MNIST_Digit_Recognition_BN-L1_GN_LN.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MNIST Digit Recognition models with Normalization (BN+ L1, GN & LN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "jqrNA_iP8PLr",
    "outputId": "13af731d-8aab-4628-9f2b-46ec8855af61"
   },
   "outputs": [],
   "source": [
    "from model import build_model\n",
    "from __future__ import print_function\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "from torchvision import datasets, transforms\n",
    "from torch.autograd import Variable\n",
    "from torch.utils.data import DataLoader, Dataset\n",
    "from torchvision.utils import make_grid\n",
    "from torch.optim.lr_scheduler import StepLR,OneCycleLR\n",
    "import numpy as np\n",
    "import os\n",
    "\n",
    "#!pip install torchsummary\n",
    "from torchsummary import summary"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "t6vQR4Yt_QFO"
   },
   "source": [
    "# Reading MNIST data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "id": "mJZ5aHywr7xO"
   },
   "outputs": [],
   "source": [
    "# Train Phase transformations\n",
    "train_transforms = transforms.Compose([\n",
    "                                       transforms.RandomRotation((-7.0, 7.0), fill=(1,)),                \n",
    "                                       transforms.ColorJitter(brightness=0.4, contrast=0.4, saturation=0.40, hue=0.1),\n",
    "                                       transforms.ToTensor(),\n",
    "                                       transforms.Normalize((0.1307,), (0.3081,))\n",
    "                                       ])\n",
    "\n",
    "# Test Phase transformations\n",
    "test_transforms = transforms.Compose([\n",
    "                                       transforms.ToTensor(),\n",
    "                                       transforms.Normalize((0.1307,), (0.3081,))\n",
    "                                       ])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create Train and Test Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "EAUtmQAS94Ig"
   },
   "outputs": [],
   "source": [
    "train_data = datasets.MNIST('../data', train=True, download=True, transform=train_transforms)\n",
    "test_data = datasets.MNIST('../data', train=False, download=True,transform=test_transforms)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "NEs3ymph-H0i",
    "outputId": "56d29ec0-68ad-4400-81ef-5242947d5f41"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of training samples: 60000\n",
      "Number of test samples: 10000\n"
     ]
    }
   ],
   "source": [
    "n_train = len(train_data)\n",
    "n_test = len(test_data)\n",
    "\n",
    "print('Number of training samples: {0}'.format(n_train))\n",
    "print('Number of test samples: {0}'.format(n_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## If available, set the device to \"cuda\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "cuda = torch.cuda.is_available()\n",
    "device = torch.device(\"cuda\" if cuda else \"cpu\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Ige0pxq5_Fw7"
   },
   "source": [
    "# Model Architecture"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "K6v6YXTi6gc2"
   },
   "source": [
    "## Build the model from model.py \"build_model\" function"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Batch Normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "id": "rZqURC3aCFUf"
   },
   "outputs": [],
   "source": [
    "model_bn = build_model('BN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0lqi0uRhLsKq",
    "outputId": "8d0bd2da-d1a1-42d1-9f82-6165bda08575"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----------------------------------------------------------------\n",
      "        Layer (type)               Output Shape         Param #\n",
      "================================================================\n",
      "            Conv2d-1            [-1, 8, 26, 26]              72\n",
      "              ReLU-2            [-1, 8, 26, 26]               0\n",
      "       BatchNorm2d-3            [-1, 8, 26, 26]              16\n",
      "            Conv2d-4           [-1, 16, 24, 24]           1,152\n",
      "              ReLU-5           [-1, 16, 24, 24]               0\n",
      "       BatchNorm2d-6           [-1, 16, 24, 24]              32\n",
      "            Conv2d-7            [-1, 8, 24, 24]             128\n",
      "         MaxPool2d-8            [-1, 8, 12, 12]               0\n",
      "            Conv2d-9           [-1, 12, 10, 10]             864\n",
      "             ReLU-10           [-1, 12, 10, 10]               0\n",
      "      BatchNorm2d-11           [-1, 12, 10, 10]              24\n",
      "           Conv2d-12             [-1, 16, 8, 8]           1,728\n",
      "             ReLU-13             [-1, 16, 8, 8]               0\n",
      "      BatchNorm2d-14             [-1, 16, 8, 8]              32\n",
      "          Dropout-15             [-1, 16, 8, 8]               0\n",
      "           Conv2d-16             [-1, 16, 6, 6]           2,304\n",
      "             ReLU-17             [-1, 16, 6, 6]               0\n",
      "      BatchNorm2d-18             [-1, 16, 6, 6]              32\n",
      "          Dropout-19             [-1, 16, 6, 6]               0\n",
      "           Conv2d-20             [-1, 16, 6, 6]           2,304\n",
      "             ReLU-21             [-1, 16, 6, 6]               0\n",
      "      BatchNorm2d-22             [-1, 16, 6, 6]              32\n",
      "        AvgPool2d-23             [-1, 16, 1, 1]               0\n",
      "           Conv2d-24             [-1, 10, 1, 1]             160\n",
      "================================================================\n",
      "Total params: 8,880\n",
      "Trainable params: 8,880\n",
      "Non-trainable params: 0\n",
      "----------------------------------------------------------------\n",
      "Input size (MB): 0.00\n",
      "Forward/backward pass size (MB): 0.47\n",
      "Params size (MB): 0.03\n",
      "Estimated Total Size (MB): 0.51\n",
      "----------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "model_bn = model_bn.to(device)\n",
    "summary(model_bn, input_size=(1, 28, 28))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Rv4sT_U_6thw"
   },
   "source": [
    "## Group Normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "iX9QVS8Y6tDb"
   },
   "outputs": [],
   "source": [
    "model_gn = build_model('GN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0g8MbEeTL3NL",
    "outputId": "80e46f6f-c99c-491c-c0fe-f107fc2d24c2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----------------------------------------------------------------\n",
      "        Layer (type)               Output Shape         Param #\n",
      "================================================================\n",
      "            Conv2d-1            [-1, 8, 26, 26]              72\n",
      "              ReLU-2            [-1, 8, 26, 26]               0\n",
      "         GroupNorm-3            [-1, 8, 26, 26]              16\n",
      "            Conv2d-4           [-1, 16, 24, 24]           1,152\n",
      "              ReLU-5           [-1, 16, 24, 24]               0\n",
      "         GroupNorm-6           [-1, 16, 24, 24]              32\n",
      "            Conv2d-7            [-1, 8, 24, 24]             128\n",
      "         MaxPool2d-8            [-1, 8, 12, 12]               0\n",
      "            Conv2d-9           [-1, 12, 10, 10]             864\n",
      "             ReLU-10           [-1, 12, 10, 10]               0\n",
      "        GroupNorm-11           [-1, 12, 10, 10]              24\n",
      "           Conv2d-12             [-1, 16, 8, 8]           1,728\n",
      "             ReLU-13             [-1, 16, 8, 8]               0\n",
      "        GroupNorm-14             [-1, 16, 8, 8]              32\n",
      "          Dropout-15             [-1, 16, 8, 8]               0\n",
      "           Conv2d-16             [-1, 16, 6, 6]           2,304\n",
      "             ReLU-17             [-1, 16, 6, 6]               0\n",
      "        GroupNorm-18             [-1, 16, 6, 6]              32\n",
      "          Dropout-19             [-1, 16, 6, 6]               0\n",
      "           Conv2d-20             [-1, 16, 6, 6]           2,304\n",
      "             ReLU-21             [-1, 16, 6, 6]               0\n",
      "        GroupNorm-22             [-1, 16, 6, 6]              32\n",
      "        AvgPool2d-23             [-1, 16, 1, 1]               0\n",
      "           Conv2d-24             [-1, 10, 1, 1]             160\n",
      "================================================================\n",
      "Total params: 8,880\n",
      "Trainable params: 8,880\n",
      "Non-trainable params: 0\n",
      "----------------------------------------------------------------\n",
      "Input size (MB): 0.00\n",
      "Forward/backward pass size (MB): 0.47\n",
      "Params size (MB): 0.03\n",
      "Estimated Total Size (MB): 0.51\n",
      "----------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "model_gn = model_gn.to(device)\n",
    "summary(model_gn, input_size=(1, 28, 28))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cA1jYw_6-d8R"
   },
   "source": [
    "## Layer Normalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "XkbdPbtP-gip"
   },
   "outputs": [],
   "source": [
    "model_ln = build_model('LN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7yYRcxfL-FLD"
   },
   "source": [
    "### Model Summary"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "rcu4v1dcCKHu",
    "outputId": "3550f26b-3180-4f37-b6cd-059dc8021360"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "----------------------------------------------------------------\n",
      "        Layer (type)               Output Shape         Param #\n",
      "================================================================\n",
      "            Conv2d-1            [-1, 8, 26, 26]              72\n",
      "              ReLU-2            [-1, 8, 26, 26]               0\n",
      "         GroupNorm-3            [-1, 8, 26, 26]              16\n",
      "            Conv2d-4           [-1, 16, 24, 24]           1,152\n",
      "              ReLU-5           [-1, 16, 24, 24]               0\n",
      "         GroupNorm-6           [-1, 16, 24, 24]              32\n",
      "            Conv2d-7            [-1, 8, 24, 24]             128\n",
      "         MaxPool2d-8            [-1, 8, 12, 12]               0\n",
      "            Conv2d-9           [-1, 12, 10, 10]             864\n",
      "             ReLU-10           [-1, 12, 10, 10]               0\n",
      "        GroupNorm-11           [-1, 12, 10, 10]              24\n",
      "           Conv2d-12             [-1, 16, 8, 8]           1,728\n",
      "             ReLU-13             [-1, 16, 8, 8]               0\n",
      "        GroupNorm-14             [-1, 16, 8, 8]              32\n",
      "          Dropout-15             [-1, 16, 8, 8]               0\n",
      "           Conv2d-16             [-1, 16, 6, 6]           2,304\n",
      "             ReLU-17             [-1, 16, 6, 6]               0\n",
      "        GroupNorm-18             [-1, 16, 6, 6]              32\n",
      "          Dropout-19             [-1, 16, 6, 6]               0\n",
      "           Conv2d-20             [-1, 16, 6, 6]           2,304\n",
      "             ReLU-21             [-1, 16, 6, 6]               0\n",
      "        GroupNorm-22             [-1, 16, 6, 6]              32\n",
      "        AvgPool2d-23             [-1, 16, 1, 1]               0\n",
      "           Conv2d-24             [-1, 10, 1, 1]             160\n",
      "================================================================\n",
      "Total params: 8,880\n",
      "Trainable params: 8,880\n",
      "Non-trainable params: 0\n",
      "----------------------------------------------------------------\n",
      "Input size (MB): 0.00\n",
      "Forward/backward pass size (MB): 0.47\n",
      "Params size (MB): 0.03\n",
      "Estimated Total Size (MB): 0.51\n",
      "----------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "model_ln = model_ln.to(device)\n",
    "summary(model_ln, input_size=(1, 28, 28))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Function to set seed for all random value generation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "UlRA6GNm6Aia"
   },
   "outputs": [],
   "source": [
    "def seed_everything(seed):\n",
    "    os.environ['PYTHONHASHSEED'] = str(seed)\n",
    "    np.random.seed(seed)\n",
    "    torch.manual_seed(seed)\n",
    "    torch.cuda.manual_seed(seed)\n",
    "    torch.backends.cudnn.deterministic = True\n",
    "    torch.backends.cudnn.benchmark = True"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1PAL1z7a-ZiO"
   },
   "source": [
    "## Train Function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "W84d88IDYI2E"
   },
   "outputs": [],
   "source": [
    "def train(model, device, train_loader, optimizer, scheduler, epoch, L1, l1_factor):\n",
    "    model.train()\n",
    "    epoch_loss = 0\n",
    "    correct = 0\n",
    "\n",
    "    for batch_idx, (data, target) in enumerate(train_loader):\n",
    "        data, target = data.to(device), target.to(device)\n",
    "        optimizer.zero_grad()\n",
    "        output = model(data)\n",
    "        loss = F.nll_loss(output, target)\n",
    "\n",
    "        if L1:\n",
    "          L1_loss = nn.L1Loss(size_average=None, reduce=None, reduction='mean')\n",
    "          reg_loss = 0 \n",
    "          for param in model.parameters():\n",
    "            zero_vector = torch.rand_like(param) * 0\n",
    "            reg_loss += L1_loss(param,zero_vector)\n",
    "          loss += l1_factor * reg_loss\n",
    "\n",
    "        epoch_loss += loss.item()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        scheduler.step()\n",
    "        pred = output.argmax(dim=1, keepdim=True)  # get the index of the max log-probability\n",
    "        correct += pred.eq(target.view_as(pred)).sum().item()\n",
    "\n",
    "    print(f'Train set: Average loss: {loss.item():.4f}, Accuracy: {100. * correct/len(train_loader.dataset):.2f}')\n",
    "    train_loss = epoch_loss / len(train_loader)\n",
    "    train_acc=100.*correct/len(train_loader.dataset)\n",
    "    return train_loss, train_acc"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "iutOWfTJ-eHu"
   },
   "source": [
    "## Test/Validation Function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "id": "zxKGXJsJoddU"
   },
   "outputs": [],
   "source": [
    "def test(model, device, test_loader):\n",
    "    model.eval()\n",
    "    test_loss = 0\n",
    "    correct = 0\n",
    "    test_pred = torch.LongTensor()\n",
    "    target_pred = torch.LongTensor()\n",
    "    target_data = torch.LongTensor()\n",
    "\n",
    "    with torch.no_grad():\n",
    "        for data, target in test_loader:\n",
    "            data, target = data.to(device), target.to(device)\n",
    "            output = model(data)\n",
    "            test_loss += F.nll_loss(output, target, reduction='sum').item()  # sum up batch loss\n",
    "            pred = output.argmax(dim=1, keepdim=True)  # get the index of the max log-probability\n",
    "            pred_cpu = output.cpu().data.max(dim=1, keepdim=True)[1]\n",
    "            correct += pred.eq(target.view_as(pred)).sum().item()\n",
    "            test_pred = torch.cat((test_pred, pred_cpu), dim=0)\n",
    "            target_pred = torch.cat((target_pred, target.cpu()), dim=0)\n",
    "            target_data = torch.cat((target_data, data.cpu()), dim=0)\n",
    "\n",
    "\n",
    "    test_loss /= len(test_loader.dataset)\n",
    "    test_acc = 100.*correct/len(test_loader.dataset)\n",
    "    print(f'\\nTest set: Average loss: {test_loss:.3f}, Accuracy: {100. * correct/len(test_loader.dataset):.2f}')\n",
    "    return test_loss, test_acc, test_pred, target_pred, target_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Nu18tbPynB-E"
   },
   "source": [
    "## Function for plotting misclassified images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "id": "UPiGoutoZ3D3"
   },
   "outputs": [],
   "source": [
    "def misclassification(predictions, targets, data, xtitle):\n",
    "  pred = predictions.view(-1)\n",
    "  target = targets.view(-1)\n",
    "\n",
    "  index = 0\n",
    "  misclassified_image = []\n",
    "\n",
    "  for label, predict in zip(target, pred):\n",
    "    if label != predict:\n",
    "      misclassified_image.append(index)\n",
    "    index += 1\n",
    "\n",
    "  plt.figure(figsize=(10,5))\n",
    "  plt.suptitle(xtitle + ' - Misclassified Images');\n",
    "\n",
    "  for plot_index, bad_index in enumerate(misclassified_image[0:10]):\n",
    "    p = plt.subplot(2, 5, plot_index+1)\n",
    "    p.imshow(data[bad_index].reshape(28,28), cmap='gray')\n",
    "    p.axis('off')\n",
    "    p.set_title(f'Pred:{pred[bad_index]}, Actual:{target[bad_index]}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EExDDf3f-rST"
   },
   "source": [
    "## Main method for training and validating the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "id": "YdkLxcqcYOHu"
   },
   "outputs": [],
   "source": [
    "def main(EPOCH, model, device, train_loader, test_loader, optimizer, scheduler, L1, l1_factor):\n",
    "  train_loss_values = []\n",
    "  test_loss_values = []\n",
    "  train_acc_values = []\n",
    "  test_acc_values = []\n",
    "\n",
    "  for epoch in range(1, EPOCH + 1):\n",
    "      print('\\nEpoch {} : '.format(epoch))\n",
    "      # train the model\n",
    "      train_loss, train_acc = train(model, device, train_loader, optimizer, scheduler, epoch, L1, l1_factor)\n",
    "      test_loss, test_acc, test_pred, target_pred, target_data = test(model, device, test_loader)\n",
    "\n",
    "      train_loss_values.append(train_loss)\n",
    "      test_loss_values.append(test_loss)\n",
    "\n",
    "      train_acc_values.append(train_acc)\n",
    "      test_acc_values.append(test_acc)\n",
    "\n",
    "  return train_loss_values, test_loss_values, train_acc_values, test_acc_values, test_pred, target_pred, target_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Invoke the main method with all training parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "y5RICcjkTa2c",
    "outputId": "742ab4d2-01ba-425a-8fed-f562081c5e2e"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "------------------------------------------\n",
      "Batch Normalization with L1 regularization\n",
      "------------------------------------------\n",
      "\n",
      "Epoch 1 : \n",
      "Train set: Average loss: 0.2748, Accuracy: 76.42\n",
      "\n",
      "Test set: Average loss: 0.148, Accuracy: 97.17\n",
      "\n",
      "Epoch 2 : \n",
      "Train set: Average loss: 0.0699, Accuracy: 96.55\n",
      "\n",
      "Test set: Average loss: 0.070, Accuracy: 98.15\n",
      "\n",
      "Epoch 3 : \n",
      "Train set: Average loss: 0.1106, Accuracy: 97.66\n",
      "\n",
      "Test set: Average loss: 0.042, Accuracy: 98.77\n",
      "\n",
      "Epoch 4 : \n",
      "Train set: Average loss: 0.0923, Accuracy: 98.09\n",
      "\n",
      "Test set: Average loss: 0.042, Accuracy: 98.73\n",
      "\n",
      "Epoch 5 : \n",
      "Train set: Average loss: 0.0054, Accuracy: 98.42\n",
      "\n",
      "Test set: Average loss: 0.030, Accuracy: 99.09\n",
      "\n",
      "Epoch 6 : \n",
      "Train set: Average loss: 0.1970, Accuracy: 98.64\n",
      "\n",
      "Test set: Average loss: 0.025, Accuracy: 99.09\n",
      "\n",
      "Epoch 7 : \n",
      "Train set: Average loss: 0.0123, Accuracy: 98.74\n",
      "\n",
      "Test set: Average loss: 0.028, Accuracy: 99.15\n",
      "\n",
      "Epoch 8 : \n",
      "Train set: Average loss: 0.0133, Accuracy: 98.86\n",
      "\n",
      "Test set: Average loss: 0.029, Accuracy: 99.11\n",
      "\n",
      "Epoch 9 : \n",
      "Train set: Average loss: 0.0441, Accuracy: 98.88\n",
      "\n",
      "Test set: Average loss: 0.022, Accuracy: 99.26\n",
      "\n",
      "Epoch 10 : \n",
      "Train set: Average loss: 0.0132, Accuracy: 98.95\n",
      "\n",
      "Test set: Average loss: 0.025, Accuracy: 99.31\n",
      "\n",
      "Epoch 11 : \n",
      "Train set: Average loss: 0.1070, Accuracy: 99.01\n",
      "\n",
      "Test set: Average loss: 0.022, Accuracy: 99.28\n",
      "\n",
      "Epoch 12 : \n",
      "Train set: Average loss: 0.1052, Accuracy: 99.07\n",
      "\n",
      "Test set: Average loss: 0.024, Accuracy: 99.25\n",
      "\n",
      "Epoch 13 : \n",
      "Train set: Average loss: 0.0320, Accuracy: 99.05\n",
      "\n",
      "Test set: Average loss: 0.020, Accuracy: 99.37\n",
      "\n",
      "Epoch 14 : \n",
      "Train set: Average loss: 0.0151, Accuracy: 99.19\n",
      "\n",
      "Test set: Average loss: 0.019, Accuracy: 99.40\n",
      "\n",
      "Epoch 15 : \n",
      "Train set: Average loss: 0.0030, Accuracy: 99.23\n",
      "\n",
      "Test set: Average loss: 0.019, Accuracy: 99.32\n",
      "\n",
      "Epoch 16 : \n",
      "Train set: Average loss: 0.1759, Accuracy: 99.29\n",
      "\n",
      "Test set: Average loss: 0.019, Accuracy: 99.36\n",
      "\n",
      "Epoch 17 : \n",
      "Train set: Average loss: 0.0052, Accuracy: 99.33\n",
      "\n",
      "Test set: Average loss: 0.018, Accuracy: 99.45\n",
      "\n",
      "Epoch 18 : \n",
      "Train set: Average loss: 0.0074, Accuracy: 99.37\n",
      "\n",
      "Test set: Average loss: 0.017, Accuracy: 99.49\n",
      "\n",
      "Epoch 19 : \n",
      "Train set: Average loss: 0.0147, Accuracy: 99.43\n",
      "\n",
      "Test set: Average loss: 0.016, Accuracy: 99.50\n",
      "\n",
      "Epoch 20 : \n",
      "Train set: Average loss: 0.0085, Accuracy: 99.46\n",
      "\n",
      "Test set: Average loss: 0.016, Accuracy: 99.47\n",
      "------------------------------------------\n",
      "Group Normalization\n",
      "------------------------------------------\n",
      "\n",
      "Epoch 1 : \n",
      "Train set: Average loss: 0.3363, Accuracy: 71.38\n",
      "\n",
      "Test set: Average loss: 0.215, Accuracy: 95.67\n",
      "\n",
      "Epoch 2 : \n",
      "Train set: Average loss: 0.0646, Accuracy: 96.22\n",
      "\n",
      "Test set: Average loss: 0.075, Accuracy: 97.94\n",
      "\n",
      "Epoch 3 : \n",
      "Train set: Average loss: 0.0198, Accuracy: 97.51\n",
      "\n",
      "Test set: Average loss: 0.047, Accuracy: 98.56\n",
      "\n",
      "Epoch 4 : \n",
      "Train set: Average loss: 0.0443, Accuracy: 98.05\n",
      "\n",
      "Test set: Average loss: 0.040, Accuracy: 98.83\n",
      "\n",
      "Epoch 5 : \n",
      "Train set: Average loss: 0.1238, Accuracy: 98.26\n",
      "\n",
      "Test set: Average loss: 0.034, Accuracy: 98.95\n",
      "\n",
      "Epoch 6 : \n",
      "Train set: Average loss: 0.0167, Accuracy: 98.59\n",
      "\n",
      "Test set: Average loss: 0.026, Accuracy: 99.22\n",
      "\n",
      "Epoch 7 : \n",
      "Train set: Average loss: 0.0367, Accuracy: 98.75\n",
      "\n",
      "Test set: Average loss: 0.029, Accuracy: 99.18\n",
      "\n",
      "Epoch 8 : \n",
      "Train set: Average loss: 0.0134, Accuracy: 98.83\n",
      "\n",
      "Test set: Average loss: 0.029, Accuracy: 99.15\n",
      "\n",
      "Epoch 9 : \n",
      "Train set: Average loss: 0.0161, Accuracy: 98.95\n",
      "\n",
      "Test set: Average loss: 0.023, Accuracy: 99.31\n",
      "\n",
      "Epoch 10 : \n",
      "Train set: Average loss: 0.0511, Accuracy: 99.03\n",
      "\n",
      "Test set: Average loss: 0.023, Accuracy: 99.24\n",
      "\n",
      "Epoch 11 : \n",
      "Train set: Average loss: 0.1394, Accuracy: 99.13\n",
      "\n",
      "Test set: Average loss: 0.021, Accuracy: 99.34\n",
      "\n",
      "Epoch 12 : \n",
      "Train set: Average loss: 0.0047, Accuracy: 99.18\n",
      "\n",
      "Test set: Average loss: 0.020, Accuracy: 99.46\n",
      "\n",
      "Epoch 13 : \n",
      "Train set: Average loss: 0.0007, Accuracy: 99.25\n",
      "\n",
      "Test set: Average loss: 0.019, Accuracy: 99.33\n",
      "\n",
      "Epoch 14 : \n",
      "Train set: Average loss: 0.1390, Accuracy: 99.34\n",
      "\n",
      "Test set: Average loss: 0.019, Accuracy: 99.43\n",
      "\n",
      "Epoch 15 : \n",
      "Train set: Average loss: 0.0597, Accuracy: 99.32\n",
      "\n",
      "Test set: Average loss: 0.021, Accuracy: 99.35\n",
      "\n",
      "Epoch 16 : \n",
      "Train set: Average loss: 0.0043, Accuracy: 99.39\n",
      "\n",
      "Test set: Average loss: 0.016, Accuracy: 99.54\n",
      "\n",
      "Epoch 17 : \n",
      "Train set: Average loss: 0.0696, Accuracy: 99.48\n",
      "\n",
      "Test set: Average loss: 0.015, Accuracy: 99.49\n",
      "\n",
      "Epoch 18 : \n",
      "Train set: Average loss: 0.0023, Accuracy: 99.52\n",
      "\n",
      "Test set: Average loss: 0.014, Accuracy: 99.53\n",
      "\n",
      "Epoch 19 : \n",
      "Train set: Average loss: 0.0035, Accuracy: 99.54\n",
      "\n",
      "Test set: Average loss: 0.014, Accuracy: 99.57\n",
      "\n",
      "Epoch 20 : \n",
      "Train set: Average loss: 0.0274, Accuracy: 99.60\n",
      "\n",
      "Test set: Average loss: 0.014, Accuracy: 99.54\n",
      "------------------------------------------\n",
      "Layer Normalization\n",
      "------------------------------------------\n",
      "\n",
      "Epoch 1 : \n",
      "Train set: Average loss: 0.1288, Accuracy: 78.42\n",
      "\n",
      "Test set: Average loss: 0.193, Accuracy: 96.30\n",
      "\n",
      "Epoch 2 : \n",
      "Train set: Average loss: 0.2052, Accuracy: 96.18\n",
      "\n",
      "Test set: Average loss: 0.079, Accuracy: 97.81\n",
      "\n",
      "Epoch 3 : \n",
      "Train set: Average loss: 0.0116, Accuracy: 97.61\n",
      "\n",
      "Test set: Average loss: 0.065, Accuracy: 97.96\n",
      "\n",
      "Epoch 4 : \n",
      "Train set: Average loss: 0.0066, Accuracy: 98.11\n",
      "\n",
      "Test set: Average loss: 0.039, Accuracy: 98.84\n",
      "\n",
      "Epoch 5 : \n",
      "Train set: Average loss: 0.0034, Accuracy: 98.47\n",
      "\n",
      "Test set: Average loss: 0.032, Accuracy: 99.02\n",
      "\n",
      "Epoch 6 : \n",
      "Train set: Average loss: 0.0874, Accuracy: 98.68\n",
      "\n",
      "Test set: Average loss: 0.030, Accuracy: 99.07\n",
      "\n",
      "Epoch 7 : \n",
      "Train set: Average loss: 0.0018, Accuracy: 98.80\n",
      "\n",
      "Test set: Average loss: 0.034, Accuracy: 99.01\n",
      "\n",
      "Epoch 8 : \n",
      "Train set: Average loss: 0.0151, Accuracy: 98.89\n",
      "\n",
      "Test set: Average loss: 0.025, Accuracy: 99.26\n",
      "\n",
      "Epoch 9 : \n",
      "Train set: Average loss: 0.0030, Accuracy: 99.01\n",
      "\n",
      "Test set: Average loss: 0.030, Accuracy: 99.01\n",
      "\n",
      "Epoch 10 : \n",
      "Train set: Average loss: 0.0231, Accuracy: 99.01\n",
      "\n",
      "Test set: Average loss: 0.026, Accuracy: 99.20\n",
      "\n",
      "Epoch 11 : \n",
      "Train set: Average loss: 0.0084, Accuracy: 99.18\n",
      "\n",
      "Test set: Average loss: 0.024, Accuracy: 99.20\n",
      "\n",
      "Epoch 12 : \n",
      "Train set: Average loss: 0.0018, Accuracy: 99.18\n",
      "\n",
      "Test set: Average loss: 0.021, Accuracy: 99.34\n",
      "\n",
      "Epoch 13 : \n",
      "Train set: Average loss: 0.0070, Accuracy: 99.24\n",
      "\n",
      "Test set: Average loss: 0.022, Accuracy: 99.24\n",
      "\n",
      "Epoch 14 : \n",
      "Train set: Average loss: 0.0247, Accuracy: 99.30\n",
      "\n",
      "Test set: Average loss: 0.018, Accuracy: 99.42\n",
      "\n",
      "Epoch 15 : \n",
      "Train set: Average loss: 0.0283, Accuracy: 99.37\n",
      "\n",
      "Test set: Average loss: 0.020, Accuracy: 99.42\n",
      "\n",
      "Epoch 16 : \n",
      "Train set: Average loss: 0.0136, Accuracy: 99.44\n",
      "\n",
      "Test set: Average loss: 0.021, Accuracy: 99.36\n",
      "\n",
      "Epoch 17 : \n",
      "Train set: Average loss: 0.0020, Accuracy: 99.47\n",
      "\n",
      "Test set: Average loss: 0.018, Accuracy: 99.43\n",
      "\n",
      "Epoch 18 : \n",
      "Train set: Average loss: 0.0033, Accuracy: 99.54\n",
      "\n",
      "Test set: Average loss: 0.018, Accuracy: 99.47\n",
      "\n",
      "Epoch 19 : \n",
      "Train set: Average loss: 0.0101, Accuracy: 99.58\n",
      "\n",
      "Test set: Average loss: 0.018, Accuracy: 99.46\n",
      "\n",
      "Epoch 20 : \n",
      "Train set: Average loss: 0.0010, Accuracy: 99.61\n",
      "\n",
      "Test set: Average loss: 0.018, Accuracy: 99.48\n"
     ]
    }
   ],
   "source": [
    "if __name__=='__main__':\n",
    "\n",
    "  EPOCH = 20\n",
    "  batch_size = 64\n",
    "  l1_factor = 0.0001\n",
    "\n",
    "  cuda = torch.cuda.is_available()\n",
    "  device = torch.device(\"cuda\" if cuda else \"cpu\")\n",
    "\n",
    "  seed_everything(1)\n",
    "\n",
    "  kwargs = {'num_workers': 1, 'pin_memory': True} if cuda else {}\n",
    "  train_loader = torch.utils.data.DataLoader(train_data, batch_size=batch_size, shuffle=True, **kwargs)\n",
    "  test_loader = torch.utils.data.DataLoader(test_data, batch_size=batch_size, shuffle=True, **kwargs)\n",
    "  \n",
    "  model = build_model('BN').to(device)\n",
    "  optimizer = optim.SGD(model.parameters(), lr=0.02, momentum=0.8)\n",
    "  scheduler = OneCycleLR(optimizer, max_lr=0.015,epochs=20,steps_per_epoch=len(train_loader))\n",
    "  \n",
    "  print(\"------------------------------------------\")\n",
    "  print(\"Batch Normalization with L1 regularization\")\n",
    "  print(\"------------------------------------------\")\n",
    "  BN_train_loss, BN_test_loss, BN_train_acc, BN_test_acc, BN_test_pred, BN_target_pred, BN_target_data = main(EPOCH, model, device, train_loader, test_loader, optimizer, scheduler, True, l1_factor)\n",
    "\n",
    "  model = build_model('GN').to(device)\n",
    "  optimizer = optim.SGD(model.parameters(), lr=0.02, momentum=0.8)\n",
    "  scheduler = OneCycleLR(optimizer, max_lr=0.015,epochs=20,steps_per_epoch=len(train_loader))\n",
    "\n",
    "  print(\"------------------------------------------\")\n",
    "  print(\"Group Normalization\")\n",
    "  print(\"------------------------------------------\")\n",
    "  GN_train_loss, GN_test_loss, GN_train_acc, GN_test_acc, GN_test_pred, GN_target_pred, GN_target_data = main(EPOCH, model, device, train_loader, test_loader, optimizer, scheduler, False, l1_factor)\n",
    "\n",
    "  model = build_model('LN').to(device)\n",
    "  optimizer = optim.SGD(model.parameters(), lr=0.02, momentum=0.8)\n",
    "  scheduler = OneCycleLR(optimizer, max_lr=0.015,epochs=20,steps_per_epoch=len(train_loader))\n",
    "\n",
    "  print(\"------------------------------------------\")\n",
    "  print(\"Layer Normalization\")\n",
    "  print(\"------------------------------------------\")\n",
    "  LN_train_loss, LN_test_loss, LN_train_acc, LN_test_acc, LN_test_pred, LN_target_pred, LN_target_data = main(EPOCH, model, device, train_loader, test_loader, optimizer, scheduler, False, l1_factor)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "10oipdea-wg1"
   },
   "source": [
    "# Plotting the train and test accuracy & loss across each epoch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 391
    },
    "id": "kDXBDZP1YmDd",
    "outputId": "e7b1c167-d200-4e91-8514-448618c53938",
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x432 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "import seaborn as sns\n",
    "\n",
    "# Use plot styling from seaborn.\n",
    "sns.set(style='whitegrid')\n",
    "\n",
    "# Increase the plot size and font size.\n",
    "sns.set(font_scale=1)\n",
    "plt.rcParams[\"figure.figsize\"] = (25,6)\n",
    "\n",
    "# Plot the learning curve.\n",
    "fig, (ax1,ax2) = plt.subplots(1,2)\n",
    "ax1.plot(np.array(BN_train_loss), 'red', label=\"BN + L1 Training Loss\")\n",
    "ax1.plot(np.array(GN_train_loss), 'green', label=\"GN Training Loss\")\n",
    "ax1.plot(np.array(LN_train_loss), 'yellow', label=\"LN Training Loss\")\n",
    "ax1.plot(np.array(BN_test_loss), 'blue', label=\"BN + L1 Test Loss\")\n",
    "ax1.plot(np.array(GN_test_loss), 'black', label=\"GN Test Loss\")\n",
    "ax1.plot(np.array(LN_test_loss), 'purple', label=\"LN Test Loss\")\n",
    "\n",
    "# Label the plot.\n",
    "ax1.set_title(\"Training & Validation Loss\")\n",
    "ax1.set_xlabel(\"Epoch\")\n",
    "ax1.set_ylabel(\"Loss\")\n",
    "ax1.set_ylim(0,0.2)\n",
    "ax1.legend()\n",
    "\n",
    "ax2.plot(np.array(BN_train_acc), 'red', label=\"BN + L1 Training Accuracy\")\n",
    "ax2.plot(np.array(GN_train_acc), 'green', label=\"GN Training Accuracy\")\n",
    "ax2.plot(np.array(LN_train_acc), 'yellow', label=\"LN Training Accuracy\")\n",
    "ax2.plot(np.array(BN_test_acc), 'blue', label=\"BN + L1 Test Accuracy\")\n",
    "ax2.plot(np.array(GN_test_acc), 'black', label=\"GN Test Accuracy\")\n",
    "ax2.plot(np.array(LN_test_acc), 'purple', label=\"LN Test Accuracy\")\n",
    "\n",
    "# Label the plot.\n",
    "ax2.set_title(\"Training & Validation Accuracy\")\n",
    "ax2.set_xlabel(\"Epoch\")\n",
    "ax2.set_ylabel(\"Loss\")\n",
    "ax2.set_ylim(95,100)\n",
    "ax2.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Identify and plot 10 misclassified images from test data for all 3 models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WM4ekTr2mv5g"
   },
   "source": [
    "## Batch Normalization + L1 Norm Misclassified Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 324
    },
    "id": "XU4u6t7C4O9F",
    "outputId": "f34c9ed0-a6c8-443c-af04-bd551a46f1cd"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "misclassification(BN_test_pred, BN_target_pred, BN_target_data, 'Batch Normalization + L1')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "UJqC87f-ngNO"
   },
   "source": [
    "## Group Normalization Misclassified Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 324
    },
    "id": "ffx6WZTOnQWW",
    "outputId": "f6299f8e-acb9-4463-c33c-05609fdf4a20"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "misclassification(GN_test_pred, GN_target_pred, GN_target_data, 'Group Normalization')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dXMR5yIjnlJe"
   },
   "source": [
    "## Layer Normalization Misclassified Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 324
    },
    "id": "_QacCHNGnQRq",
    "outputId": "8ef76c36-d1c5-4254-abb3-eac54148bdc2"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "misclassification(LN_test_pred, LN_target_pred, LN_target_data, 'Layer Normalization')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "authorship_tag": "ABX9TyOhFOuLu/Q1V6RTtgOasj1V",
   "collapsed_sections": [],
   "include_colab_link": true,
   "name": "MNIST Digit Recognition Batch Normalization and Regularization.ipynb",
   "provenance": []
  },
  "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": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}