dummy_demo.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "from torch.utils.data import DataLoader\n",
    "import numpy as np\n",
    "import time\n",
    "import os\n",
    "import sys\n",
    "from matplotlib import pyplot as plt\n",
    "from Datasets.MSASLDataset import MSASLDataset_3DCNN\n",
    "from Datasets.MSASL5Channel import MSAL5ChannelDataset\n",
    "from Models.Conv3d import Conv3d\n",
    "from utils.train import train\n",
    "from torchvision import transforms\n",
    "import os\n",
    "from PIL import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2813\n"
     ]
    }
   ],
   "source": [
    "###### loading full dataset\n",
    "root_dir = './videos/MS-ASL100'\n",
    "classes = os.listdir(root_dir)\n",
    "total = 0\n",
    "for c in classes:\n",
    "    total += len(os.listdir(os.path.join(root_dir, c)))\n",
    "print(total)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "classes_5 = ['spring', 'cousin', 'water', 'friend', 'like']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 32/32 [00:03<00:00, 10.37it/s]\n",
      "100%|██████████| 37/37 [00:03<00:00, 11.42it/s]\n",
      "100%|██████████| 34/34 [00:03<00:00, 10.68it/s]\n",
      "100%|██████████| 35/35 [00:03<00:00, 10.57it/s]\n",
      "100%|██████████| 34/34 [00:03<00:00, 10.15it/s]\n",
      "100%|██████████| 5/5 [00:16<00:00,  3.24s/it]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "172"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Loading the data\n",
    "root_dir = './videos/MS-ASL100'\n",
    "classes = classes_5\n",
    "data = MSASLDataset_3DCNN(root_dir, classes, frames=10)\n",
    "len(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Save the loaded data if needed. Uncomment the code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# data = [dataset[i] for i in range(len(dataset))]\n",
    "# print(len(data))\n",
    "# torch.save(data, './StoredData/conv3d10classData.pt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset = [(d[0]/255, d[1]) for d in data]\n",
    "len(dataset)\n",
    "del data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### visualize data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "torch.Size([10, 128, 128, 3])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "example = dataset[1][0].permute(3,1,2,0)\n",
    "print(dataset[1][1])\n",
    "plt.figure(figsize=(10,10))\n",
    "print(example.shape)\n",
    "for i in range(example.shape[0]):\n",
    "    plt.subplot(1, example.shape[0], i+1)\n",
    "    plt.title(f'frame {i}')\n",
    "    plt.imshow(example[i])\n",
    "    plt.axis('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Augment data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def augment_dataset(dataset, n, num_augmentations=10):\n",
    "    for i in range(n):\n",
    "        print(\"augmenting\", i, \"of\", n)\n",
    "        img = dataset[i][0].permute(3,1,2,0).clone()\n",
    "        label = dataset[i][1]\n",
    "\n",
    "        for j in range(num_augmentations):\n",
    "            temp_img = img.clone()\n",
    "            choices = [\"rotation\", \"flip\", \"translation\", \"noise\"]\n",
    "            choice = np.random.choice(choices)\n",
    "\n",
    "            if choice == \"rotation\":\n",
    "                angle = np.random.randint(-30, 30)\n",
    "                for i in range(10):\n",
    "                    temp_rgb = temp_img[i].permute(2, 0, 1)\n",
    "                    temp_rgb = transforms.functional.to_pil_image(temp_rgb)\n",
    "                    temp_rgb = temp_rgb.rotate(angle, resample=Image.BICUBIC)\n",
    "                    temp_rgb = transforms.ToTensor()(temp_rgb)\n",
    "                    temp_img[i] = temp_rgb.permute(1, 2, 0)\n",
    "                \n",
    "            if choice == \"flip\":\n",
    "                \n",
    "                    for i in range(10):\n",
    "                        temp_rgb = temp_img[i].permute(2, 0, 1)\n",
    "                        temp_rgb = transforms.functional.to_pil_image(temp_rgb)\n",
    "\n",
    "                        temp_rgb = temp_rgb.transpose(Image.FLIP_LEFT_RIGHT)\n",
    "                        temp_rgb = transforms.ToTensor()(temp_rgb)\n",
    "                        temp_img[i] = temp_rgb.permute(1, 2, 0)\n",
    "\n",
    "            if choice == \"translation\":\n",
    "                x = np.random.randint(-10, 10)\n",
    "                y = np.random.randint(-10, 10)\n",
    "                for i in range(10):\n",
    "                    temp_rgb = temp_img[i].permute(2, 0, 1)\n",
    "                    temp_rgb = transforms.functional.to_pil_image(temp_rgb)\n",
    "                    temp_rgb = transforms.functional.affine(temp_rgb, angle=0, translate=[x,y], scale=1, shear=0)\n",
    "                    temp_rgb = transforms.ToTensor()(temp_rgb)\n",
    "                    temp_img[i] = temp_rgb.permute(1, 2, 0)\n",
    "\n",
    "            if choice == \"noise\":\n",
    "                for i in range(10):\n",
    "                    random_noise = torch.normal(0, 0.1, size=temp_img[i].shape)\n",
    "                    temp_img[i] = temp_img[i] + random_noise\n",
    "            \n",
    "\n",
    "            dataset.append((temp_img.permute(3,1,2,0), label))\n",
    "    return dataset\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "augmenting 0 of 172\n",
      "augmenting 1 of 172\n",
      "augmenting 2 of 172\n",
      "augmenting 3 of 172\n",
      "augmenting 4 of 172\n",
      "augmenting 5 of 172\n",
      "augmenting 6 of 172\n",
      "augmenting 7 of 172\n",
      "augmenting 8 of 172\n",
      "augmenting 9 of 172\n",
      "augmenting 10 of 172\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\randy\\AppData\\Local\\Temp\\ipykernel_29004\\2634722734.py:17: DeprecationWarning: BICUBIC is deprecated and will be removed in Pillow 10 (2023-07-01). Use Resampling.BICUBIC instead.\n",
      "  temp_rgb = temp_rgb.rotate(angle, resample=Image.BICUBIC)\n",
      "C:\\Users\\randy\\AppData\\Local\\Temp\\ipykernel_29004\\2634722734.py:27: DeprecationWarning: FLIP_LEFT_RIGHT is deprecated and will be removed in Pillow 10 (2023-07-01). Use Transpose.FLIP_LEFT_RIGHT instead.\n",
      "  temp_rgb = temp_rgb.transpose(Image.FLIP_LEFT_RIGHT)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "augmenting 11 of 172\n",
      "augmenting 12 of 172\n",
      "augmenting 13 of 172\n",
      "augmenting 14 of 172\n",
      "augmenting 15 of 172\n",
      "augmenting 16 of 172\n",
      "augmenting 17 of 172\n",
      "augmenting 18 of 172\n",
      "augmenting 19 of 172\n",
      "augmenting 20 of 172\n",
      "augmenting 21 of 172\n",
      "augmenting 22 of 172\n",
      "augmenting 23 of 172\n",
      "augmenting 24 of 172\n",
      "augmenting 25 of 172\n",
      "augmenting 26 of 172\n",
      "augmenting 27 of 172\n",
      "augmenting 28 of 172\n",
      "augmenting 29 of 172\n",
      "augmenting 30 of 172\n",
      "augmenting 31 of 172\n",
      "augmenting 32 of 172\n",
      "augmenting 33 of 172\n",
      "augmenting 34 of 172\n",
      "augmenting 35 of 172\n",
      "augmenting 36 of 172\n",
      "augmenting 37 of 172\n",
      "augmenting 38 of 172\n",
      "augmenting 39 of 172\n",
      "augmenting 40 of 172\n",
      "augmenting 41 of 172\n",
      "augmenting 42 of 172\n",
      "augmenting 43 of 172\n",
      "augmenting 44 of 172\n",
      "augmenting 45 of 172\n",
      "augmenting 46 of 172\n",
      "augmenting 47 of 172\n",
      "augmenting 48 of 172\n",
      "augmenting 49 of 172\n",
      "augmenting 50 of 172\n",
      "augmenting 51 of 172\n",
      "augmenting 52 of 172\n",
      "augmenting 53 of 172\n",
      "augmenting 54 of 172\n",
      "augmenting 55 of 172\n",
      "augmenting 56 of 172\n",
      "augmenting 57 of 172\n",
      "augmenting 58 of 172\n",
      "augmenting 59 of 172\n",
      "augmenting 60 of 172\n",
      "augmenting 61 of 172\n",
      "augmenting 62 of 172\n",
      "augmenting 63 of 172\n",
      "augmenting 64 of 172\n",
      "augmenting 65 of 172\n",
      "augmenting 66 of 172\n",
      "augmenting 67 of 172\n",
      "augmenting 68 of 172\n",
      "augmenting 69 of 172\n",
      "augmenting 70 of 172\n",
      "augmenting 71 of 172\n",
      "augmenting 72 of 172\n",
      "augmenting 73 of 172\n",
      "augmenting 74 of 172\n",
      "augmenting 75 of 172\n",
      "augmenting 76 of 172\n",
      "augmenting 77 of 172\n",
      "augmenting 78 of 172\n",
      "augmenting 79 of 172\n",
      "augmenting 80 of 172\n",
      "augmenting 81 of 172\n",
      "augmenting 82 of 172\n",
      "augmenting 83 of 172\n",
      "augmenting 84 of 172\n",
      "augmenting 85 of 172\n",
      "augmenting 86 of 172\n",
      "augmenting 87 of 172\n",
      "augmenting 88 of 172\n",
      "augmenting 89 of 172\n",
      "augmenting 90 of 172\n",
      "augmenting 91 of 172\n",
      "augmenting 92 of 172\n",
      "augmenting 93 of 172\n",
      "augmenting 94 of 172\n",
      "augmenting 95 of 172\n",
      "augmenting 96 of 172\n",
      "augmenting 97 of 172\n",
      "augmenting 98 of 172\n",
      "augmenting 99 of 172\n",
      "augmenting 100 of 172\n",
      "augmenting 101 of 172\n",
      "augmenting 102 of 172\n",
      "augmenting 103 of 172\n",
      "augmenting 104 of 172\n",
      "augmenting 105 of 172\n",
      "augmenting 106 of 172\n",
      "augmenting 107 of 172\n",
      "augmenting 108 of 172\n",
      "augmenting 109 of 172\n",
      "augmenting 110 of 172\n",
      "augmenting 111 of 172\n",
      "augmenting 112 of 172\n",
      "augmenting 113 of 172\n",
      "augmenting 114 of 172\n",
      "augmenting 115 of 172\n",
      "augmenting 116 of 172\n",
      "augmenting 117 of 172\n",
      "augmenting 118 of 172\n",
      "augmenting 119 of 172\n",
      "augmenting 120 of 172\n",
      "augmenting 121 of 172\n",
      "augmenting 122 of 172\n",
      "augmenting 123 of 172\n",
      "augmenting 124 of 172\n",
      "augmenting 125 of 172\n",
      "augmenting 126 of 172\n",
      "augmenting 127 of 172\n",
      "augmenting 128 of 172\n",
      "augmenting 129 of 172\n",
      "augmenting 130 of 172\n",
      "augmenting 131 of 172\n",
      "augmenting 132 of 172\n",
      "augmenting 133 of 172\n",
      "augmenting 134 of 172\n",
      "augmenting 135 of 172\n",
      "augmenting 136 of 172\n",
      "augmenting 137 of 172\n",
      "augmenting 138 of 172\n",
      "augmenting 139 of 172\n",
      "augmenting 140 of 172\n",
      "augmenting 141 of 172\n",
      "augmenting 142 of 172\n",
      "augmenting 143 of 172\n",
      "augmenting 144 of 172\n",
      "augmenting 145 of 172\n",
      "augmenting 146 of 172\n",
      "augmenting 147 of 172\n",
      "augmenting 148 of 172\n",
      "augmenting 149 of 172\n",
      "augmenting 150 of 172\n",
      "augmenting 151 of 172\n",
      "augmenting 152 of 172\n",
      "augmenting 153 of 172\n",
      "augmenting 154 of 172\n",
      "augmenting 155 of 172\n",
      "augmenting 156 of 172\n",
      "augmenting 157 of 172\n",
      "augmenting 158 of 172\n",
      "augmenting 159 of 172\n",
      "augmenting 160 of 172\n",
      "augmenting 161 of 172\n",
      "augmenting 162 of 172\n",
      "augmenting 163 of 172\n",
      "augmenting 164 of 172\n",
      "augmenting 165 of 172\n",
      "augmenting 166 of 172\n",
      "augmenting 167 of 172\n",
      "augmenting 168 of 172\n",
      "augmenting 169 of 172\n",
      "augmenting 170 of 172\n",
      "augmenting 171 of 172\n",
      "688\n"
     ]
    }
   ],
   "source": [
    "n = len(dataset)\n",
    "dataset = augment_dataset(dataset, n, num_augmentations=3)\n",
    "print(len(dataset))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### save the augmented data if needed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# torch.save(dataset, './StoredData/conv3d10classDataAugmented.pt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([3, 128, 128, 10])\n",
      "4\n",
      "torch.Size([10, 128, 128, 3])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 10 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(dataset[500][0].shape)\n",
    "aug_exmp = dataset[500][0].permute(3,1,2,0)\n",
    "print(dataset[-1][1])\n",
    "plt.figure(figsize=(10,10))\n",
    "print(aug_exmp.shape)\n",
    "for i in range(aug_exmp.shape[0]):\n",
    "    plt.subplot(1, aug_exmp.shape[0], i+1)\n",
    "    plt.title(f'frame {i}')\n",
    "    plt.imshow(aug_exmp[i])\n",
    "    plt.axis('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train size: 550\n",
      "val size: 68\n",
      "test size: 70\n"
     ]
    }
   ],
   "source": [
    "#split the data into training, validation and test sets\n",
    "train_size = int(0.8 * len(dataset))\n",
    "val_size = int(0.1 * len(dataset))\n",
    "test_size = len(dataset) - train_size - val_size\n",
    "\n",
    "train_dataset, val_dataset, test_dataset = torch.utils.data.random_split(dataset, [train_size, val_size, test_size])\n",
    "print(\"train size:\", len(train_dataset))\n",
    "print(\"val size:\", len(val_dataset))\n",
    "print(\"test size:\", len(test_dataset))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_loader = DataLoader(train_dataset, batch_size=64, shuffle=True)\n",
    "val_loader = DataLoader(test_dataset, batch_size=64, shuffle=True)\n",
    "test_loader = DataLoader(test_dataset, batch_size=64, shuffle=True)\n",
    "\n",
    "del dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([64, 3, 128, 128, 10])\n",
      "tensor([4, 3, 0, 0, 3, 2, 1, 0, 0, 4, 4, 1, 4, 2, 3, 2, 0, 0, 4, 3, 3, 4, 3, 2,\n",
      "        3, 0, 3, 3, 4, 4, 0, 1, 4, 4, 0, 3, 1, 0, 1, 2, 2, 0, 4, 2, 4, 3, 3, 3,\n",
      "        3, 1, 3, 2, 3, 3, 1, 0, 3, 1, 0, 4, 4, 0, 3, 0])\n"
     ]
    }
   ],
   "source": [
    "for batch_idx, (data, target) in enumerate(train_loader):\n",
    "    print(data.shape)\n",
    "    print(target)\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Conv3d(nn.Module):\n",
    "    def __init__(self,in_channels, num_frames, num_classes=10):\n",
    "        super(Conv3d, self).__init__()\n",
    "        #input shape: (3, 128, 128, num_frames)\n",
    "        self.conv1 = nn.Conv3d(in_channels=in_channels, out_channels=64, kernel_size=(7,7,5))\n",
    "        #output shape: (64, 122, 122, num_frames-4)\n",
    "        self.bn1 = nn.BatchNorm3d(64)\n",
    "        self.pool3D1 = nn.MaxPool3d(kernel_size=(2,2,1))\n",
    "        #output shape: (64, 61, 61, num_frames-4)\n",
    "\n",
    "        self.conv2 = nn.Conv3d(in_channels=64, out_channels=32, kernel_size=(7,7,3))\n",
    "        #output shape: (32, 55, 55, num_frames-4-2)\n",
    "        self.bn2 = nn.BatchNorm3d(32)\n",
    "        self.pool3D2 = nn.MaxPool3d(kernel_size=(2,2,1))\n",
    "        #output shape: (32, 27, 27, num_frames-4-2)\n",
    "\n",
    "        self. conv3 = nn.Conv3d(in_channels=32, out_channels=16, kernel_size=(5,5,4))\n",
    "        #output shape: (16, 23, 23, num_frames-4-2-3)\n",
    "        self.bn3 = nn.BatchNorm3d(16)\n",
    "        self.pool3D3 = nn.MaxPool3d(kernel_size=(2,2,1))\n",
    "        #output shape: (16, 11, 11, num_frames-4-2-3)\n",
    "\n",
    "\n",
    "        self.fc1 = nn.Linear(16*11*11*(num_frames-9), 128)\n",
    "        self.dropout1 = nn.Dropout(0.5)\n",
    "        self.fc2 = nn.Linear(128, num_classes)\n",
    "\n",
    "\n",
    "        self.relu = nn.ReLU()\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.conv1(x)\n",
    "        x = self.bn1(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.pool3D1(x)\n",
    "\n",
    "        x = self.conv2(x)\n",
    "        x = self.bn2(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.pool3D2(x)\n",
    "\n",
    "        x = self.conv3(x)\n",
    "        x = self.bn3(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.pool3D3(x)\n",
    "\n",
    "\n",
    "        x = x.view(x.size(0), -1)\n",
    "        x = self.fc1(x)\n",
    "        x = self.relu(x)\n",
    "        x = self.dropout1(x)\n",
    "        x = self.fc2(x)\n",
    "        x = F.log_softmax(x, dim=1)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "\n",
    "def train(model, train_loader, test_loader, criterion, optimizer, device, epochs, model_name):\n",
    "    start_time = time.time()\n",
    "    model.train()\n",
    "\n",
    "    train_losses = []\n",
    "    val_losses = []\n",
    "    stop_cnt = 0\n",
    "    best_train_loss = 1000\n",
    "\n",
    "    for epoch in range(epochs):\n",
    "        \n",
    "        trn_corr = 0\n",
    "        total = 0\n",
    "        epoch_train_loss = 0\n",
    "\n",
    "        print(f'Epoch {epoch+1}:')\n",
    "        for batch_idx, (X_train, y_train) in enumerate(train_loader):\n",
    "            batch_idx += 1\n",
    "            X_train, y_train = X_train.to(device), y_train.to(device)\n",
    "            y_pred = model(X_train)\n",
    "\n",
    "            loss = criterion(y_pred, y_train)\n",
    "\n",
    "            predicted = torch.max(y_pred.data, 1)[1]\n",
    "            trn_corr += (predicted == y_train).sum()\n",
    "\n",
    "            optimizer.zero_grad()\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "            \n",
    "            \n",
    "            epoch_train_loss += loss.item()\n",
    "            total += len(y_train)\n",
    "            if batch_idx % 10 == 0:\n",
    "                print(f'\\tBatch {batch_idx}, loss: {loss.item():7.3f} running accuracy: {trn_corr.item()*100/total:7.3f}%')\n",
    "            \n",
    "\n",
    "        avg_epoch_train_loss = epoch_train_loss / len(train_loader)\n",
    "        train_losses.append(avg_epoch_train_loss)     \n",
    "\n",
    "        # Run the testing batches\n",
    "        tst_corr = 0\n",
    "        epoch_test_loss = 0\n",
    "\n",
    "        with torch.no_grad():\n",
    "            for batch_idx, (X_test, y_test) in enumerate(test_loader):\n",
    "                X_test, y_test = X_test.to(device), y_test.to(device)\n",
    "                y_val = model(X_test)\n",
    "                predicted = torch.max(y_val.data, 1)[1]\n",
    "                tst_corr += (predicted == y_test).sum()\n",
    "\n",
    "                loss = criterion(y_val, y_test)\n",
    "                epoch_test_loss += loss.item()\n",
    "                \n",
    "\n",
    "        avg_epoch_test_loss = epoch_test_loss / len(test_loader)\n",
    "        val_losses.append(avg_epoch_test_loss)\n",
    "\n",
    "        print(f'\\n\\tEpoch {epoch+1}, Training Loss: {avg_epoch_train_loss:.3f}, val Loss: {avg_epoch_test_loss:.3f}')\n",
    "        \n",
    "        if (epoch+1) % 10 == 0:\n",
    "            torch.save(model, f'./SavedModels/{model_name}_epoch{epoch+1}.pt')\n",
    "            plt.figure()\n",
    "            plt.plot(train_losses, label='Training loss')\n",
    "            plt.plot(val_losses, label='Validation loss')\n",
    "            plt.legend()\n",
    "            plt.savefig(f'./Figures/{model_name}_losses_epoch{epoch+1}.png')\n",
    "            \n",
    "        if avg_epoch_train_loss < best_train_loss:\n",
    "            best_train_loss = avg_epoch_train_loss\n",
    "            stop_cnt = 0\n",
    "        else:\n",
    "            stop_cnt += 1\n",
    "            if stop_cnt == 3:\n",
    "                print(f'\\nEarly stopping at epoch {epoch+1}')\n",
    "                break\n",
    "        \n",
    "        \n",
    "    print(f'\\nDuration: {time.time() - start_time:.0f} seconds') # print the time elapsed\n",
    "    \n",
    "\n",
    "    return train_losses, val_losses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1:\n",
      "\n",
      "\tEpoch 1, Training Loss: 1.633, val Loss: 1.572\n",
      "Epoch 2:\n",
      "\n",
      "\tEpoch 2, Training Loss: 1.530, val Loss: 1.456\n",
      "Epoch 3:\n",
      "\n",
      "\tEpoch 3, Training Loss: 1.445, val Loss: 1.428\n",
      "Epoch 4:\n",
      "\n",
      "\tEpoch 4, Training Loss: 1.370, val Loss: 1.398\n",
      "Epoch 5:\n",
      "\n",
      "\tEpoch 5, Training Loss: 1.276, val Loss: 1.434\n",
      "\n",
      "Duration: 14 seconds\n"
     ]
    }
   ],
   "source": [
    "model = Conv3d(in_channels=3, num_frames = 10, num_classes=len(classes))\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "# optimizer = optim.Adam(model.parameters(), lr=0.001, weight_decay=0.0001)\n",
    "optimizer = optim.SGD(model.parameters(), lr=0.001, momentum=0.9, weight_decay=0.0001)\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "model.to(device)\n",
    "\n",
    "train_losses, test_losses = train(\n",
    "    model = model, \n",
    "    device = device, \n",
    "    train_loader = train_loader, \n",
    "    test_loader=val_loader,\n",
    "    criterion=criterion,\n",
    "    optimizer=optimizer,\n",
    "    epochs=5,\n",
    "    model_name=\"test\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(train_losses, label='Training loss')\n",
    "plt.plot(test_losses, label='Validation loss')\n",
    "plt.legend(frameon=False)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of the network on the 70 test images: 54.285714285714285%\n"
     ]
    }
   ],
   "source": [
    "model.to(device)\n",
    "model.eval()\n",
    "correct = 0\n",
    "total = 0\n",
    "with torch.no_grad():\n",
    "    for i, (data, target) in enumerate(test_loader):\n",
    "        data, target = data.to(device), target.to(device)\n",
    "        outputs = model(data)\n",
    "        predicted = torch.max(outputs, 1)[1]\n",
    "        total += target.size(0)\n",
    "        correct += (predicted == target).sum().item()\n",
    "\n",
    "print(f'Accuracy of the network on the {total} test images: {100 * correct / total}%')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Evaluate unseen data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 14/14 [00:01<00:00,  8.25it/s]\n",
      "100%|██████████| 11/11 [00:01<00:00,  9.20it/s]\n",
      "100%|██████████| 10/10 [00:00<00:00, 10.59it/s]\n",
      "100%|██████████| 8/8 [00:00<00:00,  9.21it/s]\n",
      "100%|██████████| 17/17 [00:02<00:00,  7.93it/s]\n",
      "100%|██████████| 5/5 [00:06<00:00,  1.37s/it]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of the network on the 60 new test images: 20.0%\n"
     ]
    }
   ],
   "source": [
    "new_data = MSASLDataset_3DCNN(\"./videos/test-MS-ASL100\", classes, frames=10)\n",
    "new_data = [(d[0]/255, d[1]) for d in new_data]\n",
    "new_loader = DataLoader(new_data, batch_size=64, shuffle=True)\n",
    "del new_data\n",
    "model.to(device)\n",
    "model.eval()\n",
    "correct = 0\n",
    "total = 0\n",
    "\n",
    "with torch.no_grad():\n",
    "    for i, (data, target) in enumerate(new_loader):\n",
    "        data, target = data.to(device), target.to(device)\n",
    "        outputs = model(data)\n",
    "        predicted = torch.max(outputs, 1)[1]\n",
    "        total += target.size(0)\n",
    "        correct += (predicted == target).sum().item()\n",
    "\n",
    "print(f'Accuracy of the network on the {total} new test images: {100 * correct / total}%')\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "cvFinal",
   "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": "3.11.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}