10-[MobileNet]Efficient Convolutional Neural Networks for Mobile Vision Applications.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 10-[MobileNet]Efficient Convolutional Neural Networks for Mobile Vision Applications"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[1.Depth-Wise Convolution](#1)\n",
    "  - [1-1.Standard Convolution](#1-1)\n",
    "  - [1-2.Depthwise Convoluton](#1-2)\n",
    "\n",
    "[2.Point-Wise Convolution](#2)\n",
    "\n",
    "[3.Depth-Wise Separable Convolution](#3)\n",
    "\n",
    "[4.MobileNet](#4)\n",
    "\n",
    "[5.Training](#5)\n",
    " - [5-1.Hyper Parameters](#5-1)\n",
    " - [5-2.Load Dataset](#5-2)\n",
    " - [5-3.Train & Valid functions](#5-3)\n",
    " - [5-4.Training iteration](#5-4)\n",
    "\n",
    "[6.Conv MobileNet](#6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import cv2\n",
    "import torch\n",
    "import albumentations as A\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from torch import nn\n",
    "from tqdm import tqdm\n",
    "from torchsummary import summary\n",
    "from albumentations.pytorch import ToTensorV2\n",
    "from torch.optim.lr_scheduler import LambdaLR\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"1\"></a>\n",
    "## 1.Depth Wise Convoltion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "일반적인 Convolution은 입력된 이미지의 채널과 같은 크기의 채널로 구성된 필터를 적용한다.\n",
    "\n",
    "반면, Depthwise convolution은 입력 이미지의 채널마다 하나의 필터가 적용된다.\n",
    "\n",
    "따라서 더 작은 파라미터 수를 가지며, 입력의 height, width만 조절할 뿐, channel의 값은 변하지 않는다."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"1-1\"></a>\n",
    "### 1-1.Standard Convolution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 64, 224, 224])\n"
     ]
    }
   ],
   "source": [
    "dummy_input = torch.randn(1, 3, 224, 224)\n",
    "\n",
    "standard_conv = nn.Conv2d(in_channels=3, out_channels=64, kernel_size=3, stride=1, padding=1)\n",
    "standard_output = standard_conv(dummy_input)\n",
    "print(standard_output.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output shape: torch.Size([1, 64, 224, 224])\n",
      "Number of parameters: 1792\n"
     ]
    }
   ],
   "source": [
    "num_params = sum(p.numel() for p in standard_conv.parameters()) ## 각 파라미터 텐서 p에 있는 원소의 총 개수\n",
    "\n",
    "print(f\"Output shape: {standard_output.shape}\")\n",
    "print(f\"Number of parameters: {num_params}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "computation_complexity: 173408256\n"
     ]
    }
   ],
   "source": [
    "computation_complexity = (\n",
    "    2 * \n",
    "    standard_output.shape[1] * ## num filters\n",
    "    standard_output.shape[2] * ## height\n",
    "    standard_output.shape[3] * ## width\n",
    "    standard_conv.kernel_size[0] * ## filter size\n",
    "    standard_conv.kernel_size[1] * ## filter size\n",
    "    standard_conv.in_channels ## input channels\n",
    ")\n",
    "\n",
    "print(f\"computation_complexity: {computation_complexity}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"1-2\"></a>\n",
    "### 1-2.Depthwise Convolution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 3, 224, 224])\n"
     ]
    }
   ],
   "source": [
    "depth_wise_conv = nn.Conv2d(in_channels=3, out_channels=3, kernel_size=3, stride=1, padding=1, groups=3)\n",
    "depth_wise_output = depth_wise_conv(dummy_input)\n",
    "print(depth_wise_output.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output shape: torch.Size([1, 3, 224, 224])\n",
      "Number of parameters: 30\n"
     ]
    }
   ],
   "source": [
    "num_params = sum(p.numel() for p in depth_wise_conv.parameters()) ## 각 파라미터 텐서 p에 있는 원소의 총 개수\n",
    "\n",
    "print(f\"Output shape: {depth_wise_output.shape}\")\n",
    "print(f\"Number of parameters: {num_params}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Computation complexity: 1354752\n"
     ]
    }
   ],
   "source": [
    "output_height, output_width = depth_wise_output.shape[2], depth_wise_output.shape[3]\n",
    "computation_complexity = 3 * 3 * 3 * output_height * output_width\n",
    "print(f\"Computation complexity: {computation_complexity}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"2\"></a>\n",
    "## 2.Pointwise Convolution"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "이전 내용에서도 몇 번 사용했던, $1\\times 1$ convolution으로, 입력되는 채널의 값이 depthwise convolution의 out_channels와 같기 때문에, 해상도의 크기는 조정되지 않는다.\n",
    "\n",
    "단, pointwise convolution의 out_channels는 설정하는 값에 따라 변화된다.(채널의 크기를 줄이거나 증가시키거나 둘 중 하나.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 3, 224, 224])\n",
      "torch.Size([1, 1, 224, 224])\n"
     ]
    }
   ],
   "source": [
    "depth_wise_conv = nn.Conv2d(in_channels=3, out_channels=3, kernel_size=3, stride=1, padding=1, groups=3)\n",
    "depth_wise_output = depth_wise_conv(dummy_input)\n",
    "print(depth_wise_output.shape)\n",
    "\n",
    "point_wise_conv = nn.Conv2d(in_channels=3, out_channels=1, kernel_size=1)\n",
    "point_wise_output = point_wise_conv(depth_wise_output)\n",
    "print(point_wise_output.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Output shape: torch.Size([1, 1, 224, 224])\n",
      "Number of parameters: 4\n"
     ]
    }
   ],
   "source": [
    "num_params = sum(p.numel() for p in point_wise_conv.parameters()) ## 각 파라미터 텐서 p에 있는 원소의 총 개수\n",
    "\n",
    "print(f\"Output shape: {point_wise_output.shape}\")\n",
    "print(f\"Number of parameters: {num_params}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"3\"></a>\n",
    "## 3.Depthwise Separable Convolution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DepthwiseSeparableConv2d(nn.Module):\n",
    "    def __init__(self, in_channels, out_channels, stride=1, width_multiplier=1.0):\n",
    "        super().__init__()\n",
    "        in_channels = int(in_channels * width_multiplier)\n",
    "        out_channels = int(out_channels * width_multiplier)\n",
    "\n",
    "        self.depth_wise_conv = nn.Conv2d(in_channels=in_channels, out_channels=in_channels, kernel_size=3, stride=stride, padding=1, groups=in_channels, bias=False)\n",
    "        self.bn1 = nn.BatchNorm2d(in_channels)\n",
    "        \n",
    "        self.point_wise_conv = nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=1, stride=1, padding=0, bias=False)\n",
    "        self.bn2 = nn.BatchNorm2d(out_channels)\n",
    "\n",
    "        self.relu = nn.ReLU(inplace=True)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.depth_wise_conv(x)\n",
    "        x = self.bn1(x)\n",
    "        x = self.relu(x)\n",
    "\n",
    "        x = self.point_wise_conv(x)\n",
    "        x = self.bn2(x)\n",
    "        x = self.relu(x)\n",
    "\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 64, 224, 224])\n"
     ]
    }
   ],
   "source": [
    "dummy_input = torch.randn(1, 3, 224, 224)\n",
    "depth_wise_sep_conv = DepthwiseSeparableConv2d(in_channels=3, out_channels=64)\n",
    "output = depth_wise_sep_conv(dummy_input)\n",
    "print(output.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"4\"></a>\n",
    "## 4.MobileNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MobileNet(nn.Module):\n",
    "    def __init__(self, num_classes=1000, width_multiplier=1.0, init_weights=False):\n",
    "        super().__init__()\n",
    "        self.feature_extractor = nn.Sequential(\n",
    "            nn.Conv2d(in_channels=3, out_channels=32, kernel_size=3, stride=2, padding=1, bias=False),\n",
    "            nn.BatchNorm2d(32),\n",
    "            nn.ReLU(inplace=True),\n",
    "\n",
    "            DepthwiseSeparableConv2d(in_channels=32, out_channels=64, stride=1, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=64, out_channels=128, stride=2, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=128, out_channels=128, stride=1, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=128, out_channels=256, stride=2, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=256, out_channels=256, stride=1, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=256, out_channels=512, stride=2, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=512, out_channels=512, stride=1, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=512, out_channels=512, stride=1, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=512, out_channels=512, stride=1, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=512, out_channels=512, stride=1, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=512, out_channels=512, stride=1, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=512, out_channels=1024, stride=2, width_multiplier=width_multiplier),\n",
    "            DepthwiseSeparableConv2d(in_channels=1024, out_channels=1024, stride=1, width_multiplier=width_multiplier),\n",
    "        )\n",
    "\n",
    "        self.avg_pool = nn.AdaptiveAvgPool2d((1, 1))\n",
    "        self.classifier = nn.Linear(1024, num_classes)\n",
    "\n",
    "        if init_weights:\n",
    "            for m in self.modules():\n",
    "                if isinstance(m, nn.Conv2d):\n",
    "                    nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')\n",
    "                    if m.bias is not None:\n",
    "                        nn.init.constant_(m.bias, 0)\n",
    "                elif isinstance(m, nn.BatchNorm2d):\n",
    "                    nn.init.constant_(m.weight, 1)\n",
    "                    nn.init.constant_(m.bias, 0)\n",
    "                elif isinstance(m, nn.Linear):\n",
    "                    nn.init.normal_(m.weight, 0, 0.01)\n",
    "                    nn.init.constant_(m.bias, 0)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.feature_extractor(x)\n",
    "        x = self.avg_pool(x)\n",
    "        x = x.view(-1, 1024)\n",
    "        x = self.classifier(x)\n",
    "\n",
    "        return x"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"5\"></a>\n",
    "## 5.Training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"5-1\"></a>\n",
    "### 5-1.Hyper Parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dir = \"/home/pervinco/Datasets/mammals-classification/data\"\n",
    "save_dir = \"./runs/mammal-classification/resnet\"\n",
    "\n",
    "epochs = 300\n",
    "batch_size = 64\n",
    "\n",
    "learning_rate = 0.001\n",
    "lr_patience = 10\n",
    "lr_decay_rate = 0.98\n",
    "weight_decay = 0.00002\n",
    "\n",
    "early_stop_counter = 0\n",
    "early_stop_patience = 20\n",
    "\n",
    "num_workers = os.cpu_count()\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"5-2\"></a>\n",
    "### 5-2.Load Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['african_elephant', 'alpaca', 'american_bison', 'anteater', 'arctic_fox', 'armadillo', 'baboon', 'badger', 'blue_whale', 'brown_bear', 'camel', 'dolphin', 'giraffe', 'groundhog', 'highland_cattle', 'horse', 'jackal', 'kangaroo', 'koala', 'manatee', 'mongoose', 'mountain_goat', 'opossum', 'orangutan', 'otter', 'polar_bear', 'porcupine', 'red_panda', 'rhinoceros', 'sea_lion', 'seal', 'snow_leopard', 'squirrel', 'sugar_glider', 'tapir', 'vampire_bat', 'vicuna', 'walrus', 'warthog', 'water_buffalo', 'weasel', 'wildebeest', 'wombat', 'yak', 'zebra']\n",
      "Num of Classes : 45\n",
      "Num of files & labels : 13751, 13751\n",
      "/home/pervinco/Datasets/mammals-classification/data/african_elephant/african_elephant-0125.jpg\n",
      "12375 12375\n",
      "1376 1376\n"
     ]
    }
   ],
   "source": [
    "classes = sorted(os.listdir(data_dir))\n",
    "print(classes)\n",
    "\n",
    "images, labels = [], []\n",
    "for str_label in classes:\n",
    "    img_files = os.listdir(f\"{data_dir}/{str_label}\")\n",
    "    for file in img_files:\n",
    "        if file.endswith(\".jpg\"):\n",
    "            images.append(f\"{data_dir}/{str_label}/{file}\")\n",
    "            labels.append(str_label)\n",
    "\n",
    "print(f\"Num of Classes : {len(classes)}\")\n",
    "print(f\"Num of files & labels : {len(images)}, {len(labels)}\")\n",
    "print(images[0])\n",
    "\n",
    "train_x, valid_x, train_y, valid_y = train_test_split(images, labels, test_size=0.1, shuffle=True, random_state=42)\n",
    "\n",
    "print(len(train_x), len(train_y))\n",
    "print(len(valid_x), len(valid_y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MammalDataset(Dataset):\n",
    "    def __init__(self, images, labels, classes, mean=None, std=None, augmentation=False):\n",
    "        self.classes = classes\n",
    "        self.images = images\n",
    "        self.labels = labels\n",
    "\n",
    "        if mean is None:\n",
    "            mean = (0.485, 0.456, 0.406)\n",
    "\n",
    "        if std is None:\n",
    "            std = (0.229, 0.224, 0.225)\n",
    "\n",
    "        if augmentation:\n",
    "            self.augmentation = A.Compose([\n",
    "                A.OneOf([\n",
    "                    A.Resize(224, 224, p=0.5),\n",
    "                    A.SmallestMaxSize(max_size=[256, 512], p=0.5)\n",
    "                ], p=1),\n",
    "                A.RandomCrop(width=224, height=224),\n",
    "\n",
    "                A.HorizontalFlip(p=0.4),\n",
    "                A.RGBShift(p=0.3),\n",
    "\n",
    "                A.ShiftScaleRotate(p=1.0, shift_limit_x=(-0.06, 0.06), shift_limit_y=(-0.06, 0.06), \n",
    "                                   scale_limit=(-0.099, 0.100), \n",
    "                                   rotate_limit=(-180, 180), \n",
    "                                   interpolation=0, \n",
    "                                   border_mode=0, \n",
    "                                   value=(0, 0, 0)),\n",
    "                A.RandomBrightnessContrast(p=1.0, brightness_limit=(-0.2, 0.2), contrast_limit=(-0.2, 0.2), brightness_by_max=True),\n",
    "                A.AdvancedBlur(p=0.35, blur_limit=(3, 7), \n",
    "                               sigmaX_limit=(0.2, 1.0), \n",
    "                               sigmaY_limit=(0.2, 1.0), \n",
    "                               rotate_limit=(-90, 90), \n",
    "                               beta_limit=(0.5, 8.0), \n",
    "                               noise_limit=(0.9, 1.1)),\n",
    "\n",
    "                A.Normalize(mean=mean, std=std),\n",
    "                ToTensorV2()\n",
    "            ])\n",
    "        else:\n",
    "            self.augmentation = A.Compose([\n",
    "                A.Resize(224, 224, p=1),\n",
    "                A.Normalize(mean=mean, std=std),\n",
    "                ToTensorV2()\n",
    "            ])\n",
    "\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.images)\n",
    "\n",
    "\n",
    "    def __getitem__(self, idx):\n",
    "        image = self.images[idx]\n",
    "        label = self.labels[idx]\n",
    "\n",
    "        image = cv2.imread(image)\n",
    "        image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)\n",
    "        label = self.classes.index(label)\n",
    "\n",
    "        image = self.augmentation(image=image)[\"image\"]\n",
    "\n",
    "        return image, label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataset = MammalDataset(train_x, train_y, classes, augmentation=True)\n",
    "valid_dataset = MammalDataset(valid_x, valid_y, classes, augmentation=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataloder = DataLoader(train_dataset, batch_size=batch_size, shuffle=True, num_workers=num_workers)\n",
    "valid_dataloader = DataLoader(valid_dataset, batch_size=batch_size, num_workers=num_workers)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"5-3\"></a>\n",
    "### 5-3.Train & Valid functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train(model, dataloader, criterion, optimizer, device):\n",
    "    model.train()\n",
    "\n",
    "    train_cost, train_acc = 0.0, 0.0\n",
    "    for x, y in tqdm(dataloader, desc=\"Train\", leave=False):\n",
    "        x, y = x.to(device), y.to(device)\n",
    "\n",
    "        optimizer.zero_grad()\n",
    "        y_pred = model(x)\n",
    "\n",
    "        cost = criterion(y_pred, y)\n",
    "        cost.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        train_cost += cost.item() * x.size(0)\n",
    "\n",
    "        _, pred_labels = torch.max(y_pred, 1)\n",
    "        train_acc += (pred_labels == y).sum().item()\n",
    "\n",
    "    train_cost /= len(dataloader.dataset)\n",
    "    train_acc /= len(dataloader.dataset)\n",
    "\n",
    "    return train_cost, train_acc\n",
    "\n",
    "\n",
    "def valid(model, dataloader, criterion, device):\n",
    "    model.eval()\n",
    "\n",
    "    valid_cost, valid_acc = 0.0, 0.0\n",
    "    with torch.no_grad():\n",
    "        for x, y in tqdm(dataloader, desc=\"Valid\", leave=False):\n",
    "            x, y = x.to(device), y.to(device)\n",
    "            y_pred = model(x)\n",
    "\n",
    "            cost = criterion(y_pred, y)\n",
    "\n",
    "            valid_cost += cost.item() * x.size(0)\n",
    "\n",
    "            _, pred_labels = torch.max(y_pred, 1)\n",
    "            valid_acc += (pred_labels == y).sum().item()\n",
    "\n",
    "        valid_cost /= len(dataloader.dataset)\n",
    "        valid_acc /= len(dataloader.dataset)\n",
    "\n",
    "    return valid_cost, valid_acc\n",
    "\n",
    "def lr_lambda(epoch):\n",
    "    return lr_decay_rate ** epoch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"5-4\"></a>\n",
    "### 5-4.Training Iteration"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "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",
      "              ReLU-3         [-1, 32, 112, 112]               0\n",
      "            Conv2d-4         [-1, 32, 112, 112]             288\n",
      "       BatchNorm2d-5         [-1, 32, 112, 112]              64\n",
      "              ReLU-6         [-1, 32, 112, 112]               0\n",
      "            Conv2d-7         [-1, 64, 112, 112]           2,048\n",
      "       BatchNorm2d-8         [-1, 64, 112, 112]             128\n",
      "              ReLU-9         [-1, 64, 112, 112]               0\n",
      "DepthwiseSeparableConv2d-10         [-1, 64, 112, 112]               0\n",
      "           Conv2d-11           [-1, 64, 56, 56]             576\n",
      "      BatchNorm2d-12           [-1, 64, 56, 56]             128\n",
      "             ReLU-13           [-1, 64, 56, 56]               0\n",
      "           Conv2d-14          [-1, 128, 56, 56]           8,192\n",
      "      BatchNorm2d-15          [-1, 128, 56, 56]             256\n",
      "             ReLU-16          [-1, 128, 56, 56]               0\n",
      "DepthwiseSeparableConv2d-17          [-1, 128, 56, 56]               0\n",
      "           Conv2d-18          [-1, 128, 56, 56]           1,152\n",
      "      BatchNorm2d-19          [-1, 128, 56, 56]             256\n",
      "             ReLU-20          [-1, 128, 56, 56]               0\n",
      "           Conv2d-21          [-1, 128, 56, 56]          16,384\n",
      "      BatchNorm2d-22          [-1, 128, 56, 56]             256\n",
      "             ReLU-23          [-1, 128, 56, 56]               0\n",
      "DepthwiseSeparableConv2d-24          [-1, 128, 56, 56]               0\n",
      "           Conv2d-25          [-1, 128, 28, 28]           1,152\n",
      "      BatchNorm2d-26          [-1, 128, 28, 28]             256\n",
      "             ReLU-27          [-1, 128, 28, 28]               0\n",
      "           Conv2d-28          [-1, 256, 28, 28]          32,768\n",
      "      BatchNorm2d-29          [-1, 256, 28, 28]             512\n",
      "             ReLU-30          [-1, 256, 28, 28]               0\n",
      "DepthwiseSeparableConv2d-31          [-1, 256, 28, 28]               0\n",
      "           Conv2d-32          [-1, 256, 28, 28]           2,304\n",
      "      BatchNorm2d-33          [-1, 256, 28, 28]             512\n",
      "             ReLU-34          [-1, 256, 28, 28]               0\n",
      "           Conv2d-35          [-1, 256, 28, 28]          65,536\n",
      "      BatchNorm2d-36          [-1, 256, 28, 28]             512\n",
      "             ReLU-37          [-1, 256, 28, 28]               0\n",
      "DepthwiseSeparableConv2d-38          [-1, 256, 28, 28]               0\n",
      "           Conv2d-39          [-1, 256, 14, 14]           2,304\n",
      "      BatchNorm2d-40          [-1, 256, 14, 14]             512\n",
      "             ReLU-41          [-1, 256, 14, 14]               0\n",
      "           Conv2d-42          [-1, 512, 14, 14]         131,072\n",
      "      BatchNorm2d-43          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-44          [-1, 512, 14, 14]               0\n",
      "DepthwiseSeparableConv2d-45          [-1, 512, 14, 14]               0\n",
      "           Conv2d-46          [-1, 512, 14, 14]           4,608\n",
      "      BatchNorm2d-47          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-48          [-1, 512, 14, 14]               0\n",
      "           Conv2d-49          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-50          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-51          [-1, 512, 14, 14]               0\n",
      "DepthwiseSeparableConv2d-52          [-1, 512, 14, 14]               0\n",
      "           Conv2d-53          [-1, 512, 14, 14]           4,608\n",
      "      BatchNorm2d-54          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-55          [-1, 512, 14, 14]               0\n",
      "           Conv2d-56          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-57          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-58          [-1, 512, 14, 14]               0\n",
      "DepthwiseSeparableConv2d-59          [-1, 512, 14, 14]               0\n",
      "           Conv2d-60          [-1, 512, 14, 14]           4,608\n",
      "      BatchNorm2d-61          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-62          [-1, 512, 14, 14]               0\n",
      "           Conv2d-63          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-64          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-65          [-1, 512, 14, 14]               0\n",
      "DepthwiseSeparableConv2d-66          [-1, 512, 14, 14]               0\n",
      "           Conv2d-67          [-1, 512, 14, 14]           4,608\n",
      "      BatchNorm2d-68          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-69          [-1, 512, 14, 14]               0\n",
      "           Conv2d-70          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-71          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-72          [-1, 512, 14, 14]               0\n",
      "DepthwiseSeparableConv2d-73          [-1, 512, 14, 14]               0\n",
      "           Conv2d-74          [-1, 512, 14, 14]           4,608\n",
      "      BatchNorm2d-75          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-76          [-1, 512, 14, 14]               0\n",
      "           Conv2d-77          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-78          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-79          [-1, 512, 14, 14]               0\n",
      "DepthwiseSeparableConv2d-80          [-1, 512, 14, 14]               0\n",
      "           Conv2d-81            [-1, 512, 7, 7]           4,608\n",
      "      BatchNorm2d-82            [-1, 512, 7, 7]           1,024\n",
      "             ReLU-83            [-1, 512, 7, 7]               0\n",
      "           Conv2d-84           [-1, 1024, 7, 7]         524,288\n",
      "      BatchNorm2d-85           [-1, 1024, 7, 7]           2,048\n",
      "             ReLU-86           [-1, 1024, 7, 7]               0\n",
      "DepthwiseSeparableConv2d-87           [-1, 1024, 7, 7]               0\n",
      "           Conv2d-88           [-1, 1024, 7, 7]           9,216\n",
      "      BatchNorm2d-89           [-1, 1024, 7, 7]           2,048\n",
      "             ReLU-90           [-1, 1024, 7, 7]               0\n",
      "           Conv2d-91           [-1, 1024, 7, 7]       1,048,576\n",
      "      BatchNorm2d-92           [-1, 1024, 7, 7]           2,048\n",
      "             ReLU-93           [-1, 1024, 7, 7]               0\n",
      "DepthwiseSeparableConv2d-94           [-1, 1024, 7, 7]               0\n",
      "AdaptiveAvgPool2d-95           [-1, 1024, 1, 1]               0\n",
      "           Linear-96                   [-1, 45]          46,125\n",
      "================================================================\n",
      "Total params: 3,253,101\n",
      "Trainable params: 3,253,101\n",
      "Non-trainable params: 0\n",
      "----------------------------------------------------------------\n",
      "Input size (MB): 0.57\n",
      "Forward/backward pass size (MB): 136.10\n",
      "Params size (MB): 12.41\n",
      "Estimated Total Size (MB): 149.08\n",
      "----------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "model = MobileNet(num_classes=len(classes), width_multiplier=1.0, init_weights=True)\n",
    "summary(model, input_size=(3, 224, 224), device=\"cpu\")\n",
    "model = model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 0 | 300\n",
      "\tTrain Loss : 3.6479, Train Acc : 0.0641\n",
      "\tValid Loss : 3.4345, Valid Acc : 0.1076\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 10 | 300\n",
      "\tTrain Loss : 2.6363, Train Acc : 0.2899\n",
      "\tValid Loss : 2.5038, Valid Acc : 0.3299\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 20 | 300\n",
      "\tTrain Loss : 1.9937, Train Acc : 0.4522\n",
      "\tValid Loss : 1.8426, Valid Acc : 0.4913\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 30 | 300\n",
      "\tTrain Loss : 1.5735, Train Acc : 0.5589\n",
      "\tValid Loss : 1.4035, Valid Acc : 0.5850\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 40 | 300\n",
      "\tTrain Loss : 1.2707, Train Acc : 0.6388\n",
      "\tValid Loss : 1.2177, Valid Acc : 0.6497\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 50 | 300\n",
      "\tTrain Loss : 1.0567, Train Acc : 0.6997\n",
      "\tValid Loss : 1.1414, Valid Acc : 0.6969\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 60 | 300\n",
      "\tTrain Loss : 0.8817, Train Acc : 0.7443\n",
      "\tValid Loss : 1.0390, Valid Acc : 0.7093\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 70 | 300\n",
      "\tTrain Loss : 0.7753, Train Acc : 0.7806\n",
      "\tValid Loss : 1.0242, Valid Acc : 0.7246\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Early stopping triggered after epoch 71 due to no improvement in validation loss\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r"
     ]
    }
   ],
   "source": [
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate, weight_decay=weight_decay)\n",
    "scheduler = LambdaLR(optimizer, lr_lambda)\n",
    "\n",
    "train_costs, train_accs = [], []\n",
    "valid_costs, valid_accs = [], []\n",
    "early_stop_counter = 0\n",
    "early_stop_patience = 10\n",
    "best_valid_cost = float('inf')\n",
    "for epoch in range(epochs):\n",
    "    train_cost, train_acc = train(model, train_dataloder, criterion, optimizer, device)\n",
    "    valid_cost, valid_acc = valid(model, valid_dataloader, criterion, device)\n",
    "\n",
    "    if epoch % 10 == 0 or epoch == epochs:\n",
    "        print(f\"Epoch : {epoch} | {epochs}\")\n",
    "        print(f\"\\tTrain Loss : {train_cost:.4f}, Train Acc : {train_acc:.4f}\")\n",
    "        print(f\"\\tValid Loss : {valid_cost:.4f}, Valid Acc : {valid_acc:.4f}\")\n",
    "\n",
    "    train_costs.append(train_cost)\n",
    "    valid_costs.append(valid_cost)\n",
    "    train_accs.append(train_acc)\n",
    "    valid_accs.append(valid_acc)\n",
    "\n",
    "    scheduler.step()\n",
    "\n",
    "    ## Early Stopping\n",
    "    if valid_cost < best_valid_cost:\n",
    "        best_valid_cost = valid_cost\n",
    "        early_stop_counter = 0\n",
    "    else:\n",
    "        early_stop_counter += 1\n",
    "    \n",
    "    if early_stop_counter >= early_stop_patience:\n",
    "        print(f\"Early stopping triggered after epoch {epoch} due to no improvement in validation loss\")\n",
    "        break  # Break out of the loop to stop training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 5))\n",
    "plt.plot(train_costs, label='Training cost')\n",
    "plt.plot(valid_costs, label='Validation cost')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Cost')\n",
    "plt.legend()\n",
    "plt.title('MobileNet Cost')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 5))\n",
    "plt.plot(train_accs, label='Training accs')\n",
    "plt.plot(valid_accs, label='Validation accs')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.title('MobileNet Accuracy')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id=\"6\"></a>\n",
    "## 6.Conv MobileNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ConvBNReLU(nn.Module):\n",
    "    def __init__(self, in_channels, out_channels, stride=1):\n",
    "        super(ConvBNReLU, self).__init__()\n",
    "        self.conv1 = nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=stride, padding=1, bias=False)\n",
    "        self.bn1 = nn.BatchNorm2d(in_channels)\n",
    "\n",
    "        self.conv2 = nn.Conv2d(in_channels=in_channels, out_channels=out_channels, kernel_size=1, stride=1, padding=0, bias=False)\n",
    "        self.bn2 = nn.BatchNorm2d(out_channels)\n",
    "        self.relu = nn.ReLU(inplace=True)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.conv1(x)\n",
    "        x = self.bn1(x)\n",
    "        x = self.relu(x)\n",
    "\n",
    "        x = self.conv2(x)\n",
    "        x = self.bn2(x)\n",
    "        x = self.relu(x)\n",
    "        \n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ConvMobileNet(nn.Module):\n",
    "    def __init__(self, num_classes=1000, width_multiplier=1.0, init_weights=False):\n",
    "        super().__init__()\n",
    "        self.feature_extractor = nn.Sequential(\n",
    "            nn.Conv2d(in_channels=3, out_channels=32, kernel_size=3, stride=2, padding=1, bias=False),\n",
    "            nn.BatchNorm2d(32),\n",
    "            nn.ReLU(inplace=True),\n",
    "\n",
    "            ConvBNReLU(in_channels=32, out_channels=64, stride=1),\n",
    "            ConvBNReLU(in_channels=64, out_channels=128, stride=2),\n",
    "            ConvBNReLU(in_channels=128, out_channels=128, stride=1),\n",
    "            ConvBNReLU(in_channels=128, out_channels=256, stride=2),\n",
    "            ConvBNReLU(in_channels=256, out_channels=256, stride=1),\n",
    "            ConvBNReLU(in_channels=256, out_channels=512, stride=2),\n",
    "            ConvBNReLU(in_channels=512, out_channels=512, stride=1),\n",
    "            ConvBNReLU(in_channels=512, out_channels=512, stride=1),\n",
    "            ConvBNReLU(in_channels=512, out_channels=512, stride=1),\n",
    "            ConvBNReLU(in_channels=512, out_channels=512, stride=1),\n",
    "            ConvBNReLU(in_channels=512, out_channels=512, stride=1),\n",
    "            ConvBNReLU(in_channels=512, out_channels=1024, stride=2),\n",
    "            ConvBNReLU(in_channels=1024, out_channels=1024, stride=1),\n",
    "        )\n",
    "\n",
    "        self.avg_pool = nn.AdaptiveAvgPool2d((1, 1))\n",
    "        self.classifier = nn.Linear(1024, num_classes)\n",
    "\n",
    "        if init_weights:\n",
    "            for m in self.modules():\n",
    "                if isinstance(m, nn.Conv2d):\n",
    "                    nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')\n",
    "                    if m.bias is not None:\n",
    "                        nn.init.constant_(m.bias, 0)\n",
    "                elif isinstance(m, nn.BatchNorm2d):\n",
    "                    nn.init.constant_(m.weight, 1)\n",
    "                    nn.init.constant_(m.bias, 0)\n",
    "                elif isinstance(m, nn.Linear):\n",
    "                    nn.init.normal_(m.weight, 0, 0.01)\n",
    "                    nn.init.constant_(m.bias, 0)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.feature_extractor(x)\n",
    "        x = self.avg_pool(x)\n",
    "        x = x.view(-1, 1024)\n",
    "        x = self.classifier(x)\n",
    "\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "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",
      "              ReLU-3         [-1, 32, 112, 112]               0\n",
      "            Conv2d-4         [-1, 32, 112, 112]           9,216\n",
      "       BatchNorm2d-5         [-1, 32, 112, 112]              64\n",
      "              ReLU-6         [-1, 32, 112, 112]               0\n",
      "            Conv2d-7         [-1, 64, 112, 112]           2,048\n",
      "       BatchNorm2d-8         [-1, 64, 112, 112]             128\n",
      "              ReLU-9         [-1, 64, 112, 112]               0\n",
      "       ConvBNReLU-10         [-1, 64, 112, 112]               0\n",
      "           Conv2d-11           [-1, 64, 56, 56]          36,864\n",
      "      BatchNorm2d-12           [-1, 64, 56, 56]             128\n",
      "             ReLU-13           [-1, 64, 56, 56]               0\n",
      "           Conv2d-14          [-1, 128, 56, 56]           8,192\n",
      "      BatchNorm2d-15          [-1, 128, 56, 56]             256\n",
      "             ReLU-16          [-1, 128, 56, 56]               0\n",
      "       ConvBNReLU-17          [-1, 128, 56, 56]               0\n",
      "           Conv2d-18          [-1, 128, 56, 56]         147,456\n",
      "      BatchNorm2d-19          [-1, 128, 56, 56]             256\n",
      "             ReLU-20          [-1, 128, 56, 56]               0\n",
      "           Conv2d-21          [-1, 128, 56, 56]          16,384\n",
      "      BatchNorm2d-22          [-1, 128, 56, 56]             256\n",
      "             ReLU-23          [-1, 128, 56, 56]               0\n",
      "       ConvBNReLU-24          [-1, 128, 56, 56]               0\n",
      "           Conv2d-25          [-1, 128, 28, 28]         147,456\n",
      "      BatchNorm2d-26          [-1, 128, 28, 28]             256\n",
      "             ReLU-27          [-1, 128, 28, 28]               0\n",
      "           Conv2d-28          [-1, 256, 28, 28]          32,768\n",
      "      BatchNorm2d-29          [-1, 256, 28, 28]             512\n",
      "             ReLU-30          [-1, 256, 28, 28]               0\n",
      "       ConvBNReLU-31          [-1, 256, 28, 28]               0\n",
      "           Conv2d-32          [-1, 256, 28, 28]         589,824\n",
      "      BatchNorm2d-33          [-1, 256, 28, 28]             512\n",
      "             ReLU-34          [-1, 256, 28, 28]               0\n",
      "           Conv2d-35          [-1, 256, 28, 28]          65,536\n",
      "      BatchNorm2d-36          [-1, 256, 28, 28]             512\n",
      "             ReLU-37          [-1, 256, 28, 28]               0\n",
      "       ConvBNReLU-38          [-1, 256, 28, 28]               0\n",
      "           Conv2d-39          [-1, 256, 14, 14]         589,824\n",
      "      BatchNorm2d-40          [-1, 256, 14, 14]             512\n",
      "             ReLU-41          [-1, 256, 14, 14]               0\n",
      "           Conv2d-42          [-1, 512, 14, 14]         131,072\n",
      "      BatchNorm2d-43          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-44          [-1, 512, 14, 14]               0\n",
      "       ConvBNReLU-45          [-1, 512, 14, 14]               0\n",
      "           Conv2d-46          [-1, 512, 14, 14]       2,359,296\n",
      "      BatchNorm2d-47          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-48          [-1, 512, 14, 14]               0\n",
      "           Conv2d-49          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-50          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-51          [-1, 512, 14, 14]               0\n",
      "       ConvBNReLU-52          [-1, 512, 14, 14]               0\n",
      "           Conv2d-53          [-1, 512, 14, 14]       2,359,296\n",
      "      BatchNorm2d-54          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-55          [-1, 512, 14, 14]               0\n",
      "           Conv2d-56          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-57          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-58          [-1, 512, 14, 14]               0\n",
      "       ConvBNReLU-59          [-1, 512, 14, 14]               0\n",
      "           Conv2d-60          [-1, 512, 14, 14]       2,359,296\n",
      "      BatchNorm2d-61          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-62          [-1, 512, 14, 14]               0\n",
      "           Conv2d-63          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-64          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-65          [-1, 512, 14, 14]               0\n",
      "       ConvBNReLU-66          [-1, 512, 14, 14]               0\n",
      "           Conv2d-67          [-1, 512, 14, 14]       2,359,296\n",
      "      BatchNorm2d-68          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-69          [-1, 512, 14, 14]               0\n",
      "           Conv2d-70          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-71          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-72          [-1, 512, 14, 14]               0\n",
      "       ConvBNReLU-73          [-1, 512, 14, 14]               0\n",
      "           Conv2d-74          [-1, 512, 14, 14]       2,359,296\n",
      "      BatchNorm2d-75          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-76          [-1, 512, 14, 14]               0\n",
      "           Conv2d-77          [-1, 512, 14, 14]         262,144\n",
      "      BatchNorm2d-78          [-1, 512, 14, 14]           1,024\n",
      "             ReLU-79          [-1, 512, 14, 14]               0\n",
      "       ConvBNReLU-80          [-1, 512, 14, 14]               0\n",
      "           Conv2d-81            [-1, 512, 7, 7]       2,359,296\n",
      "      BatchNorm2d-82            [-1, 512, 7, 7]           1,024\n",
      "             ReLU-83            [-1, 512, 7, 7]               0\n",
      "           Conv2d-84           [-1, 1024, 7, 7]         524,288\n",
      "      BatchNorm2d-85           [-1, 1024, 7, 7]           2,048\n",
      "             ReLU-86           [-1, 1024, 7, 7]               0\n",
      "       ConvBNReLU-87           [-1, 1024, 7, 7]               0\n",
      "           Conv2d-88           [-1, 1024, 7, 7]       9,437,184\n",
      "      BatchNorm2d-89           [-1, 1024, 7, 7]           2,048\n",
      "             ReLU-90           [-1, 1024, 7, 7]               0\n",
      "           Conv2d-91           [-1, 1024, 7, 7]       1,048,576\n",
      "      BatchNorm2d-92           [-1, 1024, 7, 7]           2,048\n",
      "             ReLU-93           [-1, 1024, 7, 7]               0\n",
      "       ConvBNReLU-94           [-1, 1024, 7, 7]               0\n",
      "AdaptiveAvgPool2d-95           [-1, 1024, 1, 1]               0\n",
      "           Linear-96                   [-1, 45]          46,125\n",
      "================================================================\n",
      "Total params: 28,322,061\n",
      "Trainable params: 28,322,061\n",
      "Non-trainable params: 0\n",
      "----------------------------------------------------------------\n",
      "Input size (MB): 0.57\n",
      "Forward/backward pass size (MB): 136.10\n",
      "Params size (MB): 108.04\n",
      "Estimated Total Size (MB): 244.71\n",
      "----------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "model = ConvMobileNet(num_classes=len(classes), init_weights=True)\n",
    "summary(model, input_size=(3, 224, 224), device=\"cpu\")\n",
    "model = model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 0 | 300\n",
      "\tTrain Loss : 3.7001, Train Acc : 0.0482\n",
      "\tValid Loss : 3.6043, Valid Acc : 0.0647\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 10 | 300\n",
      "\tTrain Loss : 3.4330, Train Acc : 0.0955\n",
      "\tValid Loss : 3.3266, Valid Acc : 0.1156\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 20 | 300\n",
      "\tTrain Loss : 3.3112, Train Acc : 0.1269\n",
      "\tValid Loss : 3.4209, Valid Acc : 0.1097\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 30 | 300\n",
      "\tTrain Loss : 3.1460, Train Acc : 0.1670\n",
      "\tValid Loss : 3.0433, Valid Acc : 0.1846\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 40 | 300\n",
      "\tTrain Loss : 2.9436, Train Acc : 0.2112\n",
      "\tValid Loss : 2.7929, Valid Acc : 0.2522\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 50 | 300\n",
      "\tTrain Loss : 2.7171, Train Acc : 0.2724\n",
      "\tValid Loss : 2.8432, Valid Acc : 0.2442\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 60 | 300\n",
      "\tTrain Loss : 2.5117, Train Acc : 0.3213\n",
      "\tValid Loss : 2.3612, Valid Acc : 0.3488\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 70 | 300\n",
      "\tTrain Loss : 2.3805, Train Acc : 0.3487\n",
      "\tValid Loss : 2.2332, Valid Acc : 0.3837\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 80 | 300\n",
      "\tTrain Loss : 2.2604, Train Acc : 0.3847\n",
      "\tValid Loss : 2.1653, Valid Acc : 0.4012\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 90 | 300\n",
      "\tTrain Loss : 2.1208, Train Acc : 0.4185\n",
      "\tValid Loss : 2.0299, Valid Acc : 0.4433\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 100 | 300\n",
      "\tTrain Loss : 2.0479, Train Acc : 0.4375\n",
      "\tValid Loss : 1.8594, Valid Acc : 0.4891\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 110 | 300\n",
      "\tTrain Loss : 1.9248, Train Acc : 0.4666\n",
      "\tValid Loss : 1.8024, Valid Acc : 0.5007\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 120 | 300\n",
      "\tTrain Loss : 1.8400, Train Acc : 0.4913\n",
      "\tValid Loss : 1.7736, Valid Acc : 0.5087\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 130 | 300\n",
      "\tTrain Loss : 1.7827, Train Acc : 0.5046\n",
      "\tValid Loss : 1.6700, Valid Acc : 0.5436\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 140 | 300\n",
      "\tTrain Loss : 1.7199, Train Acc : 0.5186\n",
      "\tValid Loss : 1.6528, Valid Acc : 0.5487\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 150 | 300\n",
      "\tTrain Loss : 1.6590, Train Acc : 0.5371\n",
      "\tValid Loss : 1.5634, Valid Acc : 0.5705\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 160 | 300\n",
      "\tTrain Loss : 1.6080, Train Acc : 0.5485\n",
      "\tValid Loss : 1.5745, Valid Acc : 0.5610\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 170 | 300\n",
      "\tTrain Loss : 1.5792, Train Acc : 0.5573\n",
      "\tValid Loss : 1.5297, Valid Acc : 0.5698\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        \r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch : 180 | 300\n",
      "\tTrain Loss : 1.5491, Train Acc : 0.5666\n",
      "\tValid Loss : 1.5345, Valid Acc : 0.5676\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "                                                        "
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Early stopping triggered after epoch 188 due to no improvement in validation loss\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\r"
     ]
    }
   ],
   "source": [
    "criterion = nn.CrossEntropyLoss()\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate, weight_decay=weight_decay)\n",
    "scheduler = LambdaLR(optimizer, lr_lambda)\n",
    "\n",
    "train_costs, train_accs = [], []\n",
    "valid_costs, valid_accs = [], []\n",
    "early_stop_counter = 0\n",
    "early_stop_patience = 10\n",
    "best_valid_cost = float('inf')\n",
    "for epoch in range(epochs):\n",
    "    train_cost, train_acc = train(model, train_dataloder, criterion, optimizer, device)\n",
    "    valid_cost, valid_acc = valid(model, valid_dataloader, criterion, device)\n",
    "\n",
    "    if epoch % 10 == 0 or epoch == epochs:\n",
    "        print(f\"Epoch : {epoch} | {epochs}\")\n",
    "        print(f\"\\tTrain Loss : {train_cost:.4f}, Train Acc : {train_acc:.4f}\")\n",
    "        print(f\"\\tValid Loss : {valid_cost:.4f}, Valid Acc : {valid_acc:.4f}\")\n",
    "\n",
    "    train_costs.append(train_cost)\n",
    "    valid_costs.append(valid_cost)\n",
    "    train_accs.append(train_acc)\n",
    "    valid_accs.append(valid_acc)\n",
    "\n",
    "    scheduler.step()\n",
    "\n",
    "    ## Early Stopping\n",
    "    if valid_cost < best_valid_cost:\n",
    "        best_valid_cost = valid_cost\n",
    "        early_stop_counter = 0\n",
    "    else:\n",
    "        early_stop_counter += 1\n",
    "    \n",
    "    if early_stop_counter >= early_stop_patience:\n",
    "        print(f\"Early stopping triggered after epoch {epoch} due to no improvement in validation loss\")\n",
    "        break  # Break out of the loop to stop training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 5))\n",
    "plt.plot(train_costs, label='Training cost')\n",
    "plt.plot(valid_costs, label='Validation cost')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Cost')\n",
    "plt.legend()\n",
    "plt.title('ConvMobileNet Cost')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10, 5))\n",
    "plt.plot(train_accs, label='Training accs')\n",
    "plt.plot(valid_accs, label='Validation accs')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.title('ConvMobileNet Accuracy')\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "openmmlab",
   "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}