vision_classifier.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torchvision\n",
    "import torchvision.transforms as transforms\n",
    "import torch.optim as optim\n",
    "import torch.nn as nn\n",
    "from torchvision import models\n",
    "from tqdm import tqdm\n",
    "from sklearn.metrics import confusion_matrix\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import itertools\n",
    "import numpy as np\n",
    "\n",
    "torch.manual_seed(0)\n",
    "np.random.seed(0)\n",
    "\n",
    "torch.backends.cudnn.deterministic = True\n",
    "torch.backends.cudnn.benchmark = False\n",
    "\n",
    "# font = {'weight' : 'bold',\n",
    "#         'size'   : 90}\n",
    "\n",
    "# plt.rc('font', **font)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 32\n",
    "IMG_SIZE = [224,224]\n",
    "num_classes = 42\n",
    "device = \"cuda\"\n",
    "target_names = [str(i) for i in range(42)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "105392\n"
     ]
    }
   ],
   "source": [
    "transform = transforms.Compose(\n",
    "    [transforms.Resize(IMG_SIZE),\n",
    "     transforms.ToTensor(),\n",
    "     transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])\n",
    "\n",
    "trainvalset = torchvision.datasets.ImageFolder(\"train/train/\",transform=transform)\n",
    "print(len(trainvalset))\n",
    "\n",
    "trainset, valset = torch.utils.data.random_split(trainvalset, [len(trainvalset)-30000, 30000])\n",
    "\n",
    "trainloader = torch.utils.data.DataLoader(trainset, batch_size=batch_size,\n",
    "                                          shuffle=True, num_workers=1)\n",
    "\n",
    "valloader = torch.utils.data.DataLoader(valset, batch_size=batch_size,\n",
    "                                          shuffle=True, num_workers=1)\n",
    "\n",
    "testset = torchvision.datasets.ImageFolder(\"test/\",transform=transform)\n",
    "\n",
    "testloader = torch.utils.data.DataLoader(testset, batch_size=batch_size,\n",
    "                                         shuffle=False, num_workers=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([14, 40,  7, 21, 25, 11, 41, 21,  8,  8, 32, 15, 34,  8,  9, 22, 12, 14,\n",
      "         1, 25, 14,  0,  5, 25,  3, 26, 26, 22,  3, 32,  5, 31])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def imshow(img):\n",
    "    img = img / 2 + 0.5     # unnormalize\n",
    "    npimg = img.numpy()\n",
    "    plt.imshow(np.transpose(npimg, (1, 2, 0)))\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "# get some random training images\n",
    "dataiter = iter(trainloader)\n",
    "images, labels = dataiter.next()\n",
    "print(labels)\n",
    "\n",
    "# show images\n",
    "imshow(torchvision.utils.make_grid(images))\n",
    "# print labels\n",
    "#print(' '.join('%5s' % classes[labels[j]] for j in range(4)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "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, track_running_stats=True)\n",
      "  (relu): ReLU(inplace=True)\n",
      "  (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
      "  (layer1): Sequential(\n",
      "    (0): Bottleneck(\n",
      "      (conv1): Conv2d(64, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "      (downsample): Sequential(\n",
      "        (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "        (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "    )\n",
      "    (1): Bottleneck(\n",
      "      (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (2): Bottleneck(\n",
      "      (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(128, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "  )\n",
      "  (layer2): Sequential(\n",
      "    (0): Bottleneck(\n",
      "      (conv1): Conv2d(256, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv3): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=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, track_running_stats=True)\n",
      "      )\n",
      "    )\n",
      "    (1): Bottleneck(\n",
      "      (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (2): Bottleneck(\n",
      "      (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (3): Bottleneck(\n",
      "      (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(256, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "  )\n",
      "  (layer3): Sequential(\n",
      "    (0): Bottleneck(\n",
      "      (conv1): Conv2d(512, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "      (downsample): Sequential(\n",
      "        (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "        (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "    )\n",
      "    (1): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (2): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (3): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (4): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (5): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (6): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (7): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (8): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (9): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (10): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (11): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (12): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (13): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (14): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (15): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (16): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (17): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (18): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (19): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (20): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (21): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (22): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\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, track_running_stats=True)\n",
      "      (conv3): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "  )\n",
      "  (layer4): Sequential(\n",
      "    (0): Bottleneck(\n",
      "      (conv1): Conv2d(1024, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(1024, 1024, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv3): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "      (downsample): Sequential(\n",
      "        (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "        (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "    )\n",
      "    (1): Bottleneck(\n",
      "      (conv1): Conv2d(2048, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(1024, 1024, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv3): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "    (2): Bottleneck(\n",
      "      (conv1): Conv2d(2048, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(1024, 1024, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv3): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "      (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "    )\n",
      "  )\n",
      "  (avgpool): AdaptiveAvgPool2d(output_size=(1, 1))\n",
      "  (fc): Linear(in_features=2048, out_features=42, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "\n",
    "#model = models.resnet18(pretrained=True)\n",
    "model = torchvision.models.wide_resnet101_2(pretrained=True)\n",
    "num_ftrs = model.fc.in_features\n",
    "model.fc = nn.Linear(num_ftrs, num_classes)\n",
    "#model.load_state_dict(torch.load(\"wide_resnet101.pth\"))\n",
    "model.to(device)\n",
    "\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = optim.SGD(model.parameters(), lr=0.0001, momentum=0.9)\n",
    "print(model)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_confusion_matrix(cm,\n",
    "                          target_names,\n",
    "                          title='Confusion matrix',\n",
    "                          cmap=None,\n",
    "                          normalize=True):\n",
    "    \"\"\"\n",
    "    given a sklearn confusion matrix (cm), make a nice plot\n",
    "\n",
    "    Arguments\n",
    "    ---------\n",
    "    cm:           confusion matrix from sklearn.metrics.confusion_matrix\n",
    "\n",
    "    target_names: given classification classes such as [0, 1, 2]\n",
    "                  the class names, for example: ['high', 'medium', 'low']\n",
    "\n",
    "    title:        the text to display at the top of the matrix\n",
    "\n",
    "    cmap:         the gradient of the values displayed from matplotlib.pyplot.cm\n",
    "                  see http://matplotlib.org/examples/color/colormaps_reference.html\n",
    "                  plt.get_cmap('jet') or plt.cm.Blues\n",
    "\n",
    "    normalize:    If False, plot the raw numbers\n",
    "                  If True, plot the proportions\n",
    "\n",
    "    Usage\n",
    "    -----\n",
    "    plot_confusion_matrix(cm           = cm,                  # confusion matrix created by\n",
    "                                                              # sklearn.metrics.confusion_matrix\n",
    "                          normalize    = True,                # show proportions\n",
    "                          target_names = y_labels_vals,       # list of names of the classes\n",
    "                          title        = best_estimator_name) # title of graph\n",
    "\n",
    "    Citiation\n",
    "    ---------\n",
    "    http://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html\n",
    "\n",
    "    \"\"\"\n",
    "\n",
    "    accuracy = np.trace(cm) / float(np.sum(cm))\n",
    "    misclass = 1 - accuracy\n",
    "\n",
    "    if cmap is None:\n",
    "        cmap = plt.get_cmap('Blues')\n",
    "\n",
    "    plt.figure(figsize=(80, 60))\n",
    "    plt.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    plt.title(title)\n",
    "    plt.colorbar()\n",
    "    #plt.rc('xtick', labelsize=20) \n",
    "    #plt.rc('ytick', labelsize=20) \n",
    "\n",
    "    if target_names is not None:\n",
    "        tick_marks = np.arange(len(target_names))\n",
    "        plt.xticks(tick_marks, target_names, rotation=45,fontsize=50)\n",
    "        plt.yticks(tick_marks, target_names,fontsize=50)\n",
    "\n",
    "    if normalize:\n",
    "        cm = cm.astype('float') / cm.sum(axis=1)[:, np.newaxis]\n",
    "\n",
    "\n",
    "    thresh = cm.max() / 1.5 if normalize else cm.max() / 2\n",
    "    for i, j in itertools.product(range(cm.shape[0]), range(cm.shape[1])):\n",
    "        if normalize:\n",
    "            plt.text(j, i, \"{:0.4f}\".format(cm[i, j]),\n",
    "                     horizontalalignment=\"center\",\n",
    "                     color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "        else:\n",
    "            plt.text(j, i, \"{:,}\".format(cm[i, j]),\n",
    "                     horizontalalignment=\"center\",\n",
    "                     color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "\n",
    "\n",
    "    plt.tight_layout()\n",
    "    plt.ylabel('True label')\n",
    "    plt.xlabel('Predicted label\\naccuracy={:0.4f}; misclass={:0.4f}'.format(accuracy, misclass))\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate(net):\n",
    "    net.eval()\n",
    "    correct = 0\n",
    "    total = 0\n",
    "    y_true = []\n",
    "    y_pred = []\n",
    "    with torch.no_grad():\n",
    "        pbar = tqdm(valloader)\n",
    "        i_batch = 0\n",
    "        for data in pbar:\n",
    "            images, labels = data\n",
    "            images = images.to(device)\n",
    "            labels = labels.to(device)\n",
    "            outputs = net(images)\n",
    "            _, predicted = torch.max(outputs.data, 1)\n",
    "            \n",
    "            for output,label in zip(predicted,labels):\n",
    "                y_pred.append(output.item())\n",
    "                y_true.append(label.item())\n",
    "            total += labels.size(0)\n",
    "            correct += (predicted == labels).sum().item()\n",
    "            pbar.set_description(\"Val Accuracy %.2f\" % (correct/float(total)))\n",
    "            i_batch+=1\n",
    "#             if(i_batch==10):\n",
    "#                 break\n",
    "\n",
    "    print('Accuracy of the network on the 10000 test images: %d %%' % (\n",
    "        100 * correct / total))\n",
    "    \n",
    "    cm = confusion_matrix(y_true, y_pred)\n",
    "    \n",
    "    plot_confusion_matrix(cm= cm, \n",
    "                      normalize    = True,\n",
    "                      target_names = target_names,\n",
    "                      title        = \"Confusion Matrix\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Started Training\n",
      "[1,    20] loss: 3.776\n",
      "[1,    40] loss: 3.761\n",
      "[1,    60] loss: 3.735\n",
      "[1,    80] loss: 3.693\n",
      "[1,   100] loss: 3.678\n",
      "[1,   120] loss: 3.659\n",
      "[1,   140] loss: 3.625\n",
      "[1,   160] loss: 3.621\n",
      "[1,   180] loss: 3.603\n",
      "[1,   200] loss: 3.582\n",
      "[1,   220] loss: 3.557\n",
      "[1,   240] loss: 3.527\n",
      "[1,   260] loss: 3.504\n",
      "[1,   280] loss: 3.505\n",
      "[1,   300] loss: 3.473\n",
      "[1,   320] loss: 3.450\n",
      "[1,   340] loss: 3.414\n",
      "[1,   360] loss: 3.412\n",
      "[1,   380] loss: 3.395\n",
      "[1,   400] loss: 3.355\n",
      "[1,   420] loss: 3.345\n",
      "[1,   440] loss: 3.320\n",
      "[1,   460] loss: 3.293\n",
      "[1,   480] loss: 3.253\n",
      "[1,   500] loss: 3.221\n",
      "[1,   520] loss: 3.161\n",
      "[1,   540] loss: 3.162\n",
      "[1,   560] loss: 3.143\n",
      "[1,   580] loss: 3.145\n",
      "[1,   600] loss: 3.127\n",
      "[1,   620] loss: 3.089\n",
      "[1,   640] loss: 3.083\n"
     ]
    }
   ],
   "source": [
    "print(\"Started Training\")\n",
    "for epoch in range(2):  # loop over the dataset multiple times\n",
    "    running_loss = 0.0\n",
    "    for i, data in enumerate(trainloader, 0):\n",
    "        # get the inputs; data is a list of [inputs, labels]\n",
    "        inputs, labels = data\n",
    "        inputs = inputs.to(device)\n",
    "        labels = labels.to(device)\n",
    "\n",
    "        # zero the parameter gradients\n",
    "        optimizer.zero_grad()\n",
    "\n",
    "        # forward + backward + optimize\n",
    "        outputs = model(inputs)\n",
    "        loss = criterion(outputs, labels)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        # print statistics\n",
    "        running_loss += loss.item()\n",
    "        if i % 20 == 19:    # print every 2000 mini-batches\n",
    "            print('[%d, %5d] loss: %.3f' %\n",
    "                  (epoch + 1, i + 1, running_loss / 20))\n",
    "            running_loss = 0.0\n",
    "        #if i % 200 == 199:\n",
    "    evaluate(model)\n",
    "    model.train()\n",
    "    torch.save(model.state_dict(), \"wide_resnet101.pth\")\n",
    "\n",
    "print('Finished Training')"
   ]
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}