VGG16_CIFAR10_Transfer_learning_PyTorch.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "pytorch_implementaion_cifar_transfer_learning_VGG16.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "metadata": {
        "id": "-1S9ZWljI6zB",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "!pip3 install http://download.pytorch.org/whl/cu80/torch-0.3.0.post4-cp36-cp36m-linux_x86_64.whl \n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "LlbqFJLNJNCe",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "!pip3 install torchvision"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "TxjHPluzJe09",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torchvision\n",
        "from __future__ import print_function\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "from torch.autograd import Variable\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms\n",
        "from torchvision import datasets, models\n",
        "import matplotlib.pyplot as plt \n",
        "import numpy as np\n",
        "import os, time\n",
        "import copy"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "lCIJ0pItJlSw",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "transform =torchvision.transforms.Compose([torchvision.transforms.Resize((224,224)),torchvision.transforms.ToTensor()])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "PoKtHklYJyRa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "129f5def-4dec-4042-caca-d2c717ba5156"
      },
      "cell_type": "code",
      "source": [
        "train_set = torchvision.datasets.CIFAR10(root = './data', train =True, download = True, transform =transform)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Files already downloaded and verified\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "iV35kRFFJ0zz",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "train_loader = torch.utils.data.DataLoader(train_set,batch_size=32, shuffle = True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "ISluvi6BJ6jd",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "be225346-3df5-4451-8d6a-90d6dab998a8"
      },
      "cell_type": "code",
      "source": [
        "test_set = torchvision.datasets.CIFAR10(root = './data', train = False, download = True, transform=transform)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Files already downloaded and verified\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "Xugbs3i_J7lb",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "test_loader = torch.utils.data.DataLoader(test_set, batch_size = 32, shuffle=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "xR3wdXNwJ-iP",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "classes = ['plane','car','bird','cat','deer','dog','frog','horse','ship','truck']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Bm8c4OUYKB9_",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 945
        },
        "outputId": "2b0f033e-b940-4cd4-e38a-9d099fc15b8c"
      },
      "cell_type": "code",
      "source": [
        "img, label = next(iter(train_loader))\n",
        "plt.imshow(img[0].permute(1,2,0))\n",
        "print(label)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            " 6\n",
            " 4\n",
            " 0\n",
            " 1\n",
            " 5\n",
            " 9\n",
            " 7\n",
            " 5\n",
            " 3\n",
            " 5\n",
            " 4\n",
            " 8\n",
            " 7\n",
            " 6\n",
            " 0\n",
            " 8\n",
            " 4\n",
            " 7\n",
            " 3\n",
            " 3\n",
            " 3\n",
            " 3\n",
            " 3\n",
            " 7\n",
            " 0\n",
            " 3\n",
            " 7\n",
            " 7\n",
            " 4\n",
            " 1\n",
            " 9\n",
            " 4\n",
            "[torch.LongTensor of size 32]\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAVEAAAFNCAYAAAC5YlyiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUQ8PD48i4EnUw8PDowj8D2GhwJRFK2BKAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7fc127f30550>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "NwHdL7DcKJco",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 986
        },
        "outputId": "b6a0dc8f-3a90-4ac8-ef7b-3fe9a87b410e"
      },
      "cell_type": "code",
      "source": [
        "vgg16 = models.vgg16_bn()\n",
        "print(vgg16)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "VGG(\n",
            "  (features): Sequential(\n",
            "    (0): Conv2d (3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (2): ReLU(inplace)\n",
            "    (3): Conv2d (64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (5): ReLU(inplace)\n",
            "    (6): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "    (7): Conv2d (64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (8): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (9): ReLU(inplace)\n",
            "    (10): Conv2d (128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (11): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (12): ReLU(inplace)\n",
            "    (13): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "    (14): Conv2d (128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (15): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (16): ReLU(inplace)\n",
            "    (17): Conv2d (256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (18): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (19): ReLU(inplace)\n",
            "    (20): Conv2d (256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (21): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (22): ReLU(inplace)\n",
            "    (23): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "    (24): Conv2d (256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (25): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (26): ReLU(inplace)\n",
            "    (27): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (28): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (29): ReLU(inplace)\n",
            "    (30): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (31): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (32): ReLU(inplace)\n",
            "    (33): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "    (34): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (35): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (36): ReLU(inplace)\n",
            "    (37): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (38): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (39): ReLU(inplace)\n",
            "    (40): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (41): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (42): ReLU(inplace)\n",
            "    (43): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "  )\n",
            "  (classifier): Sequential(\n",
            "    (0): Linear(in_features=25088, out_features=4096)\n",
            "    (1): ReLU(inplace)\n",
            "    (2): Dropout(p=0.5)\n",
            "    (3): Linear(in_features=4096, out_features=4096)\n",
            "    (4): ReLU(inplace)\n",
            "    (5): Dropout(p=0.5)\n",
            "    (6): Linear(in_features=4096, out_features=1000)\n",
            "  )\n",
            ")\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "Y72n_WuqKXqv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "10d877a7-b0de-4f33-a1c6-e472e924bb28"
      },
      "cell_type": "code",
      "source": [
        "for param in vgg16.features.parameters():\n",
        "    param.require_grad = False\n",
        "num_features = vgg16.classifier[6].in_features\n",
        "print(num_features)\n",
        "features = list(vgg16.classifier.children())[:-1]\n",
        "num_classes=10"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "4096\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "G-1KWtZ7KeMr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 986
        },
        "outputId": "446b3f54-4819-4230-b55f-028301ec3b0c"
      },
      "cell_type": "code",
      "source": [
        "features.extend(([nn.Linear(num_features, num_classes)]))\n",
        "vgg16.classifier = nn.Sequential(*features)\n",
        "print(vgg16)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "VGG(\n",
            "  (features): Sequential(\n",
            "    (0): Conv2d (3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (2): ReLU(inplace)\n",
            "    (3): Conv2d (64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (4): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (5): ReLU(inplace)\n",
            "    (6): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "    (7): Conv2d (64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (8): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (9): ReLU(inplace)\n",
            "    (10): Conv2d (128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (11): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (12): ReLU(inplace)\n",
            "    (13): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "    (14): Conv2d (128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (15): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (16): ReLU(inplace)\n",
            "    (17): Conv2d (256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (18): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (19): ReLU(inplace)\n",
            "    (20): Conv2d (256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (21): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (22): ReLU(inplace)\n",
            "    (23): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "    (24): Conv2d (256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (25): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (26): ReLU(inplace)\n",
            "    (27): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (28): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (29): ReLU(inplace)\n",
            "    (30): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (31): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (32): ReLU(inplace)\n",
            "    (33): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "    (34): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (35): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (36): ReLU(inplace)\n",
            "    (37): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (38): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (39): ReLU(inplace)\n",
            "    (40): Conv2d (512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "    (41): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True)\n",
            "    (42): ReLU(inplace)\n",
            "    (43): MaxPool2d(kernel_size=(2, 2), stride=(2, 2), dilation=(1, 1))\n",
            "  )\n",
            "  (classifier): Sequential(\n",
            "    (0): Linear(in_features=25088, out_features=4096)\n",
            "    (1): ReLU(inplace)\n",
            "    (2): Dropout(p=0.5)\n",
            "    (3): Linear(in_features=4096, out_features=4096)\n",
            "    (4): ReLU(inplace)\n",
            "    (5): Dropout(p=0.5)\n",
            "    (6): Linear(in_features=4096, out_features=10)\n",
            "  )\n",
            ")\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "kNLWfBCjL1Pa",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "b63b7607-c7fe-4256-cfef-0c060fe87571"
      },
      "cell_type": "code",
      "source": [
        "plt.ion()  \n",
        "\n",
        "use_gpu = torch.cuda.is_available()\n",
        "if use_gpu:\n",
        "    print(\"Using CUDA\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using CUDA\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "PHzacUGoKiM7",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "if use_gpu:\n",
        "    vgg16.cuda()\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = optim.SGD(vgg16.parameters(), lr=0.01, momentum=0.9)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Iy9Zde8MKmHu",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def eval_model(vgg16, criterion):\n",
        "    since = time.time()\n",
        "    test_loss = 0\n",
        "    test_acc = 0\n",
        "    \n",
        "    \n",
        "    print(\"evaluating model\")\n",
        "    print(\"-\"*10)\n",
        "    \n",
        "    for i, data in enumerate(test_loader):\n",
        "        vgg16.train(False)\n",
        "        vgg16.eval()\n",
        "        inputs, label = data\n",
        "        dataset_sizes = len(test_set)\n",
        "        \n",
        "        if use_gpu:\n",
        "            inputs, label = Variable(inputs.cuda(), volatile=True), Variable(label.cuda(), volatile=True)\n",
        "        else:\n",
        "            inputs, label = Variable(inputs, volatile =True), Variable(label, volatile=True)\n",
        "        \n",
        "        outputs=vgg16(inputs)\n",
        "        \n",
        "        _ , preds = torch.max(outputs.data, 1)\n",
        "        loss = criterion(outputs, label)\n",
        "        test_loss += loss.data[0]\n",
        "        test_acc += torch.sum(preds==label.data)\n",
        "        \n",
        "        del inputs, label, outputs, preds\n",
        "        torch.cuda.empty_cache()\n",
        "        \n",
        "    avg_loss = test_loss / dataset_sizes\n",
        "    avg_acc = test_acc / dataset_sizes\n",
        "        \n",
        "    elapsed_time = time.time()-since\n",
        "    print(\"test_loss {:.4f}\".format(avg_loss))\n",
        "    print(\"test_acc {:.4f}\".format(avg_acc))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "7caQ2tMPKpql",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 85
        },
        "outputId": "8ddbaaef-c856-4883-efa8-0c266bc717be"
      },
      "cell_type": "code",
      "source": [
        "model = eval_model(vgg16, criterion) "
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "evaluating model\n",
            "----------\n",
            "test_loss 0.0721\n",
            "test_acc 0.1020\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "XkZW_zv5W68d",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "img, label = next(iter(test_loader))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "hqa9CTaMLHfg",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def train_model(vgg16, criterion, optimizer, no_of_epochs=5):\n",
        "    since =time.time()\n",
        "    best_model_wts = copy.deepcopy(vgg16.state_dict())\n",
        "    best_acc = 0\n",
        "    \n",
        "    avg_loss = 0 \n",
        "    avg_acc = 0\n",
        "    avg_loss_val = 0\n",
        "    avg_acc_val = 0\n",
        "    \n",
        "    \n",
        "    train_batches = len(train_loader)\n",
        "    valid_batches = len(test_loader)\n",
        "    test_batches = len(test_loader)\n",
        "    \n",
        "    vgg16.train(True)\n",
        "    \n",
        "    for i, data in enumerate(train_loader):\n",
        "        if(i % 100 == 0):\n",
        "            print(\"\\r Training batch {}/{}\".format(i, train_batches/2), end ='', flush = True)\n",
        "            \n",
        "            loss_train = 0\n",
        "            loss_val = 0\n",
        "            acc_train = 0\n",
        "            acc_val = 0\n",
        "            #half training data \n",
        "            if i>= train_batches/2:\n",
        "                break\n",
        "                \n",
        "            inputs, labels = data\n",
        "            if use_gpu:\n",
        "                inputs, labels = Variable(inputs.cuda()), Variable(labels.cuda())\n",
        "                \n",
        "            else:\n",
        "                inputs, labels = Variable(inputs), Variable(labels)\n",
        "            \n",
        "            optimizer.zero_grad()\n",
        "            outputs=vgg16(inputs)\n",
        "            \n",
        "            _ , preds = torch.max(outputs.data, 1)\n",
        "            loss = criterion(outputs, labels)\n",
        "            \n",
        "            loss.backward()\n",
        "            optimizer.step()\n",
        "            \n",
        "            loss_train += loss.data[0]\n",
        "            acc_train += torch.sum(preds == labels.data)\n",
        "            \n",
        "            avg_loss = avg_loss*2/train_batches\n",
        "            avg_acc = avg_acc*2/train_batches\n",
        "            \n",
        "            vgg16.train(False)\n",
        "            vgg16.eval()\n",
        "            \n",
        "            for i, data in enumerate(test_loader):\n",
        "                if(i % 100 == 0):\n",
        "                    print(\"validation batch {}/{}\".format(i, valid_batches), end = '', flush=True)\n",
        "                \n",
        "                inputs, labels = data\n",
        "                \n",
        "                if use_gpu:\n",
        "                    inputs, labels = Variable(inputs.cuda()), Variable(labels.cuda())\n",
        "                \n",
        "                else:\n",
        "                    inputs, labels = Variable(inputs), Variable(labels)\n",
        "                \n",
        "                \n",
        "                optimizer.zero_grad()\n",
        "                \n",
        "                outputs = vgg16(inputs)\n",
        "                \n",
        "                _, preds = torch.max(outputs.data, 1)\n",
        "                loss = criterion(outputs, labels)\n",
        "                \n",
        "                loss_val +=loss.data[0]\n",
        "                acc_val +=torch.sum(preds==labels.data)\n",
        "                \n",
        "                avg_loss_val = loss_val /valid_batches\n",
        "                avg_acc_val = acc_val / valid_batches\n",
        "                \n",
        "                print()\n",
        "                print(\"Epoch {} result\".format(epoch))\n",
        "                print(\"train loss: {:.4f}\".format(avg_loss))\n",
        "                print(\"train acc: {:.4f}\".format(avg_acc))\n",
        "                print(\"valid loss: {:.4f}\".format(avg_loss_val))\n",
        "                print(\"valid acc: {:.4f}\".format(avg_acc_val))\n",
        "                \n",
        "                \n",
        "                if(avg_loss_val >= best_acc):\n",
        "                    best_acc = avg_loss_val\n",
        "                    best_model_wts = copy.deepcopy(vgg16.state_dict())\n",
        "                \n",
        "                elapsed_time = time.time() - since\n",
        "                \n",
        "                print(\"Training completed in {:.0f} minutes, {:.0f} seconds\".format(elapsed_time//60, elapsed_time%60))\n",
        "                \n",
        "                print(\"best acc: {:.4f}\".format(best_acc))\n",
        "                \n",
        "                vgg16.load_state_dict(best_model_wts)\n",
        "                return vgg16\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "ac2RFNWdLIxT",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "train_model(vgg16, criterion, optimizer, no_of_epochs=5)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "VynEnz76XfCU",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}