzakZL+Vyb0l1ynAX/r4pVCPE6aO93T0nMIcX7x+ca320/fxvt7ktib+5ZYvUxRZzJdbv/Sh\nrwz1Nne28bum3OVf7/q6VvujBmcYKnpHJL5Y3blzZ37xi1+wfv161q9fz4YNG9iyZQs33ngjAN/9\n7ncZP348q1ev5tBDD+Wyyy77xj4ks1VX6E5WRYXxAx84MCX7Lq/ZH5qxu2Q3fzj1D/zw8B+Gcoy6\nsOzqZTTJa8JhDx3Gpq9qVqw/9/lzmf7FdGYMnkG7pu1SGl+7Ju3KCuPTVk5jw9cbUrr/RChhRKRD\nhw4sXrwYgAsvvJB//etfjB8/npKSEr7++mvefvttvvjiC9asWcO4cePYvn07eXl5NG3atOwUt337\n9qxYsYJdu3ZF+atILbW5sw0AT5/zdMr3ver6VWwasYkmeU1YvHExVmj8aOyPUrb/Lvd1YcvOLVzU\n4yJGfGdEyvYbhQYNGrD8muUAdL6/M8XFxUltf/0b1/OP+f9gwqAJHNb2sDBCBODwtofz3LnP0aFp\nh9COUalkT0nS8YcMHJJ6+eWXvUuXLt6yZUu/5557fNq0aX7KKad4q1atvF27dn7WWWf58uXLfdWq\nVX7KKad4ixYtvGXLln7qqaf6/Pnz3d19586dftZZZ3mrVq28bdu2lR4rnd+H+q42he5kTVs+zXMK\nc8qGqu6Zck+t9tdndB+nAO/7eN8URZgeFq9f7DmFOf4/9/xPwtvc9959bgXmY+eMDTGy1KIGQ1K6\nl1Q9oPchfZVeeLH9xu17NemF6c7JdzJ84nAgVlideulUjuuU3F1jf/7iz3l27rN0a94t7RrzUmHq\n8ql8+y/f5uh2RzNr6Kwq133h4xc47/nzuPuMu7/RmJfOdC8pkQxSWUd32G486UZ8pPODQ35AiZfQ\nZ0wfmt7elM07Nie0fUFRAc/OfTYtu7hTpawbfE3V3eCTP5/M+S+cX2EXdzYK/QzDzPoD9xNLTmPc\nfVS55S2AvwAHAl8Bl7r7x4lsG7cPnWFUQe9Deio9u6jpZdmpcsD9B7BkU+yL/8CWB/LpVZ9Wuu7z\n857nvBfOIy8nj5037ayrECNz95S7uWHCDQw9diiPDHhkr2UL1y6kx6M9OOuQs3jx/BcjirDm0u4M\nw8xygIeAM4EjgQvMrHw16DfATHc/GrgI+GMS24pkpDAL3clafM3issL4Zxs+q7Qw/tHqjzjvhfMw\njK3Dt0YQad379Ym/5srjruTRGY8yavKev1fXbF1D79G96fWtXhmZLGoq7CGpPsAn7r7M3XcBY4Fz\nyq1zBDAJwN0XAt3MrG2C24pknPhblw/qOSjqcABotk8ztv5mK9N+OY0cy+GfC/+JFRr3/udeANZv\nXc/RfzoagM3/u5mGDRtGGW6deuj7DzHg4AGMmDiCv8/9O9t3bqf7o93p0LQDUy5Nj7vw1pWwE0ZH\nYHnc9IpgXrzZwI8BzKwP0AXolOC2IhVK17t9Qu06usN2XKfjKL65mFH9Yn9NX//m9TS4pUHGdnGn\nyis/e4We7Xsy8MWBHPbwYTheqy7uTJUOtwa5A3jAzD4E5gAzgeQugAYKCgrKXufn55Ofn5+i8CTT\n3DzxZm6dfCsAd/e7m+tPuj7iiPaIqtCdrBtPupEbT7qRc/52DuMWjQPgpfNe4oj2R0QcWXRmXj6T\nzvd1ZvXW1ay4dgWNGzaOOqSkFBUVUVRUVKt9hFr0NrMTgAJ37x9MjyB27W+FxetgnSVAd+CoRLet\nrOjdrVs3li1blpLfJZN17dqVpUuXRh1GnckpzMHZ+/MwZ+gcjmp3VEQR7ZEuhe5kbd+1ncZ5mfUF\nGQZ3Z/uu7TRp2CTqUGqtJkXvsM8wpgMHmVlXYBUwELggfgUzaw5sd/ddZjYYeNvdt5pZtdtWpz59\nScoepcnCRzpt72zL2q/W0v3R7uRaLhuv30iTJtH8z55Ohe5kKVnEmFlWJIuaCnUAzt2LgWHAeGAe\nMNbd55vZEDO7LFjtcGCumc0ndkXU1VVtG2a8kvk+XR27JLRXh14A/PfG/7L111vJtVx2+26a3t2U\nb931rbqPKw0L3SLJyupOb6l/Gv++MV8Vf1XhkM/cNXPp/mj3sunTu53Omxe9WSdxRdHRLVKVtOvD\nEKlrXxV/Vemyo9odhY90bsu/DYAJSydghcY9k+8JNaZMKXSLVEdnGJJVrNBonNeYbb/ZVu26pz15\nGm8te6tsOqzCeKYWuiW76QxD6rWjHo592X921WcJrT/p4kn4SKfNvrFidPdHu5N3Sx7btlWfbBKV\nyYVukfJ0hiFZozZ/yW/bto0W97Rgt+8GoEPjDqy6YVWt4vl0/acc/ODB5FkeO2/O/vsuSWapyRmG\nEoZkDSs0DKNkZEmN91G+MJ4KKnRLOtKQlNRbN0+8GYA7+91Zq/3EF8ZzLZeGOQ1r9fOrY3+lZCFZ\nQ2cYkhVKu7tVWBZJjM4wpN4qfysQEUk9JQzJeOW7u0UkHBqSkoxXVXe3iFRMQ1JSL1XV3S0iqaOE\nIVlBd1MVCZ8ShmS0ZLu7RaTmlDAko81bOw+ADk07RByJSPZTwpCMZyRVtxORGlLCkIyVqu5uEUmM\nLquVjKXubpGa02W1Uq+ou1ukbilhSEZSd7dI3Qs9YZhZfzNbYGaLzGx4Bcubmdk4M5tlZnPM7OK4\nZUvNbLaZzTSzaWHHKpmjx+M9APhgyAcRRyJSf+SGuXMzywEeAvoBXwDTzexld18Qt9qVwDx3/4GZ\ntQEWmtlf3X03UALku/uGMOOUzKPubpG6F/YZRh/gE3df5u67gLHAOeXWcWC/4PV+wLogWQBYHcQo\nGUrd3SJ1K+wv447A8rjpFcG8eA8BR5jZF8Bs4Oq4ZQ68aWbTzWxwqJFKxlB3t0g0Qh2SStCZwEx3\nP83MDiSWIHq4+1bgRHdfZWZtg/nz3X1yRTspKCgoe52fn09+fn4dhC5RUHe3SPKKioooKiqq1T5C\n7cMwsxOAAnfvH0yPANzdR8Wt8wrwB3efEkxPBIa7+4xy+xoJbHH3eys4jvow6pFUPLtbpL5Lxz6M\n6cBBZtbVzBoCA4Fx5dZZBpwOYGbtgUOAxWbW2MyaBvObAGcAc0OOV9KcurtFohN6p7eZ9QceIJac\nxrj7HWY2hNiZxmgz+xbwJPCtYJM/uPvfzGx/4CVidYxc4Bl3v6OSY+gMo55Qd7dIatTkDEO3BpGM\nYoWxz7cShkjtpOOQlEjKqLtbJFo6w5CMoWd3i6SOzjAkq6m7WyRaShiSUdTdLRIdJQzJCOruFome\nEoZkBHV3i0RPCUMyhp7dLRItJQxJe+ruFkkPuqxW0p66u0VST5fVSlbSs7tF0kO1CcPM/m5mZ5qZ\nBpClzqm7WyR9VDskFdw88BKgF/B34El3/7QOYkuYhqSyl7q7RcIRypCUu7/u7ucTe9zqauAtM3vH\nzAaZWTo8gEmymLq7RdJHQjUMM2sJ/AwYBHwE/AnoC7weXmgiMeruFkkP1Z4hmNnzQHfgGeAn7r4i\nWPSMmc0MMzip39TdLZJeEhlSGg1MqKhI4O7HpD4kkRh1d4ukl0SGpA4EmpdOmFlLM7ssvJBE9lB3\nt0j6SCRhXO7uG0sn3H0DMDS8kETU3S2SjhK5rHaOu3ePm84BPnL3o8IOLlG6rDb7qLtbJFxhdXq/\naWZ/M7NTzOwUYsXvCUkE1d/MFpjZIjMbXsHyZmY2zsxmmdkcM7s40W0le6m7WyT9JFL0vgG4Arg2\nmH6T2GW11QrORh4C+gFfANPN7GV3XxC32pXAPHf/gZm1ARaa2V+BkgS2lSyk7m6R9FRtwnD3YuDB\n4CdZfYBP3H0ZgJmNBc4B4r/0HdgveL0fsM7dd5vZCQlsK1mox+M9APhgyAcRRyIi8RLpwzgQuA04\nAmhUOt/dD0lg/x2B5XHTK4glkXgPAePM7AugKXB+EttKFlJ3t0h6SmRI6kng98DdwPeI3VcqlQPM\nZwIz3f20IDm9aWY9kt1JQUFB2ev8/Hzy8/NTFqDUneFvxEpVzRo2izgSkexSVFREUVFRrfaRyFVS\nH7h77/irpcxshrsfW+3OY8NKBe7eP5geAbi7j4pb5xXgD+4+JZieCAwnlsyq3DZuH7pKKktYYeyi\nDV0dJRKusK6S2hEUrz8zs8vN7Gz21ByqMx04yMy6mllDYCAwrtw6y4DTAcysPXAIsDjBbSWLtBnV\nBoABBwyIOBIRqUgiQ1LXAk2Aq4jVMpoBlyayc3cvNrNhwHhiyWmMu883syGxxT6a2HDXk2b2UbDZ\nje6+HqCibRP/1SSTbNmyhXVfrwPglUGvRByNiFSkyiEpM2sA3ObuI+oupORpSCrzlQ5FLRi8gEP/\n59CIoxHJfikfkgouqT21VlGJVKO00L1Pg32ULETSWCJF70eADsDzwLbS+e6eNvUEnWFkNhW6Repe\nTc4wEqlh7EcsUXw/bp6jArSkgArdIpmj2jOMTKAzjMy0ZcsWmt0b67fQ2YVI3QrlDMPMRlc03931\nTAypldJksWCw7vYikgkSGZKaGPe6EfAj9r5lh0jSVOgWyTxJD0kFTXyT3b1vOCElT0NSmUeFbpFo\nhdXpXd7+QPsabCcCqNAtkqkSqWFsYM/NBnOA9UBaN/JJ+lJHt0jmSqSG0SbudYnGfqQ2VOgWyVyJ\nDEkNAJq6e7G7u5m1MLOzwg5Mso8K3SKZLZFO71nu3rPcvJnufkyokSVBRe/MoEK3SPoIq+hd0Q4T\nGcoSKaNCt0jmSyRhzDSzO4PnUnQ1s7uAmWEHJtlDhW6R7JBIwhgWrPcy8E9iV0xdEWZQkl1U6BbJ\nDtUOLbn7VuDXdRCLZCEVukWyR7VnGGb2upm1iJtuaWavhhuWZIs7p94JwNe/+zriSESkthIZkmrv\n7htLJ9x9A/A/4YUk2UKFbpHskkjCKDGzTqUTZtYlxHgkS6jQLZJ9Erk89mZgiplNInaJbT4wNNED\nmFl/4H5iyWmMu48qt/zXwM+JFdPzgMOBNu6+0cyWApuAEmCXu/dJ9LgSLRW6RbJPQnerNbP2wLeD\nyf+4+5qEdh67s+0ioB/wBTAdGOjuFX6LBB3k17j76cH0YqB3MAxW1XHUuJdGhr8xnDun3sk+DfZR\n7UIkTYV5t9qvgc+BNcBBZpborc37AJ+4+zJ33wWMBc6pYv0LgL/FTVsSMUqaUKFbJDslcpXUpcB/\ngEnAqODf2xPcf0f2ftjSimBeRcfZF+gPvBg324E3zWy6mQ1O8JgSoRZ3xC6oU6FbJPskUsO4FjgW\neM/dTzazI4FbQojlbGIPZtoYN+9Ed19lZm2JJY757j65oo0LCgrKXufn55Ofnx9CiFKdTTs2ASp0\ni6SboqIiioqKarWPRG4+ON3djzOzWUAfd99pZnPd/ahqd252AlDg7v2D6RGAly98B8v+ATzn7mMr\n2ddIYIu731vBMtUw0oQVGvvm7sv2326POhQRqUJYNYxVQePev4A3zOxFYkNLiZhOrObR1cwaAgOB\nceVXMrPmwCnEbj9SOq+xmTUNXjcBzgDmJnhciUDf0bHS1oyLZkQciYiEIZFbg/wgeHmTmfUDmgMJ\ndXq7e7GZDQPGs+ey2vlmNiS22EcHq/4QeMPdv4rbvD3wkpl5EOcz7j4+od9KIvHeqvcAOKLTERFH\nIiJhSOiy2nSnIan0oOddiGSOMC+rFUlIq0atog5BREKihCEpcdADBwGwZOiSiCMRkbAoYUhKfLbx\nMwCaNWsWcSQiEpZqi95mtoFYA128TcAM4AZ3XxpCXCIikmYSadx7GFgFPBtMXwB0A2YDTwCnhhKZ\nZIzNmzcD0Llp54gjEZEwJdK4N9vdjy43b5a796xoWRR0lVS02t/VnjXb17Dp2k0akhLJEGFdJfWV\nmf047iA/BnYEkyXJHEyy05rtsZsXK1mIZLdEEsaFwGAzW29m64DBwCAzawxcE2p0IiKSNhLp9P4U\n+F4li99ObTiSaVZsjt0l5qg21d5aTEQyXCI1jDbApcQK3WUJxt0vCzWyJKiGEZ1mtzdjy64t6u4W\nyTA1qWEkcpXUy8BUYDJQXJPAJHtt2bUl6hBEpI4kkjCauPv1oUciIiJpLZGi97/N7IzQI5GMM3XF\nVAD6de0XcSQiUhcSqWFsIHZL8+3ATmLP2XZ3T5u7zKmGEY1GtzZiR8kO1S9EMlBYNYw2NYxHstyO\nkh3VryQiWaPShGFmB7v7J8CRlazyUTghiYhIOqrqDGME8Eti95Iqz4HvhBKRZIRxH8eetHtpj0sj\njkRE6koiNYw8d99V3bwoqYZR93JvyaXYi1W/EMlQYd1L6v0E50k9UuxqyRGpbypNGGbWzsyOBvY1\ns+5m1iP4OQlonOgBzKy/mS0ws0VmNryC5b82s5lm9qGZzTGz3WbWIpFtJVpGUn+ciEiGq6qGMYDY\nLUE6EatjlH47bAFuSmTnZpYDPAT0A74AppvZy+6+oHQdd78buDtY/yzgGnffmMi2Eo17p9wLwO9O\n+l3EkYhIXUqkhnGeuz9Xo52bnQCMdPfvBdMjiPVwjKpk/WeASe4+JpltVcOoWzmFOTiu+oVIBgur\nhtHOzJoFB3jMzKaZWaKtvR2B5XHTK4J532Bm+wL9gReT3Vbqln/jib0iUh8k0rh3mbs/FNwe5FvE\nnofxF6B3imM5G5js7htrsnFBQUHZ6/z8fPLz81MTlVQoJ6G/NUQkXRQVFVFUVFSrfSSSMEr/nPw+\n8LS7zw7qC4lYCXSJm+4UzKvIQOBvNdx2r4Qh4bli3BUAPH724xFHIiLJKP+HdGFhYdL7SKSG8TSx\n24McAvQgNoz1jrv3qnbnZg2AhcQK16uAacAF7j6/3HrNgcVAJ3f/Kpltg3VVw6gjVhgb8lT9QiSz\nhXUvqUuIDT996u7bgwcq/TKRnbt7sZkNA8YTSzRj3H2+mQ2JLfbRwao/BN4oTRZVbZvwbyYiIilV\n7RkGgJkNBA5099vMrDPQzt0/CD26BOkMo+5YoZGXk8fOm3ZGHYqI1EIoV0mZ2UPAqcCFwaxtwGPJ\nhyeZ7qd/+ykA/x7474gjEZEoJDIk1dfde5nZTAB3X29mDUOOS9LQi4tiVzz3O1gPTBKpjxK52mlX\ncFWUA5hZa6Ak1KhERCTtVHUvqdKzj4eJNdO1NbNCYDJQYae2ZL99c/eNOgQRiUhVQ1LTgF7u/rSZ\nfQCcTux+Uue6+9w6iU7SRt/RfQGYcdGMiCMRkahUlTDKqufuPg+YF344kq7eW/UeAEd0OiLiSEQk\nKlUljLZmdl1lC9393hDiERGRNFVVwmgANAU99EBiWjVqFXUIIhKhqhLGKne/pc4ikbR14P0HArBk\n6JKIIxGRKFV1Wa3OLASAxZsWA9CsWbOIIxGRKFWVMNSdJSIiZSpNGO6+vi4DkfS0efNmADo37Rxx\nJCISNT0FR6p08J8OBmDuYLXeiNR3ShhSpTXb1wCqX4iIEoaIiCRICUMqtWLzCgCOanNUxJGISDpI\n6AFK6U4PUApHs9ubsWXXFj2OVSQLhfIAJam/tuzaEnUIIpJGlDBERCQhoScMM+tvZgvMbJGZDa9k\nnXwzm2lmc83srbj5S81sdrBsWtixyh5TV0wF4PSup0cciYiki1BrGMGT+hYR6xr/ApgODHT3BXHr\nNAf+A5zh7ivNrI27rw2WLQZ6u/uGao6jGkaKNbq1ETtKdqh+IZKl0rGG0Qf4xN2XufsuYCxwTrl1\nfga86O4rAUqTRcDqIEapwI6SHVGHICJpJuwv447A8rjpFcG8eIcArczsLTObbmaD4pY58GYwf3DI\nsYqISBWqur15XckFegGnAU2A98zsPXf/FDjR3VeZWVtiiWO+u0+uaCcFBQVlr/Pz88nPzw898Gw1\n7uNxAFzWFDzPAAALeklEQVR69KURRyIiqVJUVERRUVGt9hF2DeMEoMDd+wfTIwB391Fx6wwHGrl7\nYTD9OPBvd3+x3L5GAlsqetKfahiplXtLLsVerPqFSBZLxxrGdOAgM+tqZg2BgcC4cuu8DJxkZg3M\nrDFwPDDfzBqbWVMAM2sCnAHoDnh1oNiLow5BRNJQqENS7l5sZsOA8cSS0xh3n29mQ2KLfbS7LzCz\nN4CPgGJgtLt/bGb7Ay+ZmQdxPuPu48OMV/YwPT9LRMoJvYbh7q8Dh5ab96dy03cDd5ebtwToGXZ8\nsrd7p8RG/H530u8ijkRE0o3uJSV7ySnMwXHVL0SyXDrWMCTDOEoUIlIxJQz5hhx9LESkAvpmkDJD\nXx4KwONnPx5xJCKSjlTDEAA2bdpEi/tbAKh+IVIPqIYhNVaaLC7toe5uEamYzjAEK4z9kdG7XW9m\nDJ0RcTQiUhdqcoahhFHPlSaL1vu2Zu2Na6tZW0SyhYakJCm5t8T6NvMsT8lCRKqlhFFPtbqjVdk9\no3bevDPiaEQkEyhh1EPdH+7Ohh2xhxjqiigRSZQSRj1z4YsXMndt7Ka/ShYikgwljHpk9PTRPDP3\nGQA2XrMx4mhEJNPoKql6Ys7nc+jxRA8All2zjC7Nu0QckYhESVdJSYU2bdpUliyePvtpJQsRqREl\njHogvot7UK9BEUcjIplKQ1JZTl3cIlIRdXrLXtTFLSKVScsahpn1N7MFZrbIzIZXsk6+mc00s7lm\n9lYy20rF1MUtIqkW6hmGmeUAi4B+wBfAdGCguy+IW6c58B/gDHdfaWZt3H1tItvG7UNnGHFa3dFK\njXkiUqV0PMPoA3zi7svcfRcwFjin3Do/A15095UA7r42iW2lHHVxi0hYwk4YHYHlcdMrgnnxDgFa\nmdlbZjbdzAYlsa3EURe3iIQpN+oAiMXQCzgNaAK8Z2bvRRtS5nl46sPq4haRUIWdMFYC8V1inYJ5\n8VYAa939a+BrM3sHODrBbcsUFBSUvc7Pzyc/P782cWeUWUtnMeyNYQAsuXoJzZs3jzgiEUk3RUVF\nFBUV1WofYRe9GwALiRWuVwHTgAvcfX7cOocBDwL9gX2A94Hzg+2q3DZuH/W26L1x40ZaPtASgDHf\nH8Olx+kRqyJSvZoUvUM9w3D3YjMbBownVi8Z4+7zzWxIbLGPdvcFZvYG8BFQDIx2948BKto2zHgz\nUWmyuPDIC5UsRCRUatzLYKWNeT1a92D2sNkRRyMimUSd3vVIabJosU8LNozYEHE0IpJp0rEPQ0LQ\noLBB7F8aKFmISJ1RwsgwzW9vTgklAOweuTviaESkPlHCyCCH/fEwNu/aDKgxT0TqnhJGhjjvufNY\nuGEhoGQhItFQwsgAD099mOfnPw/AhqtVsxCRaOgqqTQ3a+ksjnnqGCDWxd2tRbdoAxKRrKCrpLLM\nxo0by5LFmO+PUbIQkUgpYaQxdXGLSDrRkFSaUhe3iIRJnd5ZQl3cIhI21TCygLq4RSRdKWGkEXVx\ni0g6U8JIAXdn+67ttdrHkQ8eqS5uEUlr6fCI1pQ4/k/H8/6Q9+v8uDmFOTip+4JXshCRdJU1RW8K\nYq/v7Xcv1550bejHbH1Ha9bvWF82/f7P36fPQX1CP66ISCrU66ukGt7SkJ0lO8vmrR22ltatW6f8\nWGc+dSbjl44vm76q91U8cNYDKT+OiEiY6nXCKP09Si9JLZWqIZ4nP3ySS/51Sdn0IS0PYeFVC1Oy\nbxGRuqaEEVi0bhGHPnRo2XST3CZs/e3WGu17/fr1tH5wz5lKnuWx8+adVWwhIpL+0jJhmFl/4H5i\nV2SNcfdR5ZafArwMLA5m/cPdfx8sWwpsAkqAXe5eYZGgssa9347/Lbe/d3vZdJ8OfZIqjJcvaKsg\nLSLZIu0a98wsB3gIOBM4ErjAzA6rYNV33L1X8PP7uPklQL67H1NZsqjKbWfcho90Dml5CADTVk/D\nCo37Jt9X5Xat72iNFVpZsnj/5+/Xm2RRVFQUdQhZRe9naun9jFbYfRh9gE/cfZm77wLGAudUsF5l\nWc5IQYwLr1qIj3Qa5jQE4LqJ12GFxrp16/Za78ynzsQKrezqp6t6X4WP9Hp19ZP+h0wtvZ+ppfcz\nWmH3YXQElsdNryCWRMr7tpnNAlYCN7j7x8F8B940s2JgtLv/uTbB7LhpB7CnMN7moTYAPHH2Eypo\ni4hUIx0a9z4Aurj7djP7HvBP4JBg2YnuvsrM2hJLHPPdfXJtD+gjfa/CeGmyUEFbRKRyoRa9zewE\noMDd+wfTIwAvX/gut80SoLe7ry83fySwxd3vrWCb+lFgEBFJoWSL3mGfYUwHDjKzrsAqYCBwQfwK\nZtbe3b8MXvchlsTWm1ljIMfdt5pZE+AMoLCigyT7S4uISPJCTRjuXmxmw4Dx7Lmsdr6ZDYkt9tHA\nT81sKLAL+Ao4P9i8PfBScPaQCzzj7uO/eRQREakLWdG4JyIi4cvo25ubWX8zW2Bmi8xseNTxZDoz\nW2pms81spplNizqeTGNmY8zsSzP7KG5eSzMbb2YLzewNM2seZYyZopL3cqSZrTCzD4Of/lHGmEnM\nrJOZTTKzeWY2x8yuCuYn9fnM2ISRRFOgJK5WjZLCE8Q+j/FGABPc/VBgEvC/dR5VZqrovQS4N67J\n9/W6DiqD7Qauc/cjgW8DVwbfl0l9PjM2YZB4U6AkLiWNkvVVcMl3+efqngM8Fbx+CvhhnQaVoSp5\nL6HyJl+pgruvdvdZweutwHygE0l+PjP5y6GipsCOEcWSLUobJaeb2eCog8kS7UqvAnT31UC7iOPJ\ndMPMbJaZPa7hvZoxs25AT2Aq0D6Zz2cmJwxJvRPdvRfwfWKnrCdFHVAW0lUmNfcIcIC79wRWA9/o\nyZKqmVlT4AXg6uBMo/znscrPZyYnjJVAl7jpTsE8qSF3XxX8+1/gJSq+jYsk50szaw9gZh2ANRHH\nk7Hc/b9xt6X+M3BclPFkGjPLJZYs/s/dXw5mJ/X5zOSEUdYUaGYNiTUFjos4poxlZo2Dvz6Ia5Sc\nG21UGcnYe5x9HHBx8PoiYrfyl8Ts9V4GX2ilfow+n8n6C/Cxu8c/IjSpz2dG92EEl9U9wJ6mwDsi\nDiljmdn+xM4q4hsl9X4mwcyeBfKB1sCXwEhi90Z7HugMLAPOc/eNUcWYKSp5L08lNvZeAiwFhpSO\nv0vVzOxE4B1gDrH/xx34DTANeI4EP58ZnTBERKTuZPKQlIiI1CElDBERSYgShoiIJEQJQ0REEqKE\nISIiCVHCEBGRhChhiFTCzIqD22jPDP69MYX77mpmc1K1P5G6EPYjWkUy2bbg3lphUROUZBSdYYhU\nrsJbaZvZEjMbZWYfmdlUMzsgmN/VzCYGd1N908w6BfPbmdk/gvkzzeyEYFe5ZjbazOaa2etmtk+w\n/lXBg25mBR3PImlBCUOkcvuWG5I6N27ZBnfvATxM7PY0AA8CTwR3U302mAb4I1AUzO8FzAvmHww8\n6O5HAZuAnwTzhwM9g/UvD+uXE0mWbg0iUgkz2+zuzSqYvwQ41d2XBncAXeXubc3sv0AHdy8O5n/h\n7u3MbA3QMXjQV+k+ugLjgyedEdRHct39djN7DdhG7D5U/3T3beH/tiLV0xmGSM14Ja+TsSPudTF7\naooDiD1+uBcwPXgcsUjk9EEUqVxVjwM9P/h3IPBe8HoKcEHw+kLg3eD1BOAKiD2L3sxKz1oq238X\nd3+b2POWmwFNkw9dJPV0lZRI5RqZ2YfEvtgdeN3dfxMsa2lms4Gv2ZMkrgKeMLNfA/8FLgnmXwOM\nNrNfAruBocSeGPeNM5NgKOuvQVIx4AF33xzKbyeSJNUwRJIU1DB6u/v6qGMRqUsakhJJnv7KknpJ\nZxgiIpIQnWGIiEhClDBERCQhShgiIpIQJQwREUmIEoaIiCRECUNERBLy/wHNdmvfhSDYIQAAAABJ\nRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f57779c4490>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "iterations = 20\n",
    "trainLoss = []\n",
    "testAcc = []\n",
    "start = time.time()\n",
    "for epoch in range(iterations):\n",
    "    epochStart = time.time()\n",
    "    runningLoss = 0    \n",
    "    net.train(True) # For training\n",
    "    for data in trainLoader:\n",
    "        inputs,labels = data\n",
    "        # Wrap them in Variable\n",
    "        if use_gpu:\n",
    "            inputs, labels = Variable(inputs.float().cuda()), \\\n",
    "                Variable(labels.long().cuda())\n",
    "        else:\n",
    "            inputs, labels = Variable(inputs), Variable(labelslong()) \n",
    "        inputs = inputs/255.0\n",
    "        # Initialize gradients to zero\n",
    "        optimizer.zero_grad()\n",
    "        # Feed-forward input data through the network\n",
    "        outputs = net(inputs)\n",
    "        # Compute loss/error\n",
    "        loss = criterion(outputs, labels)        \n",
    "        # Backpropagate loss and compute gradients\n",
    "        loss.backward()\n",
    "        # Update the network parameters\n",
    "        optimizer.step()\n",
    "        # Accumulate loss per batch\n",
    "        runningLoss += loss.data[0]    \n",
    "    avgTrainLoss = runningLoss/200\n",
    "    trainLoss.append(avgTrainLoss)\n",
    "    # Evaluating performance on test set for each epoch\n",
    "    net.train(False) # For testing\n",
    "    inputs = TestImages/255\n",
    "    if use_gpu:\n",
    "        inputs = Variable(inputs.cuda())\n",
    "        outputs = net(inputs)\n",
    "        _, predicted = torch.max(outputs.data, 1)\n",
    "        predicted = predicted.cpu()\n",
    "    else:\n",
    "        inputs = Variable(inputs)\n",
    "        outputs = net(inputs)\n",
    "        _, predicted = torch.max(outputs.data, 1)\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    total += TestLabels.size(0)\n",
    "    correct += (predicted == TestLabels).sum()\n",
    "    avgTestAcc = correct/60.0\n",
    "    testAcc.append(avgTestAcc)\n",
    "        \n",
    "    # Plotting Loss vs Epochs\n",
    "    fig1 = plt.figure(1)        \n",
    "    plt.plot(range(epoch+1),trainLoss,'r--',label='train')        \n",
    "    if epoch==0:\n",
    "        plt.legend(loc='upper left')\n",
    "        plt.xlabel('Epochs')\n",
    "        plt.ylabel('Loss')    \n",
    "    # Plotting testing accuracy vs Epochs\n",
    "    fig2 = plt.figure(2)        \n",
    "    plt.plot(range(epoch+1),testAcc,'g-',label='test')        \n",
    "    if epoch==0:\n",
    "        plt.legend(loc='upper left')\n",
    "        plt.xlabel('Epochs')\n",
    "        plt.ylabel('Testing accuracy')    \n",
    "    \n",
    "    epochEnd = time.time()-epochStart\n",
    "    print('At Iteration: {:.0f} /{:.0f}  ;  Training Loss: {:.6f} ; Testing Acc: {:.3f} ; Time consumed: {:.0f}m {:.0f}s '\\\n",
    "          .format(epoch + 1,iterations,avgTrainLoss,avgTestAcc*100,epochEnd//60,epochEnd%60))\n",
    "end = time.time()-start\n",
    "print('Training completed in {:.0f}m {:.0f}s'.format(end//60,end%60))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}