MobileNetV2 (1).ipynb

{"cells":[{"cell_type":"code","execution_count":null,"id":"83e573af-aa64-42c3-a8c6-13ed12ccb5b9","metadata":{"id":"83e573af-aa64-42c3-a8c6-13ed12ccb5b9"},"outputs":[],"source":["import os\n","import torch\n","import torch.nn as nn\n","import torch.optim as optim\n","from torch.utils.data import DataLoader\n","from torchvision import datasets, transforms, models\n","import matplotlib.pyplot as plt\n","from tqdm import tqdm\n","from torchsummary import summary\n","from sklearn.metrics import precision_score, recall_score, f1_score, accuracy_score, confusion_matrix, roc_curve, auc, average_precision_score, precision_recall_curve\n","import pandas as pd\n","import numpy as np\n","import seaborn as sns\n","from sklearn.preprocessing import label_binarize\n","import torch.optim as optim"]},{"cell_type":"code","execution_count":null,"id":"13e1c9f7-0252-431a-911b-80151de62a22","metadata":{"id":"13e1c9f7-0252-431a-911b-80151de62a22","outputId":"e9d6cf20-bb9b-4365-fb9c-458ef52a74a5"},"outputs":[{"name":"stdout","output_type":"stream","text":["Using device: cuda\n"]}],"source":["# Check if GPU is available\n","device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n","print(f\"Using device: {device}\")"]},{"cell_type":"code","execution_count":null,"id":"d90cdec6-42d1-4ce0-bc02-62ae26322580","metadata":{"id":"d90cdec6-42d1-4ce0-bc02-62ae26322580"},"outputs":[],"source":["# Set paths for dataset splits\n","train_dir = 'I:/RICE/final_rice_data/train'\n","val_dir = 'I:/RICE/final_rice_data/val'\n","test_dir = 'I:/RICE/final_rice_data/test'"]},{"cell_type":"code","execution_count":null,"id":"186481a5-8f51-489e-b9fd-7ab96ba0fcb3","metadata":{"id":"186481a5-8f51-489e-b9fd-7ab96ba0fcb3","outputId":"e7eecaa5-8cdf-4109-eb34-9245d8e879e7"},"outputs":[{"name":"stdout","output_type":"stream","text":["Total number of images in the training set: 10000\n","Total number of images in the validation set: 1250\n","Total number of images in the test set: 1250\n"]}],"source":["# Function to count images in a directory\n","def count_images_in_directory(directory):\n","    count = 0\n","    for _, _, files in os.walk(directory):\n","        count += len([f for f in files if f.lower().endswith(('.png', '.jpg', '.jpeg'))])\n","    return count\n","\n","# Count images in each directory\n","train_count = count_images_in_directory(train_dir)\n","val_count = count_images_in_directory(val_dir)\n","test_count = count_images_in_directory(test_dir)\n","\n","# Display the counts\n","print(f\"Total number of images in the training set: {train_count}\")\n","print(f\"Total number of images in the validation set: {val_count}\")\n","print(f\"Total number of images in the test set: {test_count}\")"]},{"cell_type":"code","execution_count":null,"id":"d29614dd-8eb8-41aa-9704-186815f43f6e","metadata":{"id":"d29614dd-8eb8-41aa-9704-186815f43f6e"},"outputs":[],"source":["# Data augmentation and normalization for training\n","train_transforms = transforms.Compose([\n","    transforms.Resize((224, 224)),          # Resize images\n","    transforms.RandomHorizontalFlip(),        # Random horizontal flip\n","    transforms.RandomRotation(20),            # Random rotation\n","    transforms.ToTensor(),                    # Convert image to tensor\n","    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]), # Normalize\n","])"]},{"cell_type":"code","execution_count":null,"id":"47314d31-ac3e-481b-8bd8-20eccd191e33","metadata":{"id":"47314d31-ac3e-481b-8bd8-20eccd191e33"},"outputs":[],"source":["# Normalization for validation and test\n","val_test_transforms = transforms.Compose([\n","    transforms.Resize((224, 224)),\n","    transforms.ToTensor(),\n","    transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),\n","])"]},{"cell_type":"code","execution_count":null,"id":"615fd829-e718-45bb-9a0f-b0b3745ea279","metadata":{"id":"615fd829-e718-45bb-9a0f-b0b3745ea279"},"outputs":[],"source":["# Load datasets\n","train_dataset = datasets.ImageFolder(root=train_dir, transform=train_transforms)\n","val_dataset = datasets.ImageFolder(root=val_dir, transform=val_test_transforms)\n","test_dataset = datasets.ImageFolder(root=test_dir, transform=val_test_transforms)"]},{"cell_type":"code","execution_count":null,"id":"e9ac0e7e-8ba3-4408-917c-e1e266169ee6","metadata":{"id":"e9ac0e7e-8ba3-4408-917c-e1e266169ee6"},"outputs":[],"source":["# Data loaders\n","train_loader = DataLoader(train_dataset, batch_size=32, shuffle=True)\n","val_loader = DataLoader(val_dataset, batch_size=32, shuffle=False)\n","test_loader = DataLoader(test_dataset, batch_size=32, shuffle=False)"]},{"cell_type":"code","execution_count":null,"id":"39631a9c-add4-4572-96e9-30a935c14f6e","metadata":{"id":"39631a9c-add4-4572-96e9-30a935c14f6e","outputId":"13af10b1-8727-4c60-e900-9c15a2665369"},"outputs":[{"name":"stderr","output_type":"stream","text":["C:\\Users\\Administrator\\anaconda3\\envs\\torch\\lib\\site-packages\\torchvision\\models\\_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and may be removed in the future, please use 'weights' instead.\n","  warnings.warn(\n","C:\\Users\\Administrator\\anaconda3\\envs\\torch\\lib\\site-packages\\torchvision\\models\\_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and may be removed in the future. The current behavior is equivalent to passing `weights=MobileNet_V2_Weights.IMAGENET1K_V1`. You can also use `weights=MobileNet_V2_Weights.DEFAULT` to get the most up-to-date weights.\n","  warnings.warn(msg)\n"]}],"source":["# Define the MobileNetV2 model architecture\n","class MobileNetV2Model(nn.Module):\n","    def __init__(self, num_classes):\n","        super(MobileNetV2Model, self).__init__()\n","        self.mobilenet = models.mobilenet_v2(pretrained=True)  # Load MobileNetV2 with pretrained weights\n","        self.mobilenet.classifier[1] = nn.Linear(self.mobilenet.classifier[1].in_features, num_classes)  # Modify classifier\n","\n","    def forward(self, x):\n","        return self.mobilenet(x)\n","\n","# Initialize the model\n","num_classes = 5  # Adjust based on your dataset\n","device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n","model = MobileNetV2Model(num_classes).to(device)"]},{"cell_type":"code","execution_count":null,"id":"06cfef83-d14a-4237-9606-f87ea739bf4f","metadata":{"scrolled":true,"id":"06cfef83-d14a-4237-9606-f87ea739bf4f","outputId":"a4c0f622-3c08-4704-cc58-f5232925ee65"},"outputs":[{"name":"stdout","output_type":"stream","text":["----------------------------------------------------------------\n","        Layer (type)               Output Shape         Param #\n","================================================================\n","            Conv2d-1         [-1, 32, 112, 112]             864\n","       BatchNorm2d-2         [-1, 32, 112, 112]              64\n","             ReLU6-3         [-1, 32, 112, 112]               0\n","            Conv2d-4         [-1, 32, 112, 112]             288\n","       BatchNorm2d-5         [-1, 32, 112, 112]              64\n","             ReLU6-6         [-1, 32, 112, 112]               0\n","            Conv2d-7         [-1, 16, 112, 112]             512\n","       BatchNorm2d-8         [-1, 16, 112, 112]              32\n","  InvertedResidual-9         [-1, 16, 112, 112]               0\n","           Conv2d-10         [-1, 96, 112, 112]           1,536\n","      BatchNorm2d-11         [-1, 96, 112, 112]             192\n","            ReLU6-12         [-1, 96, 112, 112]               0\n","           Conv2d-13           [-1, 96, 56, 56]             864\n","      BatchNorm2d-14           [-1, 96, 56, 56]             192\n","            ReLU6-15           [-1, 96, 56, 56]               0\n","           Conv2d-16           [-1, 24, 56, 56]           2,304\n","      BatchNorm2d-17           [-1, 24, 56, 56]              48\n"," InvertedResidual-18           [-1, 24, 56, 56]               0\n","           Conv2d-19          [-1, 144, 56, 56]           3,456\n","      BatchNorm2d-20          [-1, 144, 56, 56]             288\n","            ReLU6-21          [-1, 144, 56, 56]               0\n","           Conv2d-22          [-1, 144, 56, 56]           1,296\n","      BatchNorm2d-23          [-1, 144, 56, 56]             288\n","            ReLU6-24          [-1, 144, 56, 56]               0\n","           Conv2d-25           [-1, 24, 56, 56]           3,456\n","      BatchNorm2d-26           [-1, 24, 56, 56]              48\n"," InvertedResidual-27           [-1, 24, 56, 56]               0\n","           Conv2d-28          [-1, 144, 56, 56]           3,456\n","      BatchNorm2d-29          [-1, 144, 56, 56]             288\n","            ReLU6-30          [-1, 144, 56, 56]               0\n","           Conv2d-31          [-1, 144, 28, 28]           1,296\n","      BatchNorm2d-32          [-1, 144, 28, 28]             288\n","            ReLU6-33          [-1, 144, 28, 28]               0\n","           Conv2d-34           [-1, 32, 28, 28]           4,608\n","      BatchNorm2d-35           [-1, 32, 28, 28]              64\n"," InvertedResidual-36           [-1, 32, 28, 28]               0\n","           Conv2d-37          [-1, 192, 28, 28]           6,144\n","      BatchNorm2d-38          [-1, 192, 28, 28]             384\n","            ReLU6-39          [-1, 192, 28, 28]               0\n","           Conv2d-40          [-1, 192, 28, 28]           1,728\n","      BatchNorm2d-41          [-1, 192, 28, 28]             384\n","            ReLU6-42          [-1, 192, 28, 28]               0\n","           Conv2d-43           [-1, 32, 28, 28]           6,144\n","      BatchNorm2d-44           [-1, 32, 28, 28]              64\n"," InvertedResidual-45           [-1, 32, 28, 28]               0\n","           Conv2d-46          [-1, 192, 28, 28]           6,144\n","      BatchNorm2d-47          [-1, 192, 28, 28]             384\n","            ReLU6-48          [-1, 192, 28, 28]               0\n","           Conv2d-49          [-1, 192, 28, 28]           1,728\n","      BatchNorm2d-50          [-1, 192, 28, 28]             384\n","            ReLU6-51          [-1, 192, 28, 28]               0\n","           Conv2d-52           [-1, 32, 28, 28]           6,144\n","      BatchNorm2d-53           [-1, 32, 28, 28]              64\n"," InvertedResidual-54           [-1, 32, 28, 28]               0\n","           Conv2d-55          [-1, 192, 28, 28]           6,144\n","      BatchNorm2d-56          [-1, 192, 28, 28]             384\n","            ReLU6-57          [-1, 192, 28, 28]               0\n","           Conv2d-58          [-1, 192, 14, 14]           1,728\n","      BatchNorm2d-59          [-1, 192, 14, 14]             384\n","            ReLU6-60          [-1, 192, 14, 14]               0\n","           Conv2d-61           [-1, 64, 14, 14]          12,288\n","      BatchNorm2d-62           [-1, 64, 14, 14]             128\n"," InvertedResidual-63           [-1, 64, 14, 14]               0\n","           Conv2d-64          [-1, 384, 14, 14]          24,576\n","      BatchNorm2d-65          [-1, 384, 14, 14]             768\n","            ReLU6-66          [-1, 384, 14, 14]               0\n","           Conv2d-67          [-1, 384, 14, 14]           3,456\n","      BatchNorm2d-68          [-1, 384, 14, 14]             768\n","            ReLU6-69          [-1, 384, 14, 14]               0\n","           Conv2d-70           [-1, 64, 14, 14]          24,576\n","      BatchNorm2d-71           [-1, 64, 14, 14]             128\n"," InvertedResidual-72           [-1, 64, 14, 14]               0\n","           Conv2d-73          [-1, 384, 14, 14]          24,576\n","      BatchNorm2d-74          [-1, 384, 14, 14]             768\n","            ReLU6-75          [-1, 384, 14, 14]               0\n","           Conv2d-76          [-1, 384, 14, 14]           3,456\n","      BatchNorm2d-77          [-1, 384, 14, 14]             768\n","            ReLU6-78          [-1, 384, 14, 14]               0\n","           Conv2d-79           [-1, 64, 14, 14]          24,576\n","      BatchNorm2d-80           [-1, 64, 14, 14]             128\n"," InvertedResidual-81           [-1, 64, 14, 14]               0\n","           Conv2d-82          [-1, 384, 14, 14]          24,576\n","      BatchNorm2d-83          [-1, 384, 14, 14]             768\n","            ReLU6-84          [-1, 384, 14, 14]               0\n","           Conv2d-85          [-1, 384, 14, 14]           3,456\n","      BatchNorm2d-86          [-1, 384, 14, 14]             768\n","            ReLU6-87          [-1, 384, 14, 14]               0\n","           Conv2d-88           [-1, 64, 14, 14]          24,576\n","      BatchNorm2d-89           [-1, 64, 14, 14]             128\n"," InvertedResidual-90           [-1, 64, 14, 14]               0\n","           Conv2d-91          [-1, 384, 14, 14]          24,576\n","      BatchNorm2d-92          [-1, 384, 14, 14]             768\n","            ReLU6-93          [-1, 384, 14, 14]               0\n","           Conv2d-94          [-1, 384, 14, 14]           3,456\n","      BatchNorm2d-95          [-1, 384, 14, 14]             768\n","            ReLU6-96          [-1, 384, 14, 14]               0\n","           Conv2d-97           [-1, 96, 14, 14]          36,864\n","      BatchNorm2d-98           [-1, 96, 14, 14]             192\n"," InvertedResidual-99           [-1, 96, 14, 14]               0\n","          Conv2d-100          [-1, 576, 14, 14]          55,296\n","     BatchNorm2d-101          [-1, 576, 14, 14]           1,152\n","           ReLU6-102          [-1, 576, 14, 14]               0\n","          Conv2d-103          [-1, 576, 14, 14]           5,184\n","     BatchNorm2d-104          [-1, 576, 14, 14]           1,152\n","           ReLU6-105          [-1, 576, 14, 14]               0\n","          Conv2d-106           [-1, 96, 14, 14]          55,296\n","     BatchNorm2d-107           [-1, 96, 14, 14]             192\n","InvertedResidual-108           [-1, 96, 14, 14]               0\n","          Conv2d-109          [-1, 576, 14, 14]          55,296\n","     BatchNorm2d-110          [-1, 576, 14, 14]           1,152\n","           ReLU6-111          [-1, 576, 14, 14]               0\n","          Conv2d-112          [-1, 576, 14, 14]           5,184\n","     BatchNorm2d-113          [-1, 576, 14, 14]           1,152\n","           ReLU6-114          [-1, 576, 14, 14]               0\n","          Conv2d-115           [-1, 96, 14, 14]          55,296\n","     BatchNorm2d-116           [-1, 96, 14, 14]             192\n","InvertedResidual-117           [-1, 96, 14, 14]               0\n","          Conv2d-118          [-1, 576, 14, 14]          55,296\n","     BatchNorm2d-119          [-1, 576, 14, 14]           1,152\n","           ReLU6-120          [-1, 576, 14, 14]               0\n","          Conv2d-121            [-1, 576, 7, 7]           5,184\n","     BatchNorm2d-122            [-1, 576, 7, 7]           1,152\n","           ReLU6-123            [-1, 576, 7, 7]               0\n","          Conv2d-124            [-1, 160, 7, 7]          92,160\n","     BatchNorm2d-125            [-1, 160, 7, 7]             320\n","InvertedResidual-126            [-1, 160, 7, 7]               0\n","          Conv2d-127            [-1, 960, 7, 7]         153,600\n","     BatchNorm2d-128            [-1, 960, 7, 7]           1,920\n","           ReLU6-129            [-1, 960, 7, 7]               0\n","          Conv2d-130            [-1, 960, 7, 7]           8,640\n","     BatchNorm2d-131            [-1, 960, 7, 7]           1,920\n","           ReLU6-132            [-1, 960, 7, 7]               0\n","          Conv2d-133            [-1, 160, 7, 7]         153,600\n","     BatchNorm2d-134            [-1, 160, 7, 7]             320\n","InvertedResidual-135            [-1, 160, 7, 7]               0\n","          Conv2d-136            [-1, 960, 7, 7]         153,600\n","     BatchNorm2d-137            [-1, 960, 7, 7]           1,920\n","           ReLU6-138            [-1, 960, 7, 7]               0\n","          Conv2d-139            [-1, 960, 7, 7]           8,640\n","     BatchNorm2d-140            [-1, 960, 7, 7]           1,920\n","           ReLU6-141            [-1, 960, 7, 7]               0\n","          Conv2d-142            [-1, 160, 7, 7]         153,600\n","     BatchNorm2d-143            [-1, 160, 7, 7]             320\n","InvertedResidual-144            [-1, 160, 7, 7]               0\n","          Conv2d-145            [-1, 960, 7, 7]         153,600\n","     BatchNorm2d-146            [-1, 960, 7, 7]           1,920\n","           ReLU6-147            [-1, 960, 7, 7]               0\n","          Conv2d-148            [-1, 960, 7, 7]           8,640\n","     BatchNorm2d-149            [-1, 960, 7, 7]           1,920\n","           ReLU6-150            [-1, 960, 7, 7]               0\n","          Conv2d-151            [-1, 320, 7, 7]         307,200\n","     BatchNorm2d-152            [-1, 320, 7, 7]             640\n","InvertedResidual-153            [-1, 320, 7, 7]               0\n","          Conv2d-154           [-1, 1280, 7, 7]         409,600\n","     BatchNorm2d-155           [-1, 1280, 7, 7]           2,560\n","           ReLU6-156           [-1, 1280, 7, 7]               0\n","         Dropout-157                 [-1, 1280]               0\n","          Linear-158                    [-1, 5]           6,405\n","     MobileNetV2-159                    [-1, 5]               0\n","================================================================\n","Total params: 2,230,277\n","Trainable params: 2,230,277\n","Non-trainable params: 0\n","----------------------------------------------------------------\n","Input size (MB): 0.57\n","Forward/backward pass size (MB): 152.86\n","Params size (MB): 8.51\n","Estimated Total Size (MB): 161.94\n","----------------------------------------------------------------\n"]}],"source":["# Display model summary\n","summary(model, (3, 224, 224))"]},{"cell_type":"code","execution_count":null,"id":"fb97622b-7b38-40c2-b516-0a5ef3ea5419","metadata":{"id":"fb97622b-7b38-40c2-b516-0a5ef3ea5419"},"outputs":[],"source":["# Define loss function and optimizer\n","criterion = nn.CrossEntropyLoss()\n","optimizer = optim.Adam(model.parameters(), lr=0.00001)"]},{"cell_type":"code","execution_count":null,"id":"68bcc11c-8b54-42af-b0db-8cc02dcf1b06","metadata":{"id":"68bcc11c-8b54-42af-b0db-8cc02dcf1b06"},"outputs":[],"source":["# Training function\n","def train(model, train_loader, criterion, optimizer, device):\n","    model.train()\n","    running_loss = 0.0\n","    correct = 0\n","    total = 0\n","    for images, labels in tqdm(train_loader, desc='Training', leave=False):\n","        images, labels = images.to(device), labels.to(device)\n","\n","        optimizer.zero_grad()\n","        outputs = model(images)\n","        loss = criterion(outputs, labels)\n","        loss.backward()\n","        optimizer.step()\n","\n","        running_loss += loss.item()\n","        _, predicted = torch.max(outputs.data, 1)\n","        total += labels.size(0)\n","        correct += (predicted == labels).sum().item()\n","\n","    avg_loss = running_loss / len(train_loader)\n","    accuracy = 100 * correct / total\n","    return avg_loss, accuracy"]},{"cell_type":"code","execution_count":null,"id":"98b0e1f9-2c14-4ad5-adbd-655ab71f5cb4","metadata":{"id":"98b0e1f9-2c14-4ad5-adbd-655ab71f5cb4"},"outputs":[],"source":["# Validation function\n","def validate(model, val_loader, criterion, device):\n","    model.eval()\n","    val_loss = 0.0\n","    correct = 0\n","    total = 0\n","    with torch.no_grad():\n","        for images, labels in val_loader:\n","            images, labels = images.to(device), labels.to(device)\n","            outputs = model(images)\n","            loss = criterion(outputs, labels)\n","            val_loss += loss.item()\n","\n","            _, predicted = torch.max(outputs.data, 1)\n","            total += labels.size(0)\n","            correct += (predicted == labels).sum().item()\n","\n","    avg_val_loss = val_loss / len(val_loader)\n","    accuracy = 100 * correct / total\n","    return avg_val_loss, accuracy"]},{"cell_type":"code","execution_count":null,"id":"c0595bb3-ce00-4373-8710-dcef22c857e2","metadata":{"id":"c0595bb3-ce00-4373-8710-dcef22c857e2"},"outputs":[],"source":["# Define a function to train and validate the model with early stopping\n","def train_and_evaluate(model, train_loader, val_loader, criterion, optimizer, num_epochs, device, patience=5):\n","    # Store losses and accuracies\n","    train_losses = []\n","    val_losses = []\n","    val_accuracies = []\n","\n","    best_val_accuracy = 0\n","    epochs_without_improvement = 0\n","\n","    for epoch in range(num_epochs):\n","        # Train the model\n","        train_loss, train_accuracy = train(model, train_loader, criterion, optimizer, device)\n","        # Validate the model\n","        val_loss, val_accuracy = validate(model, val_loader, criterion, device)\n","\n","        # Append to lists for plotting later\n","        train_losses.append(train_loss)\n","        val_losses.append(val_loss)\n","        val_accuracies.append(val_accuracy)\n","\n","        # Print training and validation results\n","        print(f\"Epoch {epoch + 1}/{num_epochs} - \"\n","              f\"Training Accuracy: {train_accuracy:.4f} - \"\n","              f\"Training Loss: {train_loss:.4f} - \"\n","              f\"Validation Accuracy: {val_accuracy:.4f} - \"\n","              f\"Validation Loss: {val_loss:.4f}\")\n","\n","        # Check if we have a new best validation accuracy\n","        if val_accuracy > best_val_accuracy:\n","            best_val_accuracy = val_accuracy\n","            epochs_without_improvement = 0\n","            # Save the best model state\n","            torch.save(model.state_dict(), 'best_crack_detection_model.pth')\n","            print(f\"Best model saved with accuracy: {best_val_accuracy:.4f}%\")\n","        else:\n","            epochs_without_improvement += 1\n","\n","        # Check for early stopping\n","        if epochs_without_improvement >= patience:\n","            print(f\"Early stopping triggered. Stopping training at epoch {epoch + 1}.\")\n","            break\n","\n","    return train_losses, val_losses, val_accuracies"]},{"cell_type":"code","execution_count":null,"id":"2327728e-02c3-4494-8858-49276c7891b3","metadata":{"id":"2327728e-02c3-4494-8858-49276c7891b3","outputId":"da9f54b0-bdcb-49b4-df34-abb01612f0e5"},"outputs":[{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 1/20 - Training Accuracy: 72.4300 - Training Loss: 0.8120 - Validation Accuracy: 88.4800 - Validation Loss: 0.3887\n","Best model saved with accuracy: 88.4800%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 2/20 - Training Accuracy: 89.8000 - Training Loss: 0.3139 - Validation Accuracy: 93.4400 - Validation Loss: 0.2129\n","Best model saved with accuracy: 93.4400%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 3/20 - Training Accuracy: 93.2400 - Training Loss: 0.2052 - Validation Accuracy: 94.3200 - Validation Loss: 0.1543\n","Best model saved with accuracy: 94.3200%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 4/20 - Training Accuracy: 94.8500 - Training Loss: 0.1511 - Validation Accuracy: 95.6000 - Validation Loss: 0.1247\n","Best model saved with accuracy: 95.6000%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 5/20 - Training Accuracy: 95.0000 - Training Loss: 0.1378 - Validation Accuracy: 96.3200 - Validation Loss: 0.1132\n","Best model saved with accuracy: 96.3200%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 6/20 - Training Accuracy: 95.8100 - Training Loss: 0.1203 - Validation Accuracy: 97.0400 - Validation Loss: 0.0847\n","Best model saved with accuracy: 97.0400%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 7/20 - Training Accuracy: 95.8800 - Training Loss: 0.1145 - Validation Accuracy: 97.2000 - Validation Loss: 0.0742\n","Best model saved with accuracy: 97.2000%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 8/20 - Training Accuracy: 96.7800 - Training Loss: 0.0931 - Validation Accuracy: 97.4400 - Validation Loss: 0.0665\n","Best model saved with accuracy: 97.4400%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 9/20 - Training Accuracy: 96.7500 - Training Loss: 0.0876 - Validation Accuracy: 97.6000 - Validation Loss: 0.0749\n","Best model saved with accuracy: 97.6000%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 10/20 - Training Accuracy: 97.1500 - Training Loss: 0.0816 - Validation Accuracy: 97.6800 - Validation Loss: 0.0604\n","Best model saved with accuracy: 97.6800%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 11/20 - Training Accuracy: 97.2300 - Training Loss: 0.0828 - Validation Accuracy: 97.6800 - Validation Loss: 0.0588\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 12/20 - Training Accuracy: 97.4500 - Training Loss: 0.0728 - Validation Accuracy: 97.8400 - Validation Loss: 0.0506\n","Best model saved with accuracy: 97.8400%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 13/20 - Training Accuracy: 97.5700 - Training Loss: 0.0712 - Validation Accuracy: 98.2400 - Validation Loss: 0.0487\n","Best model saved with accuracy: 98.2400%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 14/20 - Training Accuracy: 97.8300 - Training Loss: 0.0637 - Validation Accuracy: 98.1600 - Validation Loss: 0.0465\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 15/20 - Training Accuracy: 97.9400 - Training Loss: 0.0610 - Validation Accuracy: 98.4800 - Validation Loss: 0.0506\n","Best model saved with accuracy: 98.4800%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 16/20 - Training Accuracy: 97.9000 - Training Loss: 0.0594 - Validation Accuracy: 98.4800 - Validation Loss: 0.0475\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 17/20 - Training Accuracy: 97.9500 - Training Loss: 0.0569 - Validation Accuracy: 98.5600 - Validation Loss: 0.0467\n","Best model saved with accuracy: 98.5600%\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 18/20 - Training Accuracy: 98.1600 - Training Loss: 0.0555 - Validation Accuracy: 98.0800 - Validation Loss: 0.0423\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 19/20 - Training Accuracy: 98.2300 - Training Loss: 0.0503 - Validation Accuracy: 98.3200 - Validation Loss: 0.0404\n"]},{"name":"stderr","output_type":"stream","text":["                                                                                                                       \r"]},{"name":"stdout","output_type":"stream","text":["Epoch 20/20 - Training Accuracy: 98.2800 - Training Loss: 0.0490 - Validation Accuracy: 98.7200 - Validation Loss: 0.0377\n","Best model saved with accuracy: 98.7200%\n"]}],"source":["# Train and evaluate the model\n","num_epochs = 20\n","train_losses, val_losses, val_accuracies = train_and_evaluate(model, train_loader, val_loader, criterion, optimizer, num_epochs, device)"]},{"cell_type":"code","execution_count":null,"id":"5494928b-cfbe-479d-83e4-da21449d6130","metadata":{"id":"5494928b-cfbe-479d-83e4-da21449d6130","outputId":"79f9b47a-6e08-436e-a577-852e0340b700"},"outputs":[{"data":{"image/png":"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","text/plain":["<Figure size 1200x500 with 2 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["# Plotting accuracy and loss curves\n","def plot_curves(train_losses, val_losses, val_accuracies):\n","    epochs = range(1, len(train_losses) + 1)\n","\n","    plt.figure(figsize=(12, 5))\n","\n","    # Plot Loss Curves\n","    plt.subplot(1, 2, 1)\n","    plt.plot(epochs, train_losses, 'bo-', label='Training Loss')\n","    plt.plot(epochs, val_losses, 'ro-', label='Validation Loss')\n","    plt.title('Loss Curves')\n","    plt.xlabel('Epochs')\n","    plt.ylabel('Loss')\n","    plt.legend()\n","\n","    # Plot Accuracy Curves\n","    plt.subplot(1, 2, 2)\n","    plt.plot(epochs, val_accuracies, 'go-', label='Validation Accuracy')\n","    # plt.plot(epochs, train_accuracies, 'go-', label='Train Accuracy')\n","    plt.title('Accuracy Curve')\n","    plt.xlabel('Epochs')\n","    plt.ylabel('Accuracy (%)')\n","    plt.legend()\n","\n","    plt.tight_layout()\n","    plt.show()\n","\n","# Call the function to plot the curves\n","plot_curves(train_losses, val_losses, val_accuracies)"]}],"metadata":{"kernelspec":{"display_name":"Python 3 (ipykernel)","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":"3.8.20"},"colab":{"provenance":[]}},"nbformat":4,"nbformat_minor":5}