Transfer Learning Comparisons for Apple Disease Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Import The Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import time\n",
    "import copy\n",
    "import os\n",
    "import torch\n",
    "import torch.optim as optim\n",
    "import torch.nn as nn\n",
    "import torchvision\n",
    "import matplotlib.pyplot as plt\n",
    "from torch.optim import lr_scheduler\n",
    "from torchvision import datasets, models, transforms"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "transforms = {\n",
    "    'train': transforms.Compose([\n",
    "        transforms.RandomResizedCrop(224),\n",
    "        transforms.RandomHorizontalFlip(),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "    ]),\n",
    "    'val': transforms.Compose([\n",
    "        transforms.Resize(256),\n",
    "        transforms.CenterCrop(224),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "    ]),\n",
    "    'test': transforms.Compose([\n",
    "        transforms.Resize(256),\n",
    "        transforms.CenterCrop(224),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "    ])\n",
    "}\n",
    "\n",
    "data_dir = 'data'\n",
    "image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x), transform=transforms[x])\n",
    "                  for x in ['train', 'val', 'test']}\n",
    "dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=4, shuffle=True, num_workers=4)\n",
    "               for x in ['train', 'val', 'test']}\n",
    "data_size = {x: len(image_datasets[x]) for x in ['train', 'val', 'test']}\n",
    "class_names = image_datasets['train'].classes\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataiter = iter(dataloaders['train'])\n",
    "images, labels = dataiter.next()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def imshow(inp, title=None):\n",
    "    inp = inp.numpy().transpose((1, 2, 0))\n",
    "    mean = np.array([0.485, 0.456, 0.406])\n",
    "    std = np.array([0.229, 0.224, 0.225])\n",
    "    inp = std * inp + mean\n",
    "    inp = np.clip(inp, 0, 1)\n",
    "    plt.imshow(inp)\n",
    "    if title is not None:\n",
    "        plt.title(title)\n",
    "    plt.pause(0.001)\n",
    "    \n",
    "out = torchvision.utils.make_grid(images)\n",
    "imshow(out, title=[class_names[x] for x in labels])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create a function to train and to evaluate the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_model(model, criterion, optimizer, scheduler, num_epochs=25):\n",
    "    since = time.time()\n",
    "    \n",
    "    best_model_wts = copy.deepcopy(model.state_dict())\n",
    "    best_acc = 0.0\n",
    "    \n",
    "    for epoch in range(num_epochs):\n",
    "        print('Epoch {}/{}'.format(epoch, num_epochs - 1))\n",
    "        print('-' * 10)\n",
    "        \n",
    "        for phase in ['train', 'val']:\n",
    "            if phase == 'train':\n",
    "                model.train()\n",
    "            else:\n",
    "                model.eval()\n",
    "                \n",
    "            running_loss = 0.0\n",
    "            running_corrects = 0\n",
    "            \n",
    "            for inputs, labels in dataloaders[phase]:\n",
    "                inputs = inputs.to(device)\n",
    "                labels = labels.to(device)\n",
    "                \n",
    "                optimizer.zero_grad()\n",
    "                \n",
    "                with torch.set_grad_enabled(phase == 'train'):\n",
    "                    outputs = model(inputs)\n",
    "                    _, preds = torch.max(outputs, 1)\n",
    "                    loss = criterion(outputs, labels)\n",
    "                    \n",
    "                    if phase == 'train':\n",
    "                        loss.backward()\n",
    "                        optimizer.step()\n",
    "                        \n",
    "                running_loss += loss.item() * inputs.size(0)\n",
    "                running_corrects += torch.sum(preds == labels.data)\n",
    "                \n",
    "            if phase == 'train':\n",
    "                scheduler.step()\n",
    "                \n",
    "            epoch_loss = running_loss / data_size[phase]\n",
    "            epoch_acc = running_corrects.double() / data_size[phase]\n",
    "            \n",
    "            print('{} Loss: {:.4f} Acc: {:.4f}'.format(\n",
    "                phase, epoch_loss, epoch_acc))\n",
    "            \n",
    "            if phase == 'val' and epoch_acc > best_acc:\n",
    "                best_acc = epoch_acc\n",
    "                best_model_wts = copy.deepcopy(model.state_dict())\n",
    "        print()\n",
    "    \n",
    "    time_elapsed = time.time() - since\n",
    "    print('Training complete in {:.0f}m {:.0f}s'.format(\n",
    "        time_elapsed // 60, time_elapsed % 60))\n",
    "    print('Best val Acc: {:4f}'.format(best_acc))\n",
    "    \n",
    "    model.load_state_dict(best_model_wts)\n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Declare and Download (optional) the pretrained model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "model_resnet = models.resnet18(pretrained=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading: \"https://download.pytorch.org/models/vgg16-397923af.pth\" to C:\\Users\\Irfan Alghani Khalid/.cache\\torch\\checkpoints\\vgg16-397923af.pth\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "00c1c183688a44abb592923fcf57f286",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=553433881.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "model_vgg16 = models.vgg16(pretrained=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading: \"https://download.pytorch.org/models/inception_v3_google-1a9a5a14.pth\" to C:\\Users\\Irfan Alghani Khalid/.cache\\torch\\checkpoints\\inception_v3_google-1a9a5a14.pth\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "811f7f29da0543c2b567b8d2c1323827",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=108857766.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "model_inceptionv3 = models.inception_v3(pretrained=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading: \"https://download.pytorch.org/models/alexnet-owt-4df8aa71.pth\" to C:\\Users\\Irfan Alghani Khalid/.cache\\torch\\checkpoints\\alexnet-owt-4df8aa71.pth\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6b2e438475bd4dc681b61a61a01f5d13",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=244418560.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "model_alexnet = models.alexnet(pretrained=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Set the model, parameters to tune, criterion, oprimizer, and the weight update"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### VGG-16 Architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "for param in model_vgg16.parameters():\n",
    "    param.requires_grad = False\n",
    "\n",
    "num_ftrs = model_vgg16.classifier[6].in_features\n",
    "model_vgg16.classifier[6] = nn.Linear(num_ftrs, len(class_names))\n",
    "model_vgg16 = model_vgg16.to(device)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer_conv = optim.SGD(model_vgg16.classifier[6].parameters(), lr=0.001, momentum=0.9)\n",
    "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_conv, step_size=7, gamma=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0/24\n",
      "----------\n",
      "train Loss: 0.5082 Acc: 0.8164\n",
      "val Loss: 0.2406 Acc: 0.9216\n",
      "\n",
      "Epoch 1/24\n",
      "----------\n",
      "train Loss: 0.4444 Acc: 0.8444\n",
      "val Loss: 0.1936 Acc: 0.9490\n",
      "\n",
      "Epoch 2/24\n",
      "----------\n",
      "train Loss: 0.4170 Acc: 0.8576\n",
      "val Loss: 0.1470 Acc: 0.9529\n",
      "\n",
      "Epoch 3/24\n",
      "----------\n",
      "train Loss: 0.4029 Acc: 0.8637\n",
      "val Loss: 0.1736 Acc: 0.9333\n",
      "\n",
      "Epoch 4/24\n",
      "----------\n",
      "train Loss: 0.3557 Acc: 0.8852\n",
      "val Loss: 0.1350 Acc: 0.9529\n",
      "\n",
      "Epoch 5/24\n",
      "----------\n",
      "train Loss: 0.4106 Acc: 0.8681\n",
      "val Loss: 0.1764 Acc: 0.9333\n",
      "\n",
      "Epoch 6/24\n",
      "----------\n",
      "train Loss: 0.4169 Acc: 0.8637\n",
      "val Loss: 0.1326 Acc: 0.9608\n",
      "\n",
      "Epoch 7/24\n",
      "----------\n",
      "train Loss: 0.3500 Acc: 0.8926\n",
      "val Loss: 0.1248 Acc: 0.9647\n",
      "\n",
      "Epoch 8/24\n",
      "----------\n",
      "train Loss: 0.3416 Acc: 0.8887\n",
      "val Loss: 0.1123 Acc: 0.9529\n",
      "\n",
      "Epoch 9/24\n",
      "----------\n",
      "train Loss: 0.3084 Acc: 0.8913\n",
      "val Loss: 0.1182 Acc: 0.9529\n",
      "\n",
      "Epoch 10/24\n",
      "----------\n",
      "train Loss: 0.2984 Acc: 0.9040\n",
      "val Loss: 0.1153 Acc: 0.9647\n",
      "\n",
      "Epoch 11/24\n",
      "----------\n",
      "train Loss: 0.3051 Acc: 0.8931\n",
      "val Loss: 0.1204 Acc: 0.9647\n",
      "\n",
      "Epoch 12/24\n",
      "----------\n",
      "train Loss: 0.3155 Acc: 0.8940\n",
      "val Loss: 0.1173 Acc: 0.9647\n",
      "\n",
      "Epoch 13/24\n",
      "----------\n",
      "train Loss: 0.2958 Acc: 0.9058\n",
      "val Loss: 0.0918 Acc: 0.9647\n",
      "\n",
      "Epoch 14/24\n",
      "----------\n",
      "train Loss: 0.2994 Acc: 0.9005\n",
      "val Loss: 0.0967 Acc: 0.9647\n",
      "\n",
      "Epoch 15/24\n",
      "----------\n",
      "train Loss: 0.2948 Acc: 0.9032\n",
      "val Loss: 0.0987 Acc: 0.9647\n",
      "\n",
      "Epoch 16/24\n",
      "----------\n",
      "train Loss: 0.2838 Acc: 0.9040\n",
      "val Loss: 0.0978 Acc: 0.9647\n",
      "\n",
      "Epoch 17/24\n",
      "----------\n",
      "train Loss: 0.3064 Acc: 0.8935\n",
      "val Loss: 0.1003 Acc: 0.9647\n",
      "\n",
      "Epoch 18/24\n",
      "----------\n",
      "train Loss: 0.2893 Acc: 0.8983\n",
      "val Loss: 0.0997 Acc: 0.9647\n",
      "\n",
      "Epoch 19/24\n",
      "----------\n",
      "train Loss: 0.3005 Acc: 0.8918\n",
      "val Loss: 0.0976 Acc: 0.9647\n",
      "\n",
      "Epoch 20/24\n",
      "----------\n",
      "train Loss: 0.2992 Acc: 0.8979\n",
      "val Loss: 0.0965 Acc: 0.9647\n",
      "\n",
      "Epoch 21/24\n",
      "----------\n",
      "train Loss: 0.3058 Acc: 0.8979\n",
      "val Loss: 0.0964 Acc: 0.9647\n",
      "\n",
      "Epoch 22/24\n",
      "----------\n",
      "train Loss: 0.3083 Acc: 0.9001\n",
      "val Loss: 0.0968 Acc: 0.9647\n",
      "\n",
      "Epoch 23/24\n",
      "----------\n",
      "train Loss: 0.3015 Acc: 0.8975\n",
      "val Loss: 0.0971 Acc: 0.9647\n",
      "\n",
      "Epoch 24/24\n",
      "----------\n",
      "train Loss: 0.2913 Acc: 0.9049\n",
      "val Loss: 0.0976 Acc: 0.9647\n",
      "\n",
      "Training complete in 27m 4s\n",
      "Best val Acc: 0.964706\n"
     ]
    }
   ],
   "source": [
    "model_vgg16 = train_model(model_vgg16, criterion, optimizer_conv, exp_lr_scheduler, num_epochs=25)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ResNet Architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "for param in model_resnet.parameters():\n",
    "    param.requires_grad = False\n",
    "\n",
    "num_ftrs = model_resnet.fc.in_features\n",
    "model_resnet.fc = nn.Linear(num_ftrs, len(class_names))\n",
    "model_resnet = model_resnet.to(device)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer_conv = optim.SGD(model_resnet.fc.parameters(), lr=0.001, momentum=0.9)\n",
    "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_conv, step_size=7, gamma=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0/24\n",
      "----------\n",
      "train Loss: 0.7583 Acc: 0.7077\n",
      "val Loss: 0.2599 Acc: 0.9255\n",
      "\n",
      "Epoch 1/24\n",
      "----------\n",
      "train Loss: 0.5358 Acc: 0.8103\n",
      "val Loss: 0.1720 Acc: 0.9333\n",
      "\n",
      "Epoch 2/24\n",
      "----------\n",
      "train Loss: 0.5212 Acc: 0.8098\n",
      "val Loss: 0.1459 Acc: 0.9529\n",
      "\n",
      "Epoch 3/24\n",
      "----------\n",
      "train Loss: 0.4364 Acc: 0.8418\n",
      "val Loss: 0.1577 Acc: 0.9451\n",
      "\n",
      "Epoch 4/24\n",
      "----------\n",
      "train Loss: 0.4518 Acc: 0.8344\n",
      "val Loss: 0.1248 Acc: 0.9529\n",
      "\n",
      "Epoch 5/24\n",
      "----------\n",
      "train Loss: 0.4502 Acc: 0.8383\n",
      "val Loss: 0.1211 Acc: 0.9569\n",
      "\n",
      "Epoch 6/24\n",
      "----------\n",
      "train Loss: 0.4754 Acc: 0.8335\n",
      "val Loss: 0.1378 Acc: 0.9490\n",
      "\n",
      "Epoch 7/24\n",
      "----------\n",
      "train Loss: 0.3959 Acc: 0.8554\n",
      "val Loss: 0.1172 Acc: 0.9647\n",
      "\n",
      "Epoch 8/24\n",
      "----------\n",
      "train Loss: 0.3609 Acc: 0.8690\n",
      "val Loss: 0.1277 Acc: 0.9569\n",
      "\n",
      "Epoch 9/24\n",
      "----------\n",
      "train Loss: 0.3653 Acc: 0.8751\n",
      "val Loss: 0.1263 Acc: 0.9608\n",
      "\n",
      "Epoch 10/24\n",
      "----------\n",
      "train Loss: 0.3580 Acc: 0.8681\n",
      "val Loss: 0.1059 Acc: 0.9647\n",
      "\n",
      "Epoch 11/24\n",
      "----------\n",
      "train Loss: 0.3863 Acc: 0.8580\n",
      "val Loss: 0.1313 Acc: 0.9608\n",
      "\n",
      "Epoch 12/24\n",
      "----------\n",
      "train Loss: 0.3368 Acc: 0.8777\n",
      "val Loss: 0.1142 Acc: 0.9686\n",
      "\n",
      "Epoch 13/24\n",
      "----------\n",
      "train Loss: 0.3500 Acc: 0.8755\n",
      "val Loss: 0.1120 Acc: 0.9608\n",
      "\n",
      "Epoch 14/24\n",
      "----------\n",
      "train Loss: 0.3470 Acc: 0.8729\n",
      "val Loss: 0.1105 Acc: 0.9569\n",
      "\n",
      "Epoch 15/24\n",
      "----------\n",
      "train Loss: 0.3796 Acc: 0.8580\n",
      "val Loss: 0.1232 Acc: 0.9686\n",
      "\n",
      "Epoch 16/24\n",
      "----------\n",
      "train Loss: 0.3209 Acc: 0.8812\n",
      "val Loss: 0.1284 Acc: 0.9647\n",
      "\n",
      "Epoch 17/24\n",
      "----------\n",
      "train Loss: 0.3619 Acc: 0.8729\n",
      "val Loss: 0.1378 Acc: 0.9569\n",
      "\n",
      "Epoch 18/24\n",
      "----------\n",
      "train Loss: 0.3497 Acc: 0.8703\n",
      "val Loss: 0.1223 Acc: 0.9686\n",
      "\n",
      "Epoch 19/24\n",
      "----------\n",
      "train Loss: 0.3342 Acc: 0.8826\n",
      "val Loss: 0.1167 Acc: 0.9608\n",
      "\n",
      "Epoch 20/24\n",
      "----------\n",
      "train Loss: 0.3563 Acc: 0.8729\n",
      "val Loss: 0.1306 Acc: 0.9608\n",
      "\n",
      "Epoch 21/24\n",
      "----------\n",
      "train Loss: 0.3570 Acc: 0.8747\n",
      "val Loss: 0.1151 Acc: 0.9647\n",
      "\n",
      "Epoch 22/24\n",
      "----------\n",
      "train Loss: 0.3512 Acc: 0.8712\n",
      "val Loss: 0.1392 Acc: 0.9490\n",
      "\n",
      "Epoch 23/24\n",
      "----------\n",
      "train Loss: 0.3775 Acc: 0.8655\n",
      "val Loss: 0.1080 Acc: 0.9647\n",
      "\n",
      "Epoch 24/24\n",
      "----------\n",
      "train Loss: 0.3543 Acc: 0.8712\n",
      "val Loss: 0.1183 Acc: 0.9647\n",
      "\n",
      "Training complete in 10m 7s\n",
      "Best val Acc: 0.968627\n"
     ]
    }
   ],
   "source": [
    "model_resnet = train_model(model_resnet, criterion, optimizer_conv, exp_lr_scheduler, num_epochs=25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "for param in model_resnet.parameters():\n",
    "    param.requires_grad = False\n",
    "\n",
    "num_ftrs = model_resnet.fc.in_features\n",
    "model_resnet.fc = nn.Linear(num_ftrs, len(class_names))\n",
    "model_resnet = model_resnet.to(device)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer_conv = optim.SGD(model_resnet.fc.parameters(), lr=0.001, momentum=0.9)\n",
    "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_conv, step_size=7, gamma=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0/24\n",
      "----------\n",
      "train Loss: 0.7221 Acc: 0.7182\n",
      "val Loss: 0.2825 Acc: 0.8824\n",
      "\n",
      "Epoch 1/24\n",
      "----------\n",
      "train Loss: 0.4867 Acc: 0.8098\n",
      "val Loss: 0.1549 Acc: 0.9412\n",
      "\n",
      "Epoch 2/24\n",
      "----------\n",
      "train Loss: 0.4826 Acc: 0.8164\n",
      "val Loss: 0.1943 Acc: 0.9255\n",
      "\n",
      "Epoch 3/24\n",
      "----------\n",
      "train Loss: 0.5605 Acc: 0.8138\n",
      "val Loss: 0.1159 Acc: 0.9569\n",
      "\n",
      "Epoch 4/24\n",
      "----------\n",
      "train Loss: 0.4739 Acc: 0.8392\n",
      "val Loss: 0.1291 Acc: 0.9529\n",
      "\n",
      "Epoch 5/24\n",
      "----------\n",
      "train Loss: 0.5175 Acc: 0.8155\n",
      "val Loss: 0.1160 Acc: 0.9569\n",
      "\n",
      "Epoch 6/24\n",
      "----------\n",
      "train Loss: 0.5044 Acc: 0.8234\n",
      "val Loss: 0.1401 Acc: 0.9451\n",
      "\n",
      "Epoch 7/24\n",
      "----------\n",
      "train Loss: 0.3871 Acc: 0.8650\n",
      "val Loss: 0.1084 Acc: 0.9647\n",
      "\n",
      "Epoch 8/24\n",
      "----------\n",
      "train Loss: 0.3434 Acc: 0.8747\n",
      "val Loss: 0.1257 Acc: 0.9569\n",
      "\n",
      "Epoch 9/24\n",
      "----------\n",
      "train Loss: 0.3556 Acc: 0.8760\n",
      "val Loss: 0.1021 Acc: 0.9686\n",
      "\n",
      "Epoch 10/24\n",
      "----------\n",
      "train Loss: 0.3879 Acc: 0.8585\n",
      "val Loss: 0.1121 Acc: 0.9608\n",
      "\n",
      "Epoch 11/24\n",
      "----------\n",
      "train Loss: 0.3576 Acc: 0.8699\n",
      "val Loss: 0.1359 Acc: 0.9529\n",
      "\n",
      "Epoch 12/24\n",
      "----------\n",
      "train Loss: 0.3775 Acc: 0.8663\n",
      "val Loss: 0.1129 Acc: 0.9686\n",
      "\n",
      "Epoch 13/24\n",
      "----------\n",
      "train Loss: 0.3583 Acc: 0.8755\n",
      "val Loss: 0.1269 Acc: 0.9686\n",
      "\n",
      "Epoch 14/24\n",
      "----------\n",
      "train Loss: 0.3631 Acc: 0.8716\n",
      "val Loss: 0.1208 Acc: 0.9647\n",
      "\n",
      "Epoch 15/24\n",
      "----------\n",
      "train Loss: 0.3848 Acc: 0.8585\n",
      "val Loss: 0.1155 Acc: 0.9608\n",
      "\n",
      "Epoch 16/24\n",
      "----------\n",
      "train Loss: 0.3399 Acc: 0.8725\n",
      "val Loss: 0.1134 Acc: 0.9647\n",
      "\n",
      "Epoch 17/24\n",
      "----------\n",
      "train Loss: 0.3558 Acc: 0.8712\n",
      "val Loss: 0.1287 Acc: 0.9608\n",
      "\n",
      "Epoch 18/24\n",
      "----------\n",
      "train Loss: 0.3709 Acc: 0.8642\n",
      "val Loss: 0.1532 Acc: 0.9451\n",
      "\n",
      "Epoch 19/24\n",
      "----------\n",
      "train Loss: 0.3463 Acc: 0.8725\n",
      "val Loss: 0.1227 Acc: 0.9608\n",
      "\n",
      "Epoch 20/24\n",
      "----------\n",
      "train Loss: 0.3684 Acc: 0.8685\n",
      "val Loss: 0.1231 Acc: 0.9608\n",
      "\n",
      "Epoch 21/24\n",
      "----------\n",
      "train Loss: 0.3474 Acc: 0.8786\n",
      "val Loss: 0.1265 Acc: 0.9647\n",
      "\n",
      "Epoch 22/24\n",
      "----------\n",
      "train Loss: 0.3394 Acc: 0.8869\n",
      "val Loss: 0.1111 Acc: 0.9647\n",
      "\n",
      "Epoch 23/24\n",
      "----------\n",
      "train Loss: 0.3388 Acc: 0.8852\n",
      "val Loss: 0.1058 Acc: 0.9686\n",
      "\n",
      "Epoch 24/24\n",
      "----------\n",
      "train Loss: 0.3362 Acc: 0.8812\n",
      "val Loss: 0.1284 Acc: 0.9647\n",
      "\n",
      "Training complete in 12m 16s\n",
      "Best val Acc: 0.968627\n"
     ]
    }
   ],
   "source": [
    "model_resnet = train_model(model_resnet, criterion, optimizer_conv, exp_lr_scheduler, num_epochs=25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "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): 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, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\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, track_running_stats=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, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=True)\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, track_running_stats=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, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=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",
       "      (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, track_running_stats=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, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=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",
       "    )\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, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=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",
       "      (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, track_running_stats=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, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=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",
       "    )\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, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=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",
       "      (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): 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, track_running_stats=True)\n",
       "      (relu): ReLU(inplace=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",
       "    )\n",
       "  )\n",
       "  (avgpool): AdaptiveAvgPool2d(output_size=(1, 1))\n",
       "  (fc): Linear(in_features=512, out_features=4, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_resnet"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### AlexNet Architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [],
   "source": [
    "for param in model_alexnet.parameters():\n",
    "    param.requires_grad = False\n",
    "\n",
    "num_ftrs = model_alexnet.classifier[6].in_features\n",
    "model_alexnet.classifier[6] = nn.Linear(num_ftrs, len(class_names))\n",
    "model_alexnet = model_alexnet.to(device)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer_conv = optim.SGD(model_alexnet.classifier[6].parameters(), lr=0.001, momentum=0.9)\n",
    "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_conv, step_size=7, gamma=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0/24\n",
      "----------\n",
      "train Loss: 0.8808 Acc: 0.8067\n",
      "val Loss: 0.4606 Acc: 0.8941\n",
      "\n",
      "Epoch 1/24\n",
      "----------\n",
      "train Loss: 0.6680 Acc: 0.8703\n",
      "val Loss: 0.2884 Acc: 0.9333\n",
      "\n",
      "Epoch 2/24\n",
      "----------\n",
      "train Loss: 0.7559 Acc: 0.8615\n",
      "val Loss: 0.3503 Acc: 0.9216\n",
      "\n",
      "Epoch 3/24\n",
      "----------\n",
      "train Loss: 0.6549 Acc: 0.8786\n",
      "val Loss: 0.2968 Acc: 0.9294\n",
      "\n",
      "Epoch 4/24\n",
      "----------\n",
      "train Loss: 0.6638 Acc: 0.8817\n",
      "val Loss: 0.3350 Acc: 0.9373\n",
      "\n",
      "Epoch 5/24\n",
      "----------\n",
      "train Loss: 0.6190 Acc: 0.8940\n",
      "val Loss: 0.2408 Acc: 0.9608\n",
      "\n",
      "Epoch 6/24\n",
      "----------\n",
      "train Loss: 0.5249 Acc: 0.8988\n",
      "val Loss: 0.2852 Acc: 0.9333\n",
      "\n",
      "Epoch 7/24\n",
      "----------\n",
      "train Loss: 0.3991 Acc: 0.9172\n",
      "val Loss: 0.1568 Acc: 0.9490\n",
      "\n",
      "Epoch 8/24\n",
      "----------\n",
      "train Loss: 0.3714 Acc: 0.9154\n",
      "val Loss: 0.1745 Acc: 0.9608\n",
      "\n",
      "Epoch 9/24\n",
      "----------\n",
      "train Loss: 0.3343 Acc: 0.9264\n",
      "val Loss: 0.1197 Acc: 0.9686\n",
      "\n",
      "Epoch 10/24\n",
      "----------\n",
      "train Loss: 0.3455 Acc: 0.9172\n",
      "val Loss: 0.1365 Acc: 0.9686\n",
      "\n",
      "Epoch 11/24\n",
      "----------\n",
      "train Loss: 0.3270 Acc: 0.9238\n",
      "val Loss: 0.0892 Acc: 0.9765\n",
      "\n",
      "Epoch 12/24\n",
      "----------\n",
      "train Loss: 0.2971 Acc: 0.9273\n",
      "val Loss: 0.0850 Acc: 0.9686\n",
      "\n",
      "Epoch 13/24\n",
      "----------\n",
      "train Loss: 0.2779 Acc: 0.9338\n",
      "val Loss: 0.1484 Acc: 0.9647\n",
      "\n",
      "Epoch 14/24\n",
      "----------\n",
      "train Loss: 0.3290 Acc: 0.9220\n",
      "val Loss: 0.1103 Acc: 0.9686\n",
      "\n",
      "Epoch 15/24\n",
      "----------\n",
      "train Loss: 0.2828 Acc: 0.9321\n",
      "val Loss: 0.1003 Acc: 0.9686\n",
      "\n",
      "Epoch 16/24\n",
      "----------\n",
      "train Loss: 0.3172 Acc: 0.9255\n",
      "val Loss: 0.0944 Acc: 0.9686\n",
      "\n",
      "Epoch 17/24\n",
      "----------\n",
      "train Loss: 0.2726 Acc: 0.9330\n",
      "val Loss: 0.0930 Acc: 0.9765\n",
      "\n",
      "Epoch 18/24\n",
      "----------\n",
      "train Loss: 0.3019 Acc: 0.9242\n",
      "val Loss: 0.0913 Acc: 0.9765\n",
      "\n",
      "Epoch 19/24\n",
      "----------\n",
      "train Loss: 0.2980 Acc: 0.9334\n",
      "val Loss: 0.0957 Acc: 0.9686\n",
      "\n",
      "Epoch 20/24\n",
      "----------\n",
      "train Loss: 0.2822 Acc: 0.9316\n",
      "val Loss: 0.1000 Acc: 0.9725\n",
      "\n",
      "Epoch 21/24\n",
      "----------\n",
      "train Loss: 0.2663 Acc: 0.9334\n",
      "val Loss: 0.0977 Acc: 0.9686\n",
      "\n",
      "Epoch 22/24\n",
      "----------\n",
      "train Loss: 0.3221 Acc: 0.9220\n",
      "val Loss: 0.0964 Acc: 0.9686\n",
      "\n",
      "Epoch 23/24\n",
      "----------\n",
      "train Loss: 0.3045 Acc: 0.9290\n",
      "val Loss: 0.0961 Acc: 0.9686\n",
      "\n",
      "Epoch 24/24\n",
      "----------\n",
      "train Loss: 0.3189 Acc: 0.9321\n",
      "val Loss: 0.0955 Acc: 0.9686\n",
      "\n",
      "Training complete in 7m 40s\n",
      "Best val Acc: 0.976471\n"
     ]
    }
   ],
   "source": [
    "model_alexnet = train_model(model_alexnet, criterion, optimizer_conv, exp_lr_scheduler, num_epochs=25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "AlexNet(\n",
       "  (features): Sequential(\n",
       "    (0): Conv2d(3, 64, kernel_size=(11, 11), stride=(4, 4), padding=(2, 2))\n",
       "    (1): ReLU(inplace=True)\n",
       "    (2): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
       "    (3): Conv2d(64, 192, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))\n",
       "    (4): ReLU(inplace=True)\n",
       "    (5): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
       "    (6): Conv2d(192, 384, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "    (7): ReLU(inplace=True)\n",
       "    (8): Conv2d(384, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "    (9): ReLU(inplace=True)\n",
       "    (10): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
       "    (11): ReLU(inplace=True)\n",
       "    (12): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
       "  )\n",
       "  (avgpool): AdaptiveAvgPool2d(output_size=(6, 6))\n",
       "  (classifier): Sequential(\n",
       "    (0): Dropout(p=0.5, inplace=False)\n",
       "    (1): Linear(in_features=9216, out_features=4096, bias=True)\n",
       "    (2): ReLU(inplace=True)\n",
       "    (3): Dropout(p=0.5, inplace=False)\n",
       "    (4): Linear(in_features=4096, out_features=4096, bias=True)\n",
       "    (5): ReLU(inplace=True)\n",
       "    (6): Linear(in_features=4096, out_features=4, bias=True)\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_alexnet"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "GroundTruth:  Cedar Apple Rust, Healthy, Black Rot, Healthy\n",
      "Predicted:  Cedar Apple Rust, Healthy, Black Rot, Healthy\n"
     ]
    }
   ],
   "source": [
    "dataiter = iter(dataloaders['test'])\n",
    "images, labels = dataiter.next()\n",
    "imshow(torchvision.utils.make_grid(images))\n",
    "print('GroundTruth: ', ', '.join('%5s' % class_names[x] for x in labels))\n",
    "images = images.to(device)\n",
    "labels = images.to(device)\n",
    "output = model_resnet(images)\n",
    "_, predicted = torch.max(output, 1)\n",
    "print('Predicted: ', ', '.join('%5s' % class_names[x] for x in predicted))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Save the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Save The Model\n",
    "PATH = './fix_resnet18.pth'\n",
    "torch.save(model_resnet.state_dict(), PATH)"
   ]
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}