pytorch_ipynb/gnn/gnn-basic-edge-1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Deep Learning Models -- A collection of various deep learning architectures, models, and tips for TensorFlow and PyTorch in Jupyter Notebooks.\n",
    "- Author: Sebastian Raschka\n",
    "- GitHub Repository: https://github.com/rasbt/deeplearning-models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sebastian Raschka \n",
      "\n",
      "CPython 3.7.3\n",
      "IPython 7.6.1\n",
      "\n",
      "torch 1.2.0\n"
     ]
    }
   ],
   "source": [
    "%load_ext watermark\n",
    "%watermark -a 'Sebastian Raschka' -v -p torch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- Runs on CPU or GPU (if available)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Basic Graph Neural Network with Edge Prediction on MNIST"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Implementing a very basic graph neural network (GNN) using a subnetwork for edge prediction. \n",
    "\n",
    "Here, the 28x28 image of a digit in MNIST represents the graph, where each pixel (i.e., cell in the grid) represents a particular node. The feature of that node is simply the pixel intensity in range [0, 1]. \n",
    "\n",
    "In the related notebook, [gnn-basic-1.ipyb], the adjacency matrix of the pixels was basically just determined by the neighborhood pixels. Using a Gaussian filter, pixels  were connected based on their Euclidean distance in the grid. In **this notebook**, the edges are predicted via a seperate neural network model \n",
    "\n",
    "\n",
    "```python\n",
    "        self.pred_edge_fc = nn.Sequential(nn.Linear(coord_features, 64),\n",
    "                                          nn.ReLU(),\n",
    "                                          nn.Linear(64, 1),\n",
    "                                          nn.Tanh())\n",
    "```\n",
    "\n",
    "\n",
    "Using the resulting adjacency matrix $A$, we can compute the output of a layer as \n",
    "\n",
    "$$X^{(l+1)}=A X^{(l)} W^{(l)}.$$\n",
    "\n",
    "Here, $A$ is the $N \\times N$ adjacency matrix, and $X$ is the $N \\times C$ feature matrix (a  2D coordinate array, where $N$ is the total number of pixels -- $28 \\times 28 = 784$ in MNIST). $W$ is the weight matrix of shape $N \\times P$, where $P$ would represent the number of classes if we have only a single hidden layer.\n",
    "\n",
    "\n",
    "- Inspired by and based on Boris Knyazev's tutorial at https://medium.com/@BorisAKnyazev/tutorial-on-graph-neural-networks-for-computer-vision-and-beyond-part-1-3d9fada3b80d."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "import numpy as np\n",
    "from scipy.spatial.distance import cdist\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from torchvision import datasets\n",
    "from torchvision import transforms\n",
    "from torch.utils.data import DataLoader\n",
    "from torch.utils.data.dataset import Subset\n",
    "\n",
    "\n",
    "if torch.cuda.is_available():\n",
    "    torch.backends.cudnn.deterministic = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Settings and Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "##########################\n",
    "### SETTINGS\n",
    "##########################\n",
    "\n",
    "# Device\n",
    "DEVICE = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "\n",
    "# Hyperparameters\n",
    "RANDOM_SEED = 1\n",
    "LEARNING_RATE = 0.0005\n",
    "NUM_EPOCHS = 50\n",
    "BATCH_SIZE = 128\n",
    "IMG_SIZE = 28\n",
    "\n",
    "# Architecture\n",
    "NUM_CLASSES = 10"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MNIST Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Image batch dimensions: torch.Size([128, 1, 28, 28])\n",
      "Image label dimensions: torch.Size([128])\n"
     ]
    }
   ],
   "source": [
    "train_indices = torch.arange(0, 59000)\n",
    "valid_indices = torch.arange(59000, 60000)\n",
    "\n",
    "custom_transform = transforms.Compose([transforms.ToTensor()])\n",
    "\n",
    "\n",
    "train_and_valid = datasets.MNIST(root='data', \n",
    "                                 train=True, \n",
    "                                 transform=custom_transform,\n",
    "                                 download=True)\n",
    "\n",
    "test_dataset = datasets.MNIST(root='data', \n",
    "                              train=False, \n",
    "                              transform=custom_transform,\n",
    "                              download=True)\n",
    "\n",
    "train_dataset = Subset(train_and_valid, train_indices)\n",
    "valid_dataset = Subset(train_and_valid, valid_indices)\n",
    "\n",
    "train_loader = DataLoader(dataset=train_dataset, \n",
    "                          batch_size=BATCH_SIZE,\n",
    "                          num_workers=4,\n",
    "                          shuffle=True)\n",
    "\n",
    "valid_loader = DataLoader(dataset=valid_dataset, \n",
    "                          batch_size=BATCH_SIZE,\n",
    "                          num_workers=4,\n",
    "                          shuffle=False)\n",
    "\n",
    "test_loader = DataLoader(dataset=test_dataset, \n",
    "                         batch_size=BATCH_SIZE,\n",
    "                         num_workers=4,\n",
    "                         shuffle=False)\n",
    "\n",
    "# Checking the dataset\n",
    "for images, labels in train_loader:  \n",
    "    print('Image batch dimensions:', images.shape)\n",
    "    print('Image label dimensions:', labels.shape)\n",
    "    break"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "##########################\n",
    "### MODEL\n",
    "##########################\n",
    "\n",
    "\n",
    "def make_coordinate_array(img_size, out_size=4):\n",
    "    \n",
    "    ### Make 2D coordinate array (for MNIST: 784x2)\n",
    "    n_rows = img_size * img_size\n",
    "    col, row = np.meshgrid(np.arange(img_size), np.arange(img_size))\n",
    "    coord = np.stack((col, row), axis=2).reshape(-1, 2)\n",
    "    coord = (coord - np.mean(coord, axis=0)) / (np.std(coord, axis=0) + 1e-5)\n",
    "    coord = torch.from_numpy(coord).float()\n",
    "    \n",
    "    ### Reshape to [N, N, out_size]\n",
    "    coord = torch.cat((coord.unsqueeze(0).repeat(n_rows, 1,  int(out_size/2-1)),\n",
    "                            coord.unsqueeze(1).repeat(1, n_rows, 1)), dim=2)\n",
    "    \n",
    "    \n",
    "    return coord\n",
    "\n",
    "        \n",
    "\n",
    "class GraphNet(nn.Module):\n",
    "    def __init__(self, img_size=28, coord_features=4, num_classes=10):\n",
    "        super(GraphNet, self).__init__()\n",
    "        \n",
    "        n_rows = img_size**2\n",
    "        self.fc = nn.Linear(n_rows, num_classes, bias=False)\n",
    "\n",
    "        coord = make_coordinate_array(img_size, coord_features)\n",
    "        self.register_buffer('coord', coord)\n",
    "        \n",
    "        ##########\n",
    "        # Edge Predictor\n",
    "        self.pred_edge_fc = nn.Sequential(nn.Linear(coord_features, 32), # coord -> hidden\n",
    "                                          nn.ReLU(),\n",
    "                                          nn.Linear(32, 1), # hidden -> edge\n",
    "                                          nn.Tanh())\n",
    "        \n",
    "\n",
    "        \n",
    "\n",
    "    def forward(self, x):\n",
    "        B = x.size(0)\n",
    "        \n",
    "        ### Predict edges\n",
    "        self.A = self.pred_edge_fc(self.coord).squeeze()\n",
    "\n",
    "        ### Reshape Adjacency Matrix\n",
    "        # [N, N] Adj. matrix -> [1, N, N] Adj tensor where N = HxW\n",
    "        A_tensor = self.A.unsqueeze(0)\n",
    "        # [1, N, N] Adj tensor -> [B, N, N] tensor\n",
    "        A_tensor = self.A.expand(B, -1, -1)\n",
    "        \n",
    "        ### Reshape inputs\n",
    "        # [B, C, H, W] => [B, H*W, 1]\n",
    "        x_reshape = x.view(B, -1, 1)\n",
    "        \n",
    "        # bmm = batch matrix product to sum the neighbor features\n",
    "        # Input: [B, N, N] x [B, N, 1]\n",
    "        # Output: [B, N]\n",
    "        avg_neighbor_features = (torch.bmm(A_tensor, x_reshape).view(B, -1))\n",
    "\n",
    "        logits = self.fc(avg_neighbor_features)\n",
    "        probas = F.softmax(logits, dim=1)\n",
    "        return logits, probas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.manual_seed(RANDOM_SEED)\n",
    "model = GraphNet(img_size=IMG_SIZE, num_classes=NUM_CLASSES)\n",
    "\n",
    "model = model.to(DEVICE)\n",
    "\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=LEARNING_RATE)  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 001/050 | Batch 000/461 | Cost: 24.2727\n",
      "Epoch: 001/050 | Batch 150/461 | Cost: 2.2706\n",
      "Epoch: 001/050 | Batch 300/461 | Cost: 1.8713\n",
      "Epoch: 001/050 | Batch 450/461 | Cost: 1.5048\n",
      "Epoch: 001/050\n",
      "Train ACC: 50.39 | Validation ACC: 54.80\n",
      "Time elapsed: 0.25 min\n",
      "Epoch: 002/050 | Batch 000/461 | Cost: 1.4445\n",
      "Epoch: 002/050 | Batch 150/461 | Cost: 1.3288\n",
      "Epoch: 002/050 | Batch 300/461 | Cost: 1.1868\n",
      "Epoch: 002/050 | Batch 450/461 | Cost: 1.2040\n",
      "Epoch: 002/050\n",
      "Train ACC: 67.68 | Validation ACC: 71.40\n",
      "Time elapsed: 0.49 min\n",
      "Epoch: 003/050 | Batch 000/461 | Cost: 1.2128\n",
      "Epoch: 003/050 | Batch 150/461 | Cost: 0.9953\n",
      "Epoch: 003/050 | Batch 300/461 | Cost: 0.9818\n",
      "Epoch: 003/050 | Batch 450/461 | Cost: 1.0487\n",
      "Epoch: 003/050\n",
      "Train ACC: 68.09 | Validation ACC: 73.40\n",
      "Time elapsed: 0.74 min\n",
      "Epoch: 004/050 | Batch 000/461 | Cost: 1.0444\n",
      "Epoch: 004/050 | Batch 150/461 | Cost: 0.9064\n",
      "Epoch: 004/050 | Batch 300/461 | Cost: 0.9152\n",
      "Epoch: 004/050 | Batch 450/461 | Cost: 0.7396\n",
      "Epoch: 004/050\n",
      "Train ACC: 76.10 | Validation ACC: 80.20\n",
      "Time elapsed: 0.98 min\n",
      "Epoch: 005/050 | Batch 000/461 | Cost: 0.7698\n",
      "Epoch: 005/050 | Batch 150/461 | Cost: 0.8356\n",
      "Epoch: 005/050 | Batch 300/461 | Cost: 0.6544\n",
      "Epoch: 005/050 | Batch 450/461 | Cost: 0.8700\n",
      "Epoch: 005/050\n",
      "Train ACC: 80.30 | Validation ACC: 84.10\n",
      "Time elapsed: 1.22 min\n",
      "Epoch: 006/050 | Batch 000/461 | Cost: 0.6292\n",
      "Epoch: 006/050 | Batch 150/461 | Cost: 0.7779\n",
      "Epoch: 006/050 | Batch 300/461 | Cost: 0.5978\n",
      "Epoch: 006/050 | Batch 450/461 | Cost: 0.6260\n",
      "Epoch: 006/050\n",
      "Train ACC: 82.00 | Validation ACC: 85.60\n",
      "Time elapsed: 1.46 min\n",
      "Epoch: 007/050 | Batch 000/461 | Cost: 0.7172\n",
      "Epoch: 007/050 | Batch 150/461 | Cost: 0.6444\n",
      "Epoch: 007/050 | Batch 300/461 | Cost: 0.5620\n",
      "Epoch: 007/050 | Batch 450/461 | Cost: 0.5314\n",
      "Epoch: 007/050\n",
      "Train ACC: 82.90 | Validation ACC: 86.20\n",
      "Time elapsed: 1.71 min\n",
      "Epoch: 008/050 | Batch 000/461 | Cost: 0.6211\n",
      "Epoch: 008/050 | Batch 150/461 | Cost: 0.5004\n",
      "Epoch: 008/050 | Batch 300/461 | Cost: 0.5274\n",
      "Epoch: 008/050 | Batch 450/461 | Cost: 0.6611\n",
      "Epoch: 008/050\n",
      "Train ACC: 84.86 | Validation ACC: 87.90\n",
      "Time elapsed: 1.95 min\n",
      "Epoch: 009/050 | Batch 000/461 | Cost: 0.4017\n",
      "Epoch: 009/050 | Batch 150/461 | Cost: 0.7080\n",
      "Epoch: 009/050 | Batch 300/461 | Cost: 0.4298\n",
      "Epoch: 009/050 | Batch 450/461 | Cost: 0.4516\n",
      "Epoch: 009/050\n",
      "Train ACC: 85.89 | Validation ACC: 90.20\n",
      "Time elapsed: 2.19 min\n",
      "Epoch: 010/050 | Batch 000/461 | Cost: 0.4571\n",
      "Epoch: 010/050 | Batch 150/461 | Cost: 0.4976\n",
      "Epoch: 010/050 | Batch 300/461 | Cost: 0.6208\n",
      "Epoch: 010/050 | Batch 450/461 | Cost: 0.3780\n",
      "Epoch: 010/050\n",
      "Train ACC: 85.84 | Validation ACC: 89.40\n",
      "Time elapsed: 2.43 min\n",
      "Epoch: 011/050 | Batch 000/461 | Cost: 0.5262\n",
      "Epoch: 011/050 | Batch 150/461 | Cost: 0.4255\n",
      "Epoch: 011/050 | Batch 300/461 | Cost: 0.3840\n",
      "Epoch: 011/050 | Batch 450/461 | Cost: 0.4941\n",
      "Epoch: 011/050\n",
      "Train ACC: 85.73 | Validation ACC: 90.40\n",
      "Time elapsed: 2.68 min\n",
      "Epoch: 012/050 | Batch 000/461 | Cost: 0.3425\n",
      "Epoch: 012/050 | Batch 150/461 | Cost: 0.5059\n",
      "Epoch: 012/050 | Batch 300/461 | Cost: 0.6590\n",
      "Epoch: 012/050 | Batch 450/461 | Cost: 0.5481\n",
      "Epoch: 012/050\n",
      "Train ACC: 87.50 | Validation ACC: 91.30\n",
      "Time elapsed: 2.92 min\n",
      "Epoch: 013/050 | Batch 000/461 | Cost: 0.6081\n",
      "Epoch: 013/050 | Batch 150/461 | Cost: 0.4584\n",
      "Epoch: 013/050 | Batch 300/461 | Cost: 0.2856\n",
      "Epoch: 013/050 | Batch 450/461 | Cost: 0.4324\n",
      "Epoch: 013/050\n",
      "Train ACC: 87.35 | Validation ACC: 91.20\n",
      "Time elapsed: 3.17 min\n",
      "Epoch: 014/050 | Batch 000/461 | Cost: 0.4685\n",
      "Epoch: 014/050 | Batch 150/461 | Cost: 0.4492\n",
      "Epoch: 014/050 | Batch 300/461 | Cost: 0.3913\n",
      "Epoch: 014/050 | Batch 450/461 | Cost: 0.5154\n",
      "Epoch: 014/050\n",
      "Train ACC: 86.71 | Validation ACC: 91.20\n",
      "Time elapsed: 3.41 min\n",
      "Epoch: 015/050 | Batch 000/461 | Cost: 0.4526\n",
      "Epoch: 015/050 | Batch 150/461 | Cost: 0.4834\n",
      "Epoch: 015/050 | Batch 300/461 | Cost: 0.5208\n",
      "Epoch: 015/050 | Batch 450/461 | Cost: 0.3536\n",
      "Epoch: 015/050\n",
      "Train ACC: 85.21 | Validation ACC: 89.50\n",
      "Time elapsed: 3.66 min\n",
      "Epoch: 016/050 | Batch 000/461 | Cost: 0.6614\n",
      "Epoch: 016/050 | Batch 150/461 | Cost: 0.3036\n",
      "Epoch: 016/050 | Batch 300/461 | Cost: 0.3766\n",
      "Epoch: 016/050 | Batch 450/461 | Cost: 0.4550\n",
      "Epoch: 016/050\n",
      "Train ACC: 86.97 | Validation ACC: 92.10\n",
      "Time elapsed: 3.91 min\n",
      "Epoch: 017/050 | Batch 000/461 | Cost: 0.6241\n",
      "Epoch: 017/050 | Batch 150/461 | Cost: 0.3934\n",
      "Epoch: 017/050 | Batch 300/461 | Cost: 0.4330\n",
      "Epoch: 017/050 | Batch 450/461 | Cost: 0.5914\n",
      "Epoch: 017/050\n",
      "Train ACC: 88.12 | Validation ACC: 91.60\n",
      "Time elapsed: 4.15 min\n",
      "Epoch: 018/050 | Batch 000/461 | Cost: 0.3769\n",
      "Epoch: 018/050 | Batch 150/461 | Cost: 0.4817\n",
      "Epoch: 018/050 | Batch 300/461 | Cost: 0.4103\n",
      "Epoch: 018/050 | Batch 450/461 | Cost: 0.3727\n",
      "Epoch: 018/050\n",
      "Train ACC: 86.58 | Validation ACC: 90.90\n",
      "Time elapsed: 4.40 min\n",
      "Epoch: 019/050 | Batch 000/461 | Cost: 0.4098\n",
      "Epoch: 019/050 | Batch 150/461 | Cost: 0.4435\n",
      "Epoch: 019/050 | Batch 300/461 | Cost: 0.2952\n",
      "Epoch: 019/050 | Batch 450/461 | Cost: 0.3328\n",
      "Epoch: 019/050\n",
      "Train ACC: 88.65 | Validation ACC: 92.00\n",
      "Time elapsed: 4.64 min\n",
      "Epoch: 020/050 | Batch 000/461 | Cost: 0.5363\n",
      "Epoch: 020/050 | Batch 150/461 | Cost: 0.3143\n",
      "Epoch: 020/050 | Batch 300/461 | Cost: 0.5186\n",
      "Epoch: 020/050 | Batch 450/461 | Cost: 0.3806\n",
      "Epoch: 020/050\n",
      "Train ACC: 88.95 | Validation ACC: 92.70\n",
      "Time elapsed: 4.89 min\n",
      "Epoch: 021/050 | Batch 000/461 | Cost: 0.3810\n",
      "Epoch: 021/050 | Batch 150/461 | Cost: 0.2470\n",
      "Epoch: 021/050 | Batch 300/461 | Cost: 0.6154\n",
      "Epoch: 021/050 | Batch 450/461 | Cost: 0.3651\n",
      "Epoch: 021/050\n",
      "Train ACC: 88.31 | Validation ACC: 92.40\n",
      "Time elapsed: 5.13 min\n",
      "Epoch: 022/050 | Batch 000/461 | Cost: 0.3704\n",
      "Epoch: 022/050 | Batch 150/461 | Cost: 0.4338\n",
      "Epoch: 022/050 | Batch 300/461 | Cost: 0.4197\n",
      "Epoch: 022/050 | Batch 450/461 | Cost: 0.3304\n",
      "Epoch: 022/050\n",
      "Train ACC: 88.62 | Validation ACC: 91.90\n",
      "Time elapsed: 5.31 min\n",
      "Epoch: 023/050 | Batch 000/461 | Cost: 0.2825\n",
      "Epoch: 023/050 | Batch 150/461 | Cost: 0.4302\n",
      "Epoch: 023/050 | Batch 300/461 | Cost: 0.4738\n",
      "Epoch: 023/050 | Batch 450/461 | Cost: 0.4362\n",
      "Epoch: 023/050\n",
      "Train ACC: 89.02 | Validation ACC: 92.80\n",
      "Time elapsed: 5.44 min\n",
      "Epoch: 024/050 | Batch 000/461 | Cost: 0.2097\n",
      "Epoch: 024/050 | Batch 150/461 | Cost: 0.4440\n",
      "Epoch: 024/050 | Batch 300/461 | Cost: 0.4467\n",
      "Epoch: 024/050 | Batch 450/461 | Cost: 0.2744\n",
      "Epoch: 024/050\n",
      "Train ACC: 88.82 | Validation ACC: 92.40\n",
      "Time elapsed: 5.57 min\n",
      "Epoch: 025/050 | Batch 000/461 | Cost: 0.2734\n",
      "Epoch: 025/050 | Batch 150/461 | Cost: 0.3980\n",
      "Epoch: 025/050 | Batch 300/461 | Cost: 0.4395\n",
      "Epoch: 025/050 | Batch 450/461 | Cost: 0.2336\n",
      "Epoch: 025/050\n",
      "Train ACC: 89.59 | Validation ACC: 93.90\n",
      "Time elapsed: 5.70 min\n",
      "Epoch: 026/050 | Batch 000/461 | Cost: 0.3138\n",
      "Epoch: 026/050 | Batch 150/461 | Cost: 0.3772\n",
      "Epoch: 026/050 | Batch 300/461 | Cost: 0.2955\n",
      "Epoch: 026/050 | Batch 450/461 | Cost: 0.3747\n",
      "Epoch: 026/050\n",
      "Train ACC: 88.71 | Validation ACC: 92.70\n",
      "Time elapsed: 5.82 min\n",
      "Epoch: 027/050 | Batch 000/461 | Cost: 0.4107\n",
      "Epoch: 027/050 | Batch 150/461 | Cost: 0.4375\n",
      "Epoch: 027/050 | Batch 300/461 | Cost: 0.3802\n",
      "Epoch: 027/050 | Batch 450/461 | Cost: 0.3240\n",
      "Epoch: 027/050\n",
      "Train ACC: 87.90 | Validation ACC: 91.60\n",
      "Time elapsed: 5.95 min\n",
      "Epoch: 028/050 | Batch 000/461 | Cost: 0.5124\n",
      "Epoch: 028/050 | Batch 150/461 | Cost: 0.4980\n",
      "Epoch: 028/050 | Batch 300/461 | Cost: 0.3937\n",
      "Epoch: 028/050 | Batch 450/461 | Cost: 0.2704\n",
      "Epoch: 028/050\n",
      "Train ACC: 89.08 | Validation ACC: 92.30\n",
      "Time elapsed: 6.08 min\n",
      "Epoch: 029/050 | Batch 000/461 | Cost: 0.3328\n",
      "Epoch: 029/050 | Batch 150/461 | Cost: 0.3022\n",
      "Epoch: 029/050 | Batch 300/461 | Cost: 0.3222\n",
      "Epoch: 029/050 | Batch 450/461 | Cost: 0.3084\n",
      "Epoch: 029/050\n",
      "Train ACC: 89.30 | Validation ACC: 93.90\n",
      "Time elapsed: 6.21 min\n",
      "Epoch: 030/050 | Batch 000/461 | Cost: 0.4667\n",
      "Epoch: 030/050 | Batch 150/461 | Cost: 0.3290\n",
      "Epoch: 030/050 | Batch 300/461 | Cost: 0.3261\n",
      "Epoch: 030/050 | Batch 450/461 | Cost: 0.3347\n",
      "Epoch: 030/050\n",
      "Train ACC: 89.17 | Validation ACC: 93.60\n",
      "Time elapsed: 6.33 min\n",
      "Epoch: 031/050 | Batch 000/461 | Cost: 0.3486\n",
      "Epoch: 031/050 | Batch 150/461 | Cost: 0.2426\n",
      "Epoch: 031/050 | Batch 300/461 | Cost: 0.2748\n",
      "Epoch: 031/050 | Batch 450/461 | Cost: 0.2072\n",
      "Epoch: 031/050\n",
      "Train ACC: 89.17 | Validation ACC: 93.20\n",
      "Time elapsed: 6.46 min\n",
      "Epoch: 032/050 | Batch 000/461 | Cost: 0.3423\n",
      "Epoch: 032/050 | Batch 150/461 | Cost: 0.4924\n",
      "Epoch: 032/050 | Batch 300/461 | Cost: 0.4072\n",
      "Epoch: 032/050 | Batch 450/461 | Cost: 0.3611\n",
      "Epoch: 032/050\n",
      "Train ACC: 89.83 | Validation ACC: 94.30\n",
      "Time elapsed: 6.59 min\n",
      "Epoch: 033/050 | Batch 000/461 | Cost: 0.2461\n",
      "Epoch: 033/050 | Batch 150/461 | Cost: 0.2343\n",
      "Epoch: 033/050 | Batch 300/461 | Cost: 0.2891\n",
      "Epoch: 033/050 | Batch 450/461 | Cost: 0.3772\n",
      "Epoch: 033/050\n",
      "Train ACC: 88.81 | Validation ACC: 92.40\n",
      "Time elapsed: 6.72 min\n",
      "Epoch: 034/050 | Batch 000/461 | Cost: 0.3052\n",
      "Epoch: 034/050 | Batch 150/461 | Cost: 0.5129\n",
      "Epoch: 034/050 | Batch 300/461 | Cost: 0.3810\n",
      "Epoch: 034/050 | Batch 450/461 | Cost: 0.2906\n",
      "Epoch: 034/050\n",
      "Train ACC: 89.34 | Validation ACC: 93.10\n",
      "Time elapsed: 6.85 min\n",
      "Epoch: 035/050 | Batch 000/461 | Cost: 0.3604\n",
      "Epoch: 035/050 | Batch 150/461 | Cost: 0.3832\n",
      "Epoch: 035/050 | Batch 300/461 | Cost: 0.3632\n",
      "Epoch: 035/050 | Batch 450/461 | Cost: 0.3345\n",
      "Epoch: 035/050\n",
      "Train ACC: 89.74 | Validation ACC: 93.10\n",
      "Time elapsed: 6.98 min\n",
      "Epoch: 036/050 | Batch 000/461 | Cost: 0.3382\n",
      "Epoch: 036/050 | Batch 150/461 | Cost: 0.3754\n",
      "Epoch: 036/050 | Batch 300/461 | Cost: 0.4120\n",
      "Epoch: 036/050 | Batch 450/461 | Cost: 0.4710\n",
      "Epoch: 036/050\n",
      "Train ACC: 89.10 | Validation ACC: 93.90\n",
      "Time elapsed: 7.10 min\n",
      "Epoch: 037/050 | Batch 000/461 | Cost: 0.4466\n",
      "Epoch: 037/050 | Batch 150/461 | Cost: 0.3427\n",
      "Epoch: 037/050 | Batch 300/461 | Cost: 0.3301\n",
      "Epoch: 037/050 | Batch 450/461 | Cost: 0.4110\n",
      "Epoch: 037/050\n",
      "Train ACC: 89.95 | Validation ACC: 93.90\n",
      "Time elapsed: 7.23 min\n",
      "Epoch: 038/050 | Batch 000/461 | Cost: 0.2470\n",
      "Epoch: 038/050 | Batch 150/461 | Cost: 0.4719\n",
      "Epoch: 038/050 | Batch 300/461 | Cost: 0.3253\n",
      "Epoch: 038/050 | Batch 450/461 | Cost: 0.4324\n",
      "Epoch: 038/050\n",
      "Train ACC: 89.35 | Validation ACC: 93.50\n",
      "Time elapsed: 7.36 min\n",
      "Epoch: 039/050 | Batch 000/461 | Cost: 0.3058\n",
      "Epoch: 039/050 | Batch 150/461 | Cost: 0.4755\n",
      "Epoch: 039/050 | Batch 300/461 | Cost: 0.2981\n",
      "Epoch: 039/050 | Batch 450/461 | Cost: 0.4293\n",
      "Epoch: 039/050\n",
      "Train ACC: 89.51 | Validation ACC: 92.90\n",
      "Time elapsed: 7.48 min\n",
      "Epoch: 040/050 | Batch 000/461 | Cost: 0.3378\n",
      "Epoch: 040/050 | Batch 150/461 | Cost: 0.5137\n",
      "Epoch: 040/050 | Batch 300/461 | Cost: 0.2680\n",
      "Epoch: 040/050 | Batch 450/461 | Cost: 0.3397\n",
      "Epoch: 040/050\n",
      "Train ACC: 90.01 | Validation ACC: 93.70\n",
      "Time elapsed: 7.61 min\n",
      "Epoch: 041/050 | Batch 000/461 | Cost: 0.2766\n",
      "Epoch: 041/050 | Batch 150/461 | Cost: 0.2959\n",
      "Epoch: 041/050 | Batch 300/461 | Cost: 0.1930\n",
      "Epoch: 041/050 | Batch 450/461 | Cost: 0.3735\n",
      "Epoch: 041/050\n",
      "Train ACC: 89.45 | Validation ACC: 93.60\n",
      "Time elapsed: 7.74 min\n",
      "Epoch: 042/050 | Batch 000/461 | Cost: 0.2694\n",
      "Epoch: 042/050 | Batch 150/461 | Cost: 0.3575\n",
      "Epoch: 042/050 | Batch 300/461 | Cost: 0.4267\n",
      "Epoch: 042/050 | Batch 450/461 | Cost: 0.3332\n",
      "Epoch: 042/050\n",
      "Train ACC: 89.96 | Validation ACC: 93.30\n",
      "Time elapsed: 7.86 min\n",
      "Epoch: 043/050 | Batch 000/461 | Cost: 0.2288\n",
      "Epoch: 043/050 | Batch 150/461 | Cost: 0.4260\n",
      "Epoch: 043/050 | Batch 300/461 | Cost: 0.2835\n",
      "Epoch: 043/050 | Batch 450/461 | Cost: 0.2882\n",
      "Epoch: 043/050\n",
      "Train ACC: 89.91 | Validation ACC: 93.40\n",
      "Time elapsed: 7.99 min\n",
      "Epoch: 044/050 | Batch 000/461 | Cost: 0.3211\n",
      "Epoch: 044/050 | Batch 150/461 | Cost: 0.3061\n",
      "Epoch: 044/050 | Batch 300/461 | Cost: 0.3137\n",
      "Epoch: 044/050 | Batch 450/461 | Cost: 0.2978\n",
      "Epoch: 044/050\n",
      "Train ACC: 89.63 | Validation ACC: 94.30\n",
      "Time elapsed: 8.12 min\n",
      "Epoch: 045/050 | Batch 000/461 | Cost: 0.2325\n",
      "Epoch: 045/050 | Batch 150/461 | Cost: 0.3013\n",
      "Epoch: 045/050 | Batch 300/461 | Cost: 0.3732\n",
      "Epoch: 045/050 | Batch 450/461 | Cost: 0.3229\n",
      "Epoch: 045/050\n",
      "Train ACC: 90.00 | Validation ACC: 93.80\n",
      "Time elapsed: 8.25 min\n",
      "Epoch: 046/050 | Batch 000/461 | Cost: 0.2521\n",
      "Epoch: 046/050 | Batch 150/461 | Cost: 0.4440\n",
      "Epoch: 046/050 | Batch 300/461 | Cost: 0.3420\n",
      "Epoch: 046/050 | Batch 450/461 | Cost: 0.4288\n",
      "Epoch: 046/050\n",
      "Train ACC: 89.97 | Validation ACC: 93.40\n",
      "Time elapsed: 8.38 min\n",
      "Epoch: 047/050 | Batch 000/461 | Cost: 0.4605\n",
      "Epoch: 047/050 | Batch 150/461 | Cost: 0.3261\n",
      "Epoch: 047/050 | Batch 300/461 | Cost: 0.4493\n",
      "Epoch: 047/050 | Batch 450/461 | Cost: 0.4902\n",
      "Epoch: 047/050\n",
      "Train ACC: 89.60 | Validation ACC: 93.70\n",
      "Time elapsed: 8.50 min\n",
      "Epoch: 048/050 | Batch 000/461 | Cost: 0.4136\n",
      "Epoch: 048/050 | Batch 150/461 | Cost: 0.2952\n",
      "Epoch: 048/050 | Batch 300/461 | Cost: 0.4784\n",
      "Epoch: 048/050 | Batch 450/461 | Cost: 0.3044\n",
      "Epoch: 048/050\n",
      "Train ACC: 90.15 | Validation ACC: 94.60\n",
      "Time elapsed: 8.63 min\n",
      "Epoch: 049/050 | Batch 000/461 | Cost: 0.3802\n",
      "Epoch: 049/050 | Batch 150/461 | Cost: 0.4018\n",
      "Epoch: 049/050 | Batch 300/461 | Cost: 0.3197\n",
      "Epoch: 049/050 | Batch 450/461 | Cost: 0.4157\n",
      "Epoch: 049/050\n",
      "Train ACC: 89.91 | Validation ACC: 93.70\n",
      "Time elapsed: 8.76 min\n",
      "Epoch: 050/050 | Batch 000/461 | Cost: 0.4057\n",
      "Epoch: 050/050 | Batch 150/461 | Cost: 0.3687\n",
      "Epoch: 050/050 | Batch 300/461 | Cost: 0.3552\n",
      "Epoch: 050/050 | Batch 450/461 | Cost: 0.2707\n",
      "Epoch: 050/050\n",
      "Train ACC: 89.91 | Validation ACC: 93.00\n",
      "Time elapsed: 8.88 min\n",
      "Total Training Time: 8.88 min\n"
     ]
    }
   ],
   "source": [
    "def compute_acc(model, data_loader, device):\n",
    "    correct_pred, num_examples = 0, 0\n",
    "    for features, targets in data_loader:\n",
    "        features = features.to(device)\n",
    "        targets = targets.to(device)\n",
    "        logits, probas = model(features)\n",
    "        _, predicted_labels = torch.max(probas, 1)\n",
    "        num_examples += targets.size(0)\n",
    "        correct_pred += (predicted_labels == targets).sum()\n",
    "    return correct_pred.float()/num_examples * 100\n",
    "    \n",
    "\n",
    "start_time = time.time()\n",
    "\n",
    "cost_list = []\n",
    "train_acc_list, valid_acc_list = [], []\n",
    "\n",
    "\n",
    "for epoch in range(NUM_EPOCHS):\n",
    "    \n",
    "    model.train()\n",
    "    for batch_idx, (features, targets) in enumerate(train_loader):\n",
    "        \n",
    "        features = features.to(DEVICE)\n",
    "        targets = targets.to(DEVICE)\n",
    "            \n",
    "        ### FORWARD AND BACK PROP\n",
    "        logits, probas = model(features)\n",
    "        cost = F.cross_entropy(logits, targets)\n",
    "        optimizer.zero_grad()\n",
    "        \n",
    "        cost.backward()\n",
    "        \n",
    "        ### UPDATE MODEL PARAMETERS\n",
    "        optimizer.step()\n",
    "        \n",
    "        #################################################\n",
    "        ### CODE ONLY FOR LOGGING BEYOND THIS POINT\n",
    "        ################################################\n",
    "        cost_list.append(cost.item())\n",
    "        if not batch_idx % 150:\n",
    "            print (f'Epoch: {epoch+1:03d}/{NUM_EPOCHS:03d} | '\n",
    "                   f'Batch {batch_idx:03d}/{len(train_loader):03d} |' \n",
    "                   f' Cost: {cost:.4f}')\n",
    "\n",
    "        \n",
    "\n",
    "    model.eval()\n",
    "    with torch.set_grad_enabled(False): # save memory during inference\n",
    "        \n",
    "        train_acc = compute_acc(model, train_loader, device=DEVICE)\n",
    "        valid_acc = compute_acc(model, valid_loader, device=DEVICE)\n",
    "        \n",
    "        print(f'Epoch: {epoch+1:03d}/{NUM_EPOCHS:03d}\\n'\n",
    "              f'Train ACC: {train_acc:.2f} | Validation ACC: {valid_acc:.2f}')\n",
    "        \n",
    "        train_acc_list.append(train_acc)\n",
    "        valid_acc_list.append(valid_acc)\n",
    "        \n",
    "    elapsed = (time.time() - start_time)/60\n",
    "    print(f'Time elapsed: {elapsed:.2f} min')\n",
    "  \n",
    "elapsed = (time.time() - start_time)/60\n",
    "print(f'Total Training Time: {elapsed:.2f} min')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# last adjacency matrix\n",
    "\n",
    "plt.imshow(model.A.to('cpu'));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(cost_list, label='Minibatch cost')\n",
    "plt.plot(np.convolve(cost_list, \n",
    "                     np.ones(200,)/200, mode='valid'), \n",
    "         label='Running average')\n",
    "\n",
    "plt.ylabel('Cross Entropy')\n",
    "plt.xlabel('Iteration')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.arange(1, NUM_EPOCHS+1), train_acc_list, label='Training')\n",
    "plt.plot(np.arange(1, NUM_EPOCHS+1), valid_acc_list, label='Validation')\n",
    "\n",
    "plt.xlabel('Epoch')\n",
    "plt.ylabel('Accuracy')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Validation ACC: 93.00%\n",
      "Test ACC: 90.36%\n"
     ]
    }
   ],
   "source": [
    "with torch.set_grad_enabled(False):\n",
    "    test_acc = compute_acc(model=model,\n",
    "                           data_loader=test_loader,\n",
    "                           device=DEVICE)\n",
    "    \n",
    "    valid_acc = compute_acc(model=model,\n",
    "                            data_loader=valid_loader,\n",
    "                            device=DEVICE)\n",
    "    \n",
    "\n",
    "print(f'Validation ACC: {valid_acc:.2f}%')\n",
    "print(f'Test ACC: {test_acc:.2f}%')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torchvision 0.4.0a0+6b959ee\n",
      "matplotlib  3.1.0\n",
      "torch       1.2.0\n",
      "numpy       1.16.4\n",
      "\n"
     ]
    }
   ],
   "source": [
    "%watermark -iv"
   ]
  }
 ],
 "metadata": {
  "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": "3.7.3"
  },
  "toc": {
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}