ImageClassifierPipeline.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torchvision \n",
    "import torch \n",
    "import torchvision.models as models\n",
    "import torchvision.transforms as transforms\n",
    "import torchvision.datasets as datasets\n",
    "import torch.nn as nn\n",
    "from time import time\n",
    "import PIL\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def trainModel(lr, epoch_count, train_ratio, val_ratio, train_batch_size, validation_batch_size, test_batch_size, num_classes):\n",
    "    \n",
    "    train_transforms = transforms.Compose([\n",
    "    transforms.Grayscale(num_output_channels=3), \n",
    "    transforms.Resize((70,70)),\n",
    "    transforms.ToTensor(),\n",
    "    transforms.Normalize(mean=[0.485, 0.456, 0.406],\n",
    "                                 std=[0.229, 0.224, 0.225])\n",
    "    ])\n",
    "    \n",
    "    \n",
    "    #loading and spliting data\n",
    "    dataset = datasets.ImageFolder('./dataset', transform = train_transforms)\n",
    "    \n",
    "    print(\"Class and Label Ids:\",dataset.class_to_idx)\n",
    "    dataset_size = len(dataset)\n",
    "    train_set_size = int(dataset_size * train_ratio)\n",
    "    validation_set_size = int(dataset_size * val_ratio)\n",
    "    test_set_size = dataset_size - train_set_size - validation_set_size\n",
    "    \n",
    "    print(\"Training set size:\", train_set_size)\n",
    "    print(\"Validation set size:\", validation_set_size)\n",
    "    print(\"Test set size:\", test_set_size)\n",
    "    \n",
    "    \n",
    "    train_set, validation_set, test_set = torch.utils.data.random_split(dataset, [train_set_size, validation_set_size, test_set_size])\n",
    "    \n",
    "    train_loader = torch.utils.data.DataLoader(train_set, batch_size = train_batch_size, shuffle=True, drop_last=True)\n",
    "\n",
    "    validation_loader = torch.utils.data.DataLoader(validation_set, batch_size = validation_batch_size, shuffle=True, drop_last=True)\n",
    "    \n",
    "    test_loader = torch.utils.data.DataLoader(test_set, batch_size = test_batch_size, shuffle=True, drop_last=True)\n",
    "    \n",
    "    train_loader = [batch for batch in train_loader]\n",
    "    train_batch_count = len(train_loader)\n",
    "    print(\"Traning Set Batch Count:\",train_batch_count)\n",
    "    print(\"Training Set Batch Size:\",train_batch_size)\n",
    "    \n",
    "    validation_loader = [batch for batch in validation_loader]\n",
    "    validation_batch_count = len(validation_loader)\n",
    "    print(\"Validation Set Batch Count:\",validation_batch_count)\n",
    "    print(\"Validation Set Batch Size:\",validation_batch_size)\n",
    "    \n",
    "    test_loader = [batch for batch in test_loader]\n",
    "    test_batch_count = len(test_loader)\n",
    "    print(\"Test Set Batch Count\",test_batch_count)\n",
    "    print(\"Test Set Batch Size:\",test_batch_size)\n",
    "    \n",
    "    print(\"*************************************************************************************\")\n",
    "    \n",
    "    \n",
    "    \n",
    "    training_loss_list = []\n",
    "    validation_loss_list = []\n",
    "    training_accuracy_list = []\n",
    "    validation_accuracy_list = []\n",
    "    test_accuracy_list = []\n",
    "    train_error_list = []\n",
    "    validation_error_list = []\n",
    "    test_error_list = []\n",
    "    \n",
    "    \n",
    "    \n",
    "    def TrueCount(pred, y_true, epoch, batch_id, set_type):\n",
    "        true_count = 0\n",
    "        for i in range(pred.size(0)):\n",
    "            if torch.argmax(pred[i])==torch.argmax(y_true[i]):\n",
    "                true_count += 1\n",
    "            else:#yanlış tahmin yaptıysa ve son epochta ise:\n",
    "                if epoch==epoch_count-1:\n",
    "                    vector = [batch_id, i, torch.argmax(pred[i]), torch.argmax(y_true[i])]\n",
    "                    if set_type == \"train\":\n",
    "                        train_error_list.append(vector)\n",
    "                    if set_type == \"validation\":\n",
    "                        validation_error_list.append(vector)\n",
    "                    if set_type == \"test\":\n",
    "                        test_error_list.append(vector)\n",
    "        return true_count    \n",
    "    \n",
    "    \n",
    "    \n",
    "\n",
    "    class Net(nn.Module):\n",
    "        def __init__(self,num_classes):\n",
    "            super().__init__()\n",
    "            \n",
    "            self.channel = 8\n",
    "            self.conv1 = torch.nn.Conv2d(in_channels = 3, out_channels = self.channel, kernel_size = 7) \n",
    "            torch.nn.init.xavier_uniform_(self.conv1.weight)\n",
    "            self.conv2 = torch.nn.Conv2d(in_channels = self.channel, out_channels = self.channel+2, kernel_size = 5)\n",
    "            torch.nn.init.xavier_uniform_(self.conv2.weight)\n",
    "            self.conv3 = torch.nn.Conv2d(in_channels = self.channel+2, out_channels = self.channel+7, kernel_size = 3)\n",
    "            torch.nn.init.xavier_uniform_(self.conv3.weight)\n",
    "            self.conv4 = torch.nn.Conv2d(in_channels = self.channel+7, out_channels = self.channel+9, kernel_size = 1)\n",
    "            torch.nn.init.xavier_uniform_(self.conv4.weight)\n",
    "\n",
    "            self.bn1 = nn.BatchNorm1d(num_features=150)\n",
    "            \n",
    "            self.dropout1 = nn.Dropout2d(0.25)\n",
    "            self.dropout2 = nn.Dropout2d(0.5)\n",
    "            self.pool = nn.MaxPool2d(2, 2)\n",
    "\n",
    "            self.beta1 = nn.Linear((self.channel+9)*3*3, 150)\n",
    "            torch.nn.init.xavier_uniform_(self.beta1.weight)\n",
    "            self.beta2 = nn.Linear(150, 150)\n",
    "            torch.nn.init.xavier_uniform_(self.beta2.weight)\n",
    "            self.beta3 = nn.Linear(150, 150)\n",
    "            torch.nn.init.xavier_uniform_(self.beta3.weight)\n",
    "            self.beta4 = nn.Linear(150, num_classes)\n",
    "            torch.nn.init.xavier_uniform_(self.beta4.weight)\n",
    "            \n",
    "            \n",
    "\n",
    "        def forward(self, X):\n",
    "\n",
    "            y=self.conv1(X) \n",
    "            y=torch.nn.functional.relu(y)\n",
    "            y=self.pool(y) \n",
    "            y=self.conv2(y) \n",
    "            y=torch.nn.functional.relu(y)\n",
    "            y=self.pool(y) \n",
    "            y=self.conv3(y) \n",
    "            y=torch.nn.functional.relu(y)\n",
    "            y=self.pool(y) \n",
    "            y=self.conv4(y) \n",
    "            y=torch.nn.functional.relu(y)\n",
    "            y=self.pool(y) \n",
    "            #y=self.dropout1(y)\n",
    "\n",
    "            #y = y.view(X.shape[0], self.channel*25*25)\n",
    "            #print(y.shape)\n",
    "            y = y.view(X.shape[0], (self.channel+9)*3*3)\n",
    "            \n",
    "            y=self.beta1(y)\n",
    "            y=self.dropout1(y)\n",
    "            y=torch.nn.functional.relu(y)\n",
    "            y=self.beta2(y)\n",
    "            y=self.dropout2(y)\n",
    "            y=torch.nn.functional.relu(y)\n",
    "            y=self.beta2(y)\n",
    "            y=torch.nn.functional.relu(y)\n",
    "            y=self.beta3(y)\n",
    "            y=self.bn1(y) \n",
    "            y=torch.nn.functional.relu(y)\n",
    "            y=self.beta4(y)\n",
    "            \n",
    "            y = torch.nn.functional.softmax(y, dim=1)\n",
    "            \n",
    "            return y\n",
    "        \n",
    "    \n",
    "    \n",
    "    model = Net(num_classes)\n",
    "    criterion = torch.nn.MSELoss()\n",
    "    #criterion = torch.nn.CrossEntropyLoss()\n",
    "    optimizer = torch.optim.Adam(model.parameters(), lr=lr, betas=(0.9, 0.999), eps=1e-08, weight_decay=0, amsgrad=False)\n",
    "    \n",
    "    t = time()\n",
    "    \n",
    "    for epoch in range(epoch_count):\n",
    "        \n",
    "        training_loss = 0\n",
    "        toplam = 0\n",
    "        dogru = 0\n",
    "        \n",
    "        for batch_id in range(len(train_loader)):\n",
    "            X,y = train_loader[batch_id]\n",
    "            #multi-label\n",
    "            y = y.view(y.size(0),1)\n",
    "            y = (y == torch.arange(num_classes).reshape(1, num_classes)).float()\n",
    "            model.train()\n",
    "            optimizer.zero_grad()\n",
    "            cikti = model(X)\n",
    "            loss = criterion(cikti, y)\n",
    "            training_loss += loss.item()\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "            \n",
    "            #Evaluation\n",
    "            with torch.no_grad():\n",
    "                model.eval()\n",
    "                cikti = model(X)\n",
    "                dogru = dogru + TrueCount(cikti,y,epoch,batch_id,\"train\")\n",
    "                toplam = toplam + y.size(0)\n",
    "                \n",
    "        training_accuracy = (dogru/toplam)*100\n",
    "        training_loss = training_loss/train_batch_count\n",
    "        training_loss_list.append(training_loss)  \n",
    "        training_accuracy_list.append(training_accuracy)\n",
    "        print(\"Epoch {} | Training Loss: {} | Training Accuracy: {}%\".format(epoch,training_loss,str(training_accuracy)))  \n",
    "        print(\"Geçen süre: {} mins\".format(round((time() - t) / 60, 2)))\n",
    "        #****************************************************************************     \n",
    "        \n",
    "        \n",
    "        validation_loss = 0\n",
    "        toplam = 0\n",
    "        dogru = 0                    \n",
    "        for batch_id in range(len(validation_loader)): \n",
    "            X,y = validation_loader[batch_id]\n",
    "            with torch.no_grad():\n",
    "                #multi-label\n",
    "                y = y.view(y.size(0),1)\n",
    "                y = (y == torch.arange(num_classes).reshape(1, num_classes)).float()\n",
    "                model.eval()\n",
    "                cikti = model(X)\n",
    "                loss = criterion(cikti, y)\n",
    "                validation_loss += loss.item() \n",
    "                dogru = dogru + TrueCount(cikti,y,epoch,batch_id,\"validation\")\n",
    "                toplam = toplam + y.size(0)\n",
    "                \n",
    "        validation_accuracy = (dogru/toplam)*100\n",
    "        validation_loss = validation_loss/validation_batch_count\n",
    "        validation_loss_list.append(validation_loss)  \n",
    "        validation_accuracy_list.append(validation_accuracy)\n",
    "        print(\"Epoch {} | Validation Loss: {} | Validation Accuracy: {}%\".format(epoch,validation_loss,str(validation_accuracy))) \n",
    "        print(\"Geçen süre: {} mins\".format(round((time() - t) / 60, 2)))\n",
    "        #****************************************************************************      \n",
    "        \n",
    "        \n",
    "        toplam = 0\n",
    "        dogru = 0                    \n",
    "        for batch_id in range(len(test_loader)): \n",
    "            X,y = test_loader[batch_id]\n",
    "            with torch.no_grad():\n",
    "                #multi-label\n",
    "                y = y.view(y.size(0),1)\n",
    "                y = (y == torch.arange(num_classes).reshape(1, num_classes)).float()\n",
    "                model.eval()\n",
    "                cikti = model(X)\n",
    "                dogru = dogru + TrueCount(cikti,y,epoch,batch_id,\"test\")\n",
    "                toplam = toplam + y.size(0)\n",
    "                \n",
    "        test_accuracy = (dogru/toplam)*100 \n",
    "        test_accuracy_list.append(test_accuracy)\n",
    "        print(\"Epoch {} | Test Accuracy: {}%\".format(epoch,str(test_accuracy))) \n",
    "        print(\"Geçen süre: {} mins\".format(round((time() - t) / 60, 2)))\n",
    "        print(\"*************************************************************************************\")\n",
    "        \n",
    "        \n",
    "  \n",
    "    #ModelSaving\n",
    "    checkpoint = {'state_dict': model.state_dict(),'optimizer' :optimizer.state_dict()}\n",
    "    torch.save(checkpoint, 'Checkpoint.pth')\n",
    "    print(\"Model kaydedildi.\")\n",
    "\n",
    "    \n",
    "    model = {}\n",
    "    \n",
    "    model['training_loss_list'] = training_loss_list\n",
    "    model['validation_loss_list'] = validation_loss_list\n",
    "    model['training_accuracy_list'] = training_accuracy_list\n",
    "    model['validation_accuracy_list'] = validation_accuracy_list\n",
    "    model['test_accuracy_list'] = test_accuracy_list\n",
    "    model['train_error_list'] = train_error_list\n",
    "    model['validation_error_list'] = validation_error_list\n",
    "    model['test_error_list'] = test_error_list\n",
    "    model['train_loader'] = train_loader\n",
    "    model['validation_loader'] = validation_loader\n",
    "    model['test_loader'] = test_loader\n",
    "    \n",
    "    return model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Class and Label Ids: {'Parasitized': 0, 'Uninfected': 1}\n",
      "Training set size: 20668\n",
      "Validation set size: 3444\n",
      "Test set size: 3446\n",
      "Traning Set Batch Count: 41\n",
      "Training Set Batch Size: 500\n",
      "Validation Set Batch Count: 34\n",
      "Validation Set Batch Size: 100\n",
      "Test Set Batch Count 34\n",
      "Test Set Batch Size: 100\n",
      "*************************************************************************************\n",
      "Epoch 0 | Training Loss: 0.20752619279593956 | Training Accuracy: 70.56585365853658%\n",
      "Geçen süre: 1.97 mins\n",
      "Epoch 0 | Validation Loss: 0.14389599991195343 | Validation Accuracy: 79.61764705882352%\n",
      "Geçen süre: 2.09 mins\n",
      "Epoch 0 | Test Accuracy: 81.02941176470588%\n",
      "Geçen süre: 2.22 mins\n",
      "*************************************************************************************\n",
      "Epoch 1 | Training Loss: 0.10359408515619069 | Training Accuracy: 83.01951219512195%\n",
      "Geçen süre: 4.38 mins\n",
      "Epoch 1 | Validation Loss: 0.08835795642260243 | Validation Accuracy: 88.73529411764706%\n",
      "Geçen süre: 4.52 mins\n",
      "Epoch 1 | Test Accuracy: 88.73529411764706%\n",
      "Geçen süre: 4.65 mins\n",
      "*************************************************************************************\n",
      "Epoch 2 | Training Loss: 0.07245262957564215 | Training Accuracy: 88.65853658536585%\n",
      "Geçen süre: 6.74 mins\n",
      "Epoch 2 | Validation Loss: 0.07529054467073258 | Validation Accuracy: 90.3235294117647%\n",
      "Geçen süre: 6.88 mins\n",
      "Epoch 2 | Test Accuracy: 90.11764705882352%\n",
      "Geçen süre: 7.01 mins\n",
      "*************************************************************************************\n",
      "Epoch 3 | Training Loss: 0.0594073070622072 | Training Accuracy: 90.92682926829269%\n",
      "Geçen süre: 9.14 mins\n",
      "Epoch 3 | Validation Loss: 0.06947750685846105 | Validation Accuracy: 91.41176470588235%\n",
      "Geçen süre: 9.27 mins\n",
      "Epoch 3 | Test Accuracy: 91.52941176470588%\n",
      "Geçen süre: 9.41 mins\n",
      "*************************************************************************************\n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-6-834ec0384021>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mmodel\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtrainModel\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mlr\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0.001\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mepoch_count\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m15\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrain_ratio\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0.75\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mval_ratio\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0.125\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtrain_batch_size\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m500\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mvalidation_batch_size\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m100\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mtest_batch_size\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m100\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnum_classes\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m<ipython-input-5-3c7460a7722f>\u001b[0m in \u001b[0;36mtrainModel\u001b[1;34m(lr, epoch_count, train_ratio, val_ratio, train_batch_size, validation_batch_size, test_batch_size, num_classes)\u001b[0m\n\u001b[0;32m    174\u001b[0m             \u001b[0mloss\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcriterion\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mcikti\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0my\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    175\u001b[0m             \u001b[0mtraining_loss\u001b[0m \u001b[1;33m+=\u001b[0m \u001b[0mloss\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mitem\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 176\u001b[1;33m             \u001b[0mloss\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    177\u001b[0m             \u001b[0moptimizer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    178\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mE:\\PROGRAM FILES\\Anaconda\\envs\\staj_projesi\\lib\\site-packages\\torch\\tensor.py\u001b[0m in \u001b[0;36mbackward\u001b[1;34m(self, gradient, retain_graph, create_graph)\u001b[0m\n\u001b[0;32m    100\u001b[0m                 \u001b[0mproducts\u001b[0m\u001b[1;33m.\u001b[0m \u001b[0mDefaults\u001b[0m \u001b[0mto\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[0;31m`\u001b[0m\u001b[0;31m`\u001b[0m\u001b[1;33m.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    101\u001b[0m         \"\"\"\n\u001b[1;32m--> 102\u001b[1;33m         \u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mautograd\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mgradient\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mretain_graph\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcreate_graph\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    103\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    104\u001b[0m     \u001b[1;32mdef\u001b[0m \u001b[0mregister_hook\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mhook\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32mE:\\PROGRAM FILES\\Anaconda\\envs\\staj_projesi\\lib\\site-packages\\torch\\autograd\\__init__.py\u001b[0m in \u001b[0;36mbackward\u001b[1;34m(tensors, grad_tensors, retain_graph, create_graph, grad_variables)\u001b[0m\n\u001b[0;32m     88\u001b[0m     Variable._execution_engine.run_backward(\n\u001b[0;32m     89\u001b[0m         \u001b[0mtensors\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mgrad_tensors\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mretain_graph\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcreate_graph\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 90\u001b[1;33m         allow_unreachable=True)  # allow_unreachable flag\n\u001b[0m\u001b[0;32m     91\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     92\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "model = trainModel(lr=0.001, epoch_count=15, train_ratio=0.75, val_ratio=0.125, train_batch_size=500, validation_batch_size=100, test_batch_size=100, num_classes=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"model[\"test_error_list\"]\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize the Error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x14a51b01808>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "batch_id = 33\n",
    "data_id = 54\n",
    "x = model['test_loader'][batch_id][0][data_id]\n",
    "z = x * torch.tensor([0.229, 0.224, 0.225]).view(3, 1, 1)\n",
    "z = z + torch.tensor([0.485, 0.456, 0.406]).view(3, 1, 1)\n",
    "img = transforms.ToPILImage(mode='RGB')(z)\n",
    "plt.imshow(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "fig_size = plt.rcParams[\"figure.figsize\"]\n",
    "fig_size[0] = 30\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "l1, = plt.plot(model['training_accuracy_list'] , label= \"Training Set Accuracy\")\n",
    "l2, = plt.plot(model['validation_accuracy_list'], label= \"Validation Set Accuracy\" )\n",
    "l3, = plt.plot(model['test_accuracy_list'], label= \"Test Set Accuracy\" )\n",
    "\n",
    "plt.legend(handles=[l1,l2,l3])\n",
    "plt.xlabel(\"Epoch\", fontsize=20)\n",
    "plt.ylabel(\"Accuracy\", fontsize=20)\n",
    "plt.autoscale(axis='x',tight=False)\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig_size = plt.rcParams[\"figure.figsize\"]\n",
    "fig_size[0] = 30\n",
    "fig_size[1] = 10\n",
    "plt.rcParams[\"figure.figsize\"] = fig_size\n",
    "\n",
    "l1, = plt.plot(model['training_loss_list'] , label= \"Training Set Loss\")\n",
    "l2, = plt.plot(model['validation_loss_list'], label= \"Validation Set Loss\" )\n",
    "\n",
    "plt.legend(handles=[l1,l2])\n",
    "plt.xlabel(\"Epoch\", fontsize=20)\n",
    "plt.ylabel(\"Accuracy\", fontsize=20)\n",
    "plt.autoscale(axis='x',tight=False)\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Loading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_checkpoint(filepath):\n",
    "    checkpoint = torch.load(filepath)\n",
    "    model = checkpoint['model']\n",
    "    model.load_state_dict(checkpoint['state_dict'])\n",
    "    for parameter in model.parameters():\n",
    "        parameter.requires_grad = False\n",
    "    \n",
    "    return model.eval()\n",
    "\n",
    "\n",
    "filepath = \"Model.pth\"\n",
    "loaded_model = load_checkpoint(filepath)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Görsel Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from PIL import Image\n",
    "from IPython.display import display\n",
    "\n",
    "red = Image.open('red1.jpg')\n",
    "display(red)\n",
    "red = train_transforms(red)\n",
    "red = red.unsqueeze(0) #dimension expansion\n",
    "\n",
    "with torch.no_grad():\n",
    "    model.eval()\n",
    "    out = model(red)\n",
    "print(out)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One-Hot Encoding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# some labels\n",
    "labels = torch.tensor([0,2,4])\n",
    "print(labels.shape)\n",
    "labels = labels.reshape(3, 1)\n",
    "print(labels.shape)\n",
    "\n",
    "num_classes = 10\n",
    "one_hot_target = (labels == torch.arange(num_classes).reshape(1, num_classes)).float()\n",
    "print(one_hot_target,one_hot_target.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}